Economists are still arguing over the small impact of the changes in the inflation and disagreeing vehemently over the big upgrade brought to the GDP.
Recently, the Central Statistical Office (CSO) announced it was moving to a more internationally-accepted method of calculating GDP – from factor cost earlier to market prices now. The difference between the two is that the latter also accounts for subsidies and indirect taxes.
Also read: Why govt is wary of focusing too much on new GDP nos
Simultaneously, the CSO moved to a newer base, from 2005-06 to 2011-12, which typically done to reflect structural changes in an economy.
Both these changes mean the Indian economy grew at 6.9 percent last fiscal (2013-14) and is seen growing at 7.4 percent this fiscal. Under the old methodology, the two numbers stood at 4.8 percent and 5.5 percent, respectively.
Also read: Why economists are raising skeptical eye over CSO’s GDP nos
In an interview to CNBC-TV18, TCA Anant, chief statistician, CSO says the new method of calculating the GDP captures the underlying value added to goods and services better than the older series.
Furthermore, he says the country’s industrial output numbers and the employment data gives an idea of the value added momentum and hence the CSO is planning to announce employment data to aid the RBI governor Raghuram Rajan in his future monetary policies.
“One will find discussion on monetary policy in other countries will include discussion of the employment rate and we will have to move in the same direction. We are separately planning to bring out- and that is also a very complex exercise, regular figures on employment, we hope that will get started soon but until then this understanding of the structure of GDP will be very important to gauge what is the appropriate approach to policy,” he adds.
Below is the edited transcript of the interview.
Q: Can you tell us what is the average consumer price index (CPI) for 2014 according to the new series and the old series?
A: I am sure it is there but I do not think I have calculated it, I did not recall reading it in any of the things but it is approximate answer is about half a percent difference between the old series calculation and the new series calculations.
At any given point largely on account of two factors changes in the waiting diagram which is slammer element and changes in one element in the calculation technique of using geometric mean to arithmetic mean the average difference is about half a percent. However it varies on, in some months it is slightly higher some months it is slightly lower.
Q: If the total impact of both weights and geometric mean is half a percent what exactly is the impact of just using geometric mean instead of arithmetic mean?
A: I told you the total is about half a percent or may be 0.6-0.7 of which the weight shifts would be about 0.1 or thereabout 0.1 or 0.2 geometric mean will be above the rest that is orders of magnitude.
Q: Why did you choose 2011-2012 as the base year this is for the GDP? Some may argue that it was not a quite a normal year.
A: There is a problem of when you choose the year you have to remember you have choose it pretty close to the time when the year is still in progress. Let me explain what the criteria were. The basic recommendation is that the GDP series the base should be revised frequently and in approximately once every 5 years.
As per this recommendation we should have been trying to revise GDP in 2009-2010. Now when the year 2009-2010 issue was being discussed the feeling was this was the year immediately following the crisis. In that sense the feeling was that this may not be appropriate and that year was also marked with very large agricultural crisis in terms of failure of rains. For both of these reasons they felt they wanted to defer it. Now deferring it is a complex because then we have to schedule a number of activities in advance. So we were able to schedule household consumer expenditure and employment survey which is a very important element in GDP calculations for 2011-2012.
In fact when people saw the 2011-2012 consumer expenditure survey lots of people were asking me why did you do a poverty calculation again so quickly. The answer is the, these things have to be planned in advance and executed and they are expensive exercises so once they have been done, the base year would go to 2011-2012. In a sense the impact of the crisis by some arguments continue and would be therefore we can not defer GDP base revision for so long because then the underlying estimates become more and more unreliable.
Q: Can you detect any peculiarities about 2011-2012 itself that economist and analyst ought to be aware of?
A: So far as what are the underlying peculiarities of the year it is very hard for me to say at the moment but let me try to take a stab. If you look at 2011-2012 for example you will find that international commodity prices were at the peak during 2011-2012 and subsequently started crashing. Now that would have affected domestic value added and in past we may have had the situation where domestic value added in 2011-2012 may have been depressed. Because our inputs were being expensive, at least an imported inputs would have been expensive and our domestic prices to some extent in some cases were being regulated or not rising as fast. There was a phase that may have happened so how did 2011-2012 factor in these are things which analyst will have to factor in to look at very carefully. As a statistician what I can say is that the new GDP series captures underlying value added better than the older series.
Q: Your nominal growth for FY14 and FY15 are similar to the old series, about 113 lakh crore and about 126 lakh crore. However there is a big difference in the real rate of growth. Clearly your deflator is making a big difference. Some economists I spoke to, worry that you have used the wholesale price index (WPI) to deflate the services entries and that is why you are getting an exaggerated deflator?
A: We use WPI and CPI both in different segments. However the difference is the weighting structure of different segments has changed and the rise in manufacturing which is almost exclusively deflated by WPI has meant that the over all deflator has fallen. So you have to rework the calculations using the changes in rates and you will find a large part of the explanation lies there.
Q: But CPI also has services like education, medical, transportation. So did you use CPI deflators there?
A: Which index gets deflated with which is the matter of detail. You will have to go through the entire, they have a classification category where they look at it and you will see what the appropriate deflator but services to the extent they relate to trade, construction it is not clear that the deflator should be CPI.
Q: So do I understand you right you deflated transportation, education, medical and such services with CPI and trade and such services with WPI?
A: Fair enough, now transport not sure but personal services yes, government administration yes. So that is why I am saying that you look at the compositional structure, the change in weights and then work it out. It is a derived answer.
Q: I just want to clarify you deflated some parts of services using WPI and some parts using CPI?
A: As was done in the past. In that there is no change. We used to, earlier also deflate trade and certain elements with WPI.
Q: Central Statistics Office (CSO) has not given us any back projection of GDP and which is why we are not able to understand whether 6.9 percent is a good growth number or a bad growth number.I mean if the years 2004-2008 in the new series showed a growth of say 11-12 percent then we can understand why even this 6.9 percent feels like recession. When are we going to get that back projected data?
A: The back series is a problem and I will explain the principal element of the problem. This series incorporates tow major types of changes- one of them is in manufacturing where we are using corporate accounts much more aggressively and the data base which we are being able to tap for this purpose is the Ministry of Corporate Affairs database.
Now this database does not extend going backwards. What you have is the best of information which you will have going back ward is the RBI sample but remember there is a huge difference between the two. RBI sample is about 2000 companies and that is at the upper end of the scale. The MCA database covers virtually all companies listed and unlisted as well.
Now a careful back series which would be analytically satisfactory to an analyst is not going to be that straightforward. Rules of thumb maybe worked out which is what we hope to do in the next coming months as well but you have some indicative idea of the fact that this series is showing growth better.
Part of the reason why it is showing growth better is because in manufacturing, we have moved from an establishment to an enterprise base. So, in effect we are picking up one element of valuated growth which was earlier not being picked up in the old series. That is the improvement having arisen out of improved corporate performance, bottom line performance, not sales or production performance.
There has been academic writings which has been showing that corporates through a variety of means of belt-tightening, restructuring, efficiency gains etc have improved their bottom lines or maintained their bottom lines even when the top line was dropping off.
How long will this be sustained it is not so easy to say. How far in the past does it go again not easy to say, you will have to go back and read it, the corporate studies to work this out. That can give you some guidance as to what would have been the growth going backwards. As I said, we are in the process of doing it but because we are using a new data for which an analog does not exist, the challenge is enormous. Now just because of this you cannot use it because it obviously captures current reality much better.
Q: No, I am not saying that at all. I am just trying to get a historical perspective of this 6.9 and 7.5 percent GDP data. You intimately know this data all the way back from 2004-2005, even if you haven’t put anything on the website, can you give me any rough back calculation that you may have?
A: At the moment we have not done a back calculation. The back calculation will take some time and let me point out to everybody who looks at it, even in the previous times that base division were done, back calculations become available with a lag. They are complex exercises and you can’t begin the exercise until first the new series is put in place. So, that exercise is underway but people will have to in the meantime- we have given comparative figures, till then judgemental rules can be applied.
Q: What about the back series for the CPI, can we get that earlier?
A: CPI back series we have already released a certain extent with this and we will probably-because the CPI only goes back to 2010. There is not much of an issue, we will be giving the back series quickly but the GDP will be a problem. I do not have the timeline at the moment with me; we will put it up in the website when we have figured it out.
Q: Actually for the CPI you have given only one year back data, that is through the link factors. When can we expect the CPI back data, say in three months, six months?
A: Let’s try.
Q: When you give a tall GDP number like 6.9 percent or seven and a half percent growth, how should a monetary policy maker understand this? Should he believe that there is no output gap left in the economy? That assumption somehow does not square with the facts.
A: GDP any value added series or GDP in particular can be decomposed into two elements. A quantum index or a quantum level and a value added level. Any change in value added will have a quantum side quantity side and a value side. Now, India does not have a lot of quantum indices but we do have the index of industrial production for all its limitations and the annual survey of industries which will also give you value of output as well as employment.
These give you an idea of what the underlying quantum momentum in the economy has been. The other side is giving you what the value added momentum is. Now, what you need to understand is issue of capacity, overheating or whatever is linked to growths in quantum- now this form of analysis by the way is not new, other macro-economists in other countries do in fact do this, which is why you will find discussion on monetary policy in other countries will include discussion of the employment rate and we will have to move in the same direction. We are separately planning to bring out- and that is also a very complex exercise, regular figures on employment, we hope that will get started soon but until then this understanding of the structure of GDP will be very important to gauge what is the appropriate approach to policy.
Q: You are going to give us employment data soon, by when will we start getting it?
A: We are working; a survey based exercise is very time consuming and requires a lot of logistics to be worked out and we are working on it. At the moment we are hoping that we will be able to start the service from July buy data flow will obviously take much longer, probably a year before it can get stabilized and start coming out as regular numbers.
Q: Do you already have the back projections up to 2010, can you share something?
A: Yesterday we did a slide presentation where we gave the old and new series graphs and so on. That’s why I am saying the numbers are there. These two series in broad movement are virtually identical so, you can actually see that they are parallel. Whatever difference is there-It is not a big difference. The geometric mean to arithmetic mean is a technical difference but once you built it in the series tracks each other fairly closely. I believe Press Information Bureau (PIB) had said that that slides are also there in their website or we will be putting it up on our website, people can take a look at them.