‘Degree Of Data Openness In India Is Much Higher Than Most Countries In World’
As the country loses its ‘fastest growing economy’ tag to China, the man at the helm of India’s data crunching, shares his views. More importantly, TCA Anant, Chief Statistician of India speaks on the reliability of those numbers as the opponents raise question on India’s data authenticity. Excerpts:
Are you satisfied with the growth numbers that have come, to begin with? Your views on it.
Satisfaction is very different qualitatively from what many other agencies, departments and ministries would view. I take satisfaction from the numbers based on how much information was available to us and how we have been able to compile it. So from that view point, the answer is yes.
I am completely satisfied because we were in a position to incorporate all the information available at that time. I am also satisfied because we had completed the major task of putting in the two data series which are a part of the data compilation - WPI and IIP. This was the major statistical challenge.
There is a second way to look at satisfaction, on how do you think the numbers are. That is much more difficult and challenging to answer. I think if you look at it from the medium term perspective, over the last 4-5 years, these numbers are quite good and reflect how the economy is distinctly improving.
But if you look from the view point of where we should be headed and what needs to be done, there are obviously many more things that need to be done. We would like to reach even higher growth rates.
There has been a major decline in some of those numbers and demonetisation has been attributed as one of the reasons?
Economic data and time series is subjected to a number of trends, which have varying levels of impacts. Some are short term trends depending on the seasonal factors and some are long term trends. In the GDP numbers, there is a mixture of both, at work. Apart from this there are global factors, one of them being the commodity prices, which have seen a major crash 2 years back. They are now working themselves back. The consequences of that is clearly reflected in our GDP, in a variety of ways.
You mentioned that it is difficult to measure the impact of demonetisation in short term, citing the unavailability of more robust data. So do you think there should be an authentic way to measure these kinds of impacts or is there a way possible at all?
There is a way of course. Econometricians have techniques by which they look for impact of events on time series. But it requires them to work with time series data, both before and after the event of a reasonable length, to filter out various cyclical and time trends.
In these circumstances, working with an event like demonetisation and coming out after two quarters saying that the entire change in the fourth quarter is because of this particular event, is not appropriate. We need to look at longer trajectories of time.
Demonetisation itself puts in place a variety of dynamics, some of which operate in short term and some continuing to operate into our system in long term. The short run ones are the cash exchange dynamics that people talk about. Long run relates to the behavioral changes which has been set in place i.e. manner in which people conduct these transactions. All of these dynamics are flowing out of a single policy which are discernable in the data. If you look at the manner in which we do business, the volume of e- transactions have shown a remarkable change, in the period post demonetisation.
So under all these circumstances, trying to reach a judgement about this complex policy’s impact, with variety of effects on the economy, is premature.
So what would be the right time frame to evaluate the demonetisation’s impact?
The answer will keep changing as you go further away from this event, as you gain experience from the same. CSO’s job is not to come up with an answer on how demonetisation had an impact on the economy, but to make sure that at every time point, as much data is available as possible to help you answer that question. The answers will be done by the community of analysts and academicians in the country.
Moving aside from demonetisation, explain it to us why was there a base shift (IIP &WPI) at this stage particularly and how did that help in getting the true picture of the economy?
The better way would be to refer-‘why did it take us two years after the base change of the GDP, for us to complete this work’.
For WPI and IIP series, a lot of information which is needed to be compiled is collected from the various entities, markets etc. and when we do a revision of the base, we have to set up reporting channels for these new entities and markets from the scratch.
Secondly, not only we have to get the flow of the current data but also approach these entities to retrieve data about their decisions on production and pricing, going back all the way to 2011. All these data collection is very time consuming, as everyone is not well connected in the communication network. The records have to be examined physically and data has to be retrieved from that. It’s a cumbersome process.
We started the work on this way before the release of the GDP base. But physically completing the work, it takes a while.
The reason we are emphasising on the base change and demonetisation numbers is because the country has come under radar for the authenticity of our data? The opponents have started comparing India with China and the whole subject of number fudging. How do you shun down those voices?
The issue of reliability from the statistical view point comes by answering how much access you get for the data. In India, there is complete transparency. Virtually all the data that we use is available in the public domain. The industry analysts also work with the same data, which is equally available to us. The degree of openness in India about data is much higher order of magnitude than in most of the developing countries and many developed countries.
On the questions raised regarding the reliability of the GDP numbers, many of these people confuse reliability with the fact that when we make improvements in the data, structural breaks take place in the series. These breaks are a part and parcel of the whole process of the improving the statistics. Therefore, the discontinuity which exists between the old and the new is simply because better data sources are being used.
But the analysts, who were working with the old data, find that their assumptions have been shaken. Rather than trusting on the data, they would go back and say that why did you make the change? My answer typically to them would be that you cannot ask for both better data and consistency of the time series at the same time. It is impossible to do the both. I think that inconsistency keeps repeating itself in people’s dialogue.
India and China cannot be compared when it comes to transparency and reliability of the data. I do not wish to make observations about the Chinese data system, but I would encourage the analyst to look at the documentation and data availability in India and compare that to any other country. We will certainly be open to meeting the standards of the best and leave it open to the people to make a call.
We are a very open society and Indian government is not just open but very transparent. Arguments that run conspiratorially require such a large number of people to keep quite that anybody who knows India or worked in India, should not make such statements with any degree of seriousness.
What does the data tell you about the economy, is it robust enough to take us back to the growth trajectory of 8 percent that we used to talk about earlier? Are we on the right track, seeing the data?
Almost certainly. Let me take some of the trends which were working their way out. One of the trends, I talk about is the price trend, which plays a very funny role in the GDP at constant prices. Because of a sharp crash in the prices, it pushed our GDP numbers in 2015-16 slightly up. As it was working its way out, the base effect first went up and then came down later, creating a jigsaw. As it works itself up, the trajectory will work itself out and you will come back to what would be a long time series. Consequential effect of that one element would be a pickup in growth, forget everything else.
Other than that, we will be coming out from one good year of rain and the forecast for the next year is also positive. The improvement in the rural economies through consumption, will lead to a boost in the overall demand. Without looking at the other policies of the government which are also side by side contributing to these numbers, just these two elements would bring a major improvement in the growth.
