Karan Thapar, economist Arun Kumar, and former Chief Statistician Pronab Sen discuss the International Monetary Fund's (IMF) C-grade rating for India's national account statistics and the credibility of its Gross Domestic Product (GDP) figures. Kumar asserts that the GDP is significantly overestimated because the methodology wrongly proxies the shrinking informal sector using the growing formal sector.

Sen agrees the data is "less than reliable," citing the outdated 2011-12 base year and persistent difficulties in obtaining quarterly data for the informal economy. Both experts express serious doubt about the reported 8.2% growth for quarter 2, concluding that, despite best efforts, these data reliability issues will likely linger for years.

The following is the full text of their conversation. It has been lightly edited for style.

Karan Thapar: Hello, and welcome to a special interview for The Wire. Yesterday, the national accounts data for quarter 2 was released, and it coincides with some serious concerns expressed by the International Monetary Fund (IMF) about the way India calculates its data.

In fact, the IMF has given India's national account statistics, which includes gross domestic product and gross value added, a C grade, which is the second lowest grade.

We examine the IMF's concerns and ask how valid they are and what we should do in response to them. I shall talk to some of India's most highly regarded economists and my first guest is former professor of economics at Jawaharlal Nehru University, Arun Kumar.

Professor Kumar, the IMF has given India's national account statistics a C-grade, the second lowest grade. Let me start by asking how much of a concern this is?

Arun Kumar: It's a huge concern because the GDP data came under sort of disrepute because the 2011-12 series was not accepted by a lot of economists. Even the government did not accept it in the beginning. It set up another committee and that committee showed that the growth rate during the UPA years was better than the growth rate during the NDA years. So the government did not accept that.

Then it asked the NITI Aayog to do the calculation whereas the NITI Aayog is not the agency to actually give the GDP series. Nonetheless, it gave the series and said that during the NDA years, the growth rate was better than during the UPA years. But that was also challenged by the former Chief Economic Advisor [Arvind] Subramaniam, saying that it was overestimated by 2.5%. 

Then came this shocker that during the demonetisation, about 300,000 of the 1.8 million companies were removed as shell companies and the data did not change.

And then another shocker came, when for the survey of the services sector, they found that 35% of the companies were not where they were listed to be. So in other words 35% of the remaining companies also were not found to be there. So how was the data accurate? And then came another shocker, when the government did not accept the consumer survey of 2017-18. And then the next shocker came when they did not accept the employment data which showed that unemployment had reached a 45-year high. 

And then the census was not done in 2021. So the data became very old and the various surveys which depended on the earlier census figures, were not reflecting the population change accurately. 

So, one after the other since 2011-12, there have been various shocks, which have questioned the GDP data. Then the big problem that came was that the unorganised sector got hit by shocks. They had the demonetisation shock, which hit the unorganised sector disproportionately hard. Next year came the GST, which was faulty and hit the unorganised sector. Next year came the non-banking financer companies, and finally the pandemic, which again hit the unorganised sector very hard. 

So the old method of calculating GDP had to be changed four times but it was not even changed once. Therefore, all these projections that are being made are inaccurate and I think the IMF is reflecting some of this. 

Karan Thapar: If I understand you correctly, there are multiple reasons why the way data is calculated is inadequate if not faulty and wrong. Let's focus first on how India calculates its GDP. We use the formal organised sector as a proxy for calculating growth in the informal unorganised sector. But given the informal sector accounts for as much as 45% of GDP, do we really know how the unorganised sector is doing, or is it just an intelligent guesstimate?

Arun Kumar: If you look at the way the GDP is estimated, these are called benchmark estimates. In a benchmark year, you estimate the ratio between the organised and the unorganised sector, and then you move it every year through some high-frequency data. 

That means you are assuming that the ratio between the organised and the unorganised sector has not changed, and the high-frequency data comes mostly from the organised sector. In other words, you're proxying the unorganised sector by the organised sector.

Now the organised sector is rising but the shocks impacted the unorganised sector. So benchmark estimates went wrong and the high frequency data doesn't reflect the unorganised sector. Therefore you're making the big mistake of not calculating the unorganised sector independently and assuming that it can be proxied by the organised sector. 

So a declining sector is proxied by the rising sector and therefore GDP is overestimated as a result of that. And this 45% of the GDP that's coming from unorganised sector, you can say that about 14% comes from the agriculture sector – for which some data comes because you have crop cutting experiments and data for acreage and so on – but the remaining 30%, which is the unorganised non-agriculture sector, there the data is based entirely on these benchmark estimates and projection from the previous year to the next year. 

That makes it inaccurate.

Karan Thapar So once again if I understand you correctly, you're suggesting that we do not have a good idea of how the unorganised non-agricultural sector is doing?

Arun Kumar Yes, that's right. So now of course they say that we are getting the ASUS surveys that look at the unorganised sector, but those are not incorporated into the data calculation. The IMF has recommended that that be done and I suppose that next year when the revised data will come, then maybe the ESUSC data would be there and the unorganised sector may be represented better.

But till now it is not there and it's overestimated because you're proxying a declining sector by a rising sector and you have enough evidence for that. For instance, in the trade sector, which is the second biggest employer, now the rise of the gig economy, the rise of the deliveries has really taken over. So e-commerce is growing at 20-40% whereas the neighborhood stores are declining. The neighbourhood stores are the unorganised sector, the e-commerce is the organised sector. 

Karan Thapar: What you're saying is again deeply worrying. You said 30% of GDP is the unorganised non-agricultural sector. That's a very substantial chunk of our GDP. We're overestimating how it's performing because we are proxying it to a rising sector whereas this sector is itself shrinking and that means we have a very poor idea of how 30% of the GDP is performing.

Arun Kumar: You're right. We have scattered data – like I mentioned from the trade sector, which is the second biggest employer in the country, we have similar data from textiles, from leathers, from FMCG, from the luggage industry and so on and so forth. 

The scattered data shows that the unorganised sector is declining. My own estimate in one article was that it's declining between 5-9%. So what is being shown is rising at 6% is actually declining at 5%. An 11% error in the estimation of the unorganised non-agriculture sector.

Karan Thapar: An 11% error is enormous. Let's at this point come to some of the details that the IMF has brought out. They say there's a sizeable discrepancy between the production and expenditure approaches of measuring GDP, as a result of which, often, the two differ. How much of a concern is this?

Arun Kumar: When we look at the national income, there are two aspects of it. One, you get an income and then out of that income you do the expenditure. So there are two sides of the same coin. So therefore they should be exactly the same. What is called the income or production side and what's called the expenditure side. But what happened is that the unorganised sector impacts both these measurements.

Because the unorganised sector is in every sector. It's on the lever side of the public sector. It's in services. It's in industry. it's in agriculture. So when you don't have proper data for the unorganised sector, it affects your production data and similarly the consumption side data is definitely affected because bulk of the consumption uh items are produced by the unorganised sector. Now if you're not estimating it properly then your consumption data goes wrong and most of the investment is in the organised sector.

The organised sector may be overestimating the capital expenditure and underestimating the consumption expenditure. That's why the two don't match – the income side and the expenditure side. So these are some of the problems because of which a category called discrepancy is created in the accounts. Those discrepancies have shot up. Prior to demonetisation, the discrepancies were very small, 0.5-1%. Now they can be between 2-5% and therefore they are affecting our GDP estimates. 

Karan Thapar: And the larger the discrepancy, the greater the concern.

Arun Kumar: The greater the concern about what is going wrong. The production side is taken as a controlling total for the expenditure side that's why it's sometimes positive sometimes negative. But the production side itself is affected because each sector has an unorganised sector. 

So the value addition that you're calculating there, that goes wrong and if it that goes wrong then you cannot say that's the controlling total. So we don't know which should be the controlling total and which is the one that we should lay emphasis on. 

Karan Thapar: Now the IMF also highlights the fact that India's GDP is based upon an outdated base year 2011-2012. And secondly, the IMF says that India uses the wholesale price index as the GDP deflator because we don't have good producer price indices. Are these serious concerns? 

Arun Kumar: Here’s what the decent deflator means. In any economy, there is a rise in production and there's a rise in price sales. So what's called the nominal GDP is the sum total of the rise in prices and the rise in the output or what's called the volume. So these two together give you the nominal GDP. 

Then you deflate it by the price index to get the real change in the economy. So if you don't have the right index then what are you deflating with? The nominal GDP that you're measuring, when you deflate it by the incorrect index then you get incorrect GDP data. So the wholesale price index reflects only a certain aspect of the economy. Largely it's the agriculture and the industrial sector but it doesn't necessarily reflect the services sector very well. 

If you deflate the services sector by the wholesale price index, it will not give you the correct estimate and the services sector is the largest sector of the economy. So if the largest sector of the economy is not being deflated by the correct prices, then you will not get a good estimate of the volume or the real chain. The second point is that when you think of the wholesale prices, they differ from what the producers are facing. Because wholesale prices include trade. Therefore producer prices are supposed to be better when you want to deflate them. 

But there's another aspect of it which is called double deflation. Here, when you refer to the deflation by the wholesale price index, that's only a single deflation that is being done. The inputs that you're using to produce something, their prices may be going differently from the output at which you're selling. So the output prices and the input prices may differ and therefore there's something called double deflation – you need to deflate the inputs differently and the output differently to get the correct value addition. 

If you only do single deflation or only with the wholesale price index, then there is an error that will creep into the estimation.

Karan Thapar: This sounds like a fairly substantial error that's creeping in. 

Arun Kumar: Yes, it is and some analysts have pointed this out in the last 2-3 years that this is causing a major error. And this is the reason why suddenly the GDP growth seems to be better than it is because on the ground if you look at it, the GDP doesn't seem to be growing that well. 

For instance, the latest data that came yesterday is showing a very rapid rise of 8.2% in the GDP. But the point is how is it being projected and the whole price index change is very little and therefore most of it is being shown as this…

Karan Thapar: I'll come to the most recent data in a moment's time. Let me first put to you one more concern that the IMF has highlighted. The IMF says that India lacks seasonally adjusted data and it calls for improvement of statistical techniques used in the quarterly national accounts data. 

Do we lack seasonally adjusted data and is that a concern? 

Arun Kumar: Well, I mean it's not that we entirely lack it. For instance, we know that the fourth quarter data is always high because suddenly the production spurts. The first quarter data is low. So there is some degree of seasonality that's already there in our data. But it could be better. You need to have more seasonally adjusted data. But I would not put that much store on this particular aspect as on the other aspects of how the GDP is calculated.

Karan Thapar: In which case, let me ask you this. How reliable is our GDP calculation? You've enumerated many problems. The deflator being one of them. The fact that we cannot properly estimate the unorganised sector which is 30% of GDP, if you exclude agriculture, is a second one. So how reliable is our GDP calculation? 

Arun Kumar: So I have been arguing that since demonetisation, our GDP calculation is becoming less and less reliable. In the demonetization year itself, from November till March – that's five months of the year 2016-17 – the ground report showed that people are not going to markets, the manufacturing had declined and so on. 

So for five months there was a decline or stagnation in the GDP. But the best growth for the entire 2020s period was for that year, for 2016-17, at 8%. Whereas estimates were that it's actually -2%. So in other words what is -2% gets projected as a 8% growth because you're only basing it on the organised sector and you're also basing it on projection from the benchmark year to the next year. Now the benchmark year was a good year but the year in which you are calculating, when demonetization takes place, is a bad year. So projecting from a good year into a bad year will only overinflate your data.

Therefore it shows that there's a 8-10% error possible when there's a shock. And we have a series – you had demonetization, then the GST, then the NBFC and then the pandemic. So four big shocks mean that the GDP data has been going wrong because if you project from a good year into a bad year, and then project into the next bad year, then again both will be overestimated.

Karan Thapar: So you’re saying, and I am deliberately using layman language, that India's GDP calculation is not reliable. It's probably not very credible. 

Arun Kumar: Yes. 

Karan Thapar: Are you really seriously saying that?

Arun Kumar: Yes. Yes. So my estimate is that currently the actual GDP is about 48% less than what the official GDP says. So when the government says it's $3.8 trillion, my estimate is it's probably still $2.5 trillion because we are overestimating the unorganised sector, which is actually declining. That is building up over a period of time. 

Karan Thapar: Against this background let's come to that issue you raised, which I said I would take up later. How real is the 8.2% growth reported yesterday for quarter 2? After all, nominal GDP is only 8.7%, and yet we're claiming that real GDP has grown by 8.2%, which suggests inflation is just 0.5%. How credible is that figure? 

Arun Kumar: So inflation, officially, has come down. There's no doubt. But we know that in the month of September sales of a large number of goods had stopped because of the GST. So how is it that the rate of growth has accelerated rather than decelerated? The reason is that the quarterly data is based on benchmark year and then moved up by the high frequency data. But if you look at the high frequency data, out of the 18 or 19 items that are used for high frequency data, most of them are from the organised sector. They're not from the unorganised sector. 

And if there are other high frequency data that are being used, they should be listed. They're not listed in this. So if they're not listed, we have to go by this. And if most of them are declining compared to last year's quarter 2, then how is it that it has suddenly accelerated as compared to last year? So this is a challenge that was there in quarter 1. This is a challenge in quarter 2 – also. 

Karan Thapar: So the truth is you don't believe the 8.2% figure for quarter 2

Arun Kumar: Correct.

Karan Thapar: What do you think it should be? 

Arun Kumar: Well I haven't reestimated it. I would have to reestimate it. But certainly not accelerated over last year's Q2 data and last year also the data had similar problems. 

Karan Thapar: Last year was 5.6%.

Arun Kumar: Correct. But that also had similar problems to what it is that you know you don't get the unorganised sector data, you're moving the data based on the previous year, and the high frequency data etc. 

So my thing is that  if you take the pluses and the minuses, the economy declined during demonetization, it declined during the pandemic and then accelerated the year after. On an average [the economy] has not grown at more than 2.5%. Whereas the government says it has grown at 6.5-7%. So there's a 4% overestimate that is going on all the time because of the methodology. 

Karan Thapar: So in a nutshell you're saying that the actual growth rate for Q2 should be somewhere in the region of 5.6-6.5% rather than the 8.2% reported yesterday?

Arun Kumar: I would not even go to the extent of 5-5.6%. Because if the unorganised sector is being overestimated by something like 8-9% and then you apply that to 30% of the thing, then it'll be down by at least 3-3.5%. So it'll be closer to 5% or 4.5%. 

But again, as I said, I'd have to look at the data because even some agriculture experts say that the agriculture data is incorrect. So one would have to see what the problem is. Three years ago, the government said that there was a record production of wheat and rice. But actually they had to ban the export and reduce stockholding. Now if there was record production, why would prices rise in spite of exports being banned and stockholding being limited? That is why some experts say that even the agriculture data is incorrect. It's overestimating some of the things. 

Karan Thapar: Let's move from GDP to the consumer price index. The IMF also comments about that. Now the com consumer price index is India's main measure of inflation and the IMF has given India a B-grade, which means it's broadly adequate for surveillance. In this case, should we be pleased with the B-grade? You said that the C-grade, overall, was a huge concern. But specifically for the consumer price index, is the B-grade something we should be pleased with or should we be sorry that we haven't got an A grade?

Arun Kumar: Well B-grade is sort of okay because they say for surveillance purposes. It could be better but my concern about the consumer price index is the following. The consumer price index is measured as a basket of consumption and that basket of consumption varies for the poor, the middle class and for the rich.

We cannot have just one consumption basket reflecting the reality of what the poor are consuming or what the well-off are consuming, especially given the GST has reduced the prices of a lot of consumer durables. But most of the consumer durables are not consumed by those who are poor. So 90% [of the people] will not buy cars or other such big items on which prices have come down. Therefore, this GST has reflected a decline in the organised sector prices but not the unorganised sector prices. 

The impact of this GST cut on the consumer price index would be different for the poor from the well-off sections. So my concern would not just be monitoring, but in terms of what it means for the people of India. 

Karan Thapar: Well, in fact, the IMF has made that point as well. The IMF says that India's CPI basket and I'm quoting now “fails to accurately represent current spending habits”. Let me ask you for the audience, how large and significant is that failure? 

Arun Kumar: You see the failure is at both the ends. Because the consumer price index is based on an average basket. That average basket neither represents the well-off segments nor the poor people. So in that sense it's a problem. 

Second, people's consumption habits keep changing. So what I was consuming 10 years back and what I'm consuming today has changed. Even for the poor people, that consumption basket is changing. For instance, a rikshawalla may have a mobile phone, but having a mobile phone doesn't necessarily mean he's not poor anymore. But his consumption pattern has changed. That needs to be reflected within our consumption basket.

Karan Thapar: And those changes are not being made?

Arun Kumar: Yeah, those changes have to be made and the government says they're trying to make those changes. I think the IMF has said that those changes are coming. 

Karan Thapar: Now, the Ministry of Statistics and Program Implementation is presently working on updating GDP and consumer price index bases, as well as the methodology for calculating both. And the new series will be released sometime next year. I think the Business Standard suggests it will be released at the end of February, possibly on the 26th of the month. How hopeful are you that the concerns the IMF has raised will be adequately taken care of? 

Arun Kumar: I think they are substantially looking at these things –  the kind of criticism that has been there for the last 10 years and which is at the back of what the IMF is also saying. 

But I think the problem that India has is this huge unorganised sector that has something like 63 million units and which employs roughly 97.5% of the MSME employment – measuring that is huge. 

Similarly, in agriculture, there are, by various estimates, 80-100 million farmers. Bulk of them are actually small farmers. And 85% of them cultivate less than five acres. Measuring their output and prices, those are the real challenges. So we can have a methodology which is better than what was there earlier, but I believe that it'd be very difficult to capture that unless we have a much bigger sample. The IMF has flagged that we need to have a much much bigger sample than what we have been working with. Whether the government would be able to do that or not becomes a question then.

Karan Thapar: But there is another truth professor Kumar. The truth is that even last year, India's national account statistics received a C-grade from the IMF. 

Arun Kumar: Yes.

Karan Thapar: And the IMF has pointed out that data weaknesses have remained unchanged. In other words, data that could have been strengthened and improved has not been strengthened and improved. And the IMF is clearly critical of that. 

Arun Kumar: Yeah. So, you're right that in 2024 too they gave us a C, but in 2023 they did not give any rating. They are concerned because there have been several pointers [about] the data problems. 

For instance, the 2017-18 consumer survey data was not accepted. In 2019, the employment data was not accepted. And then the census not being done means that a large number of surveys are unreliable. The census is very crucial. As long as we don't do the census and we do not change our samples, our problem of data will continue. 

So I think the IMF is right in flagging this – the data has a problem. But the government says it's trying to do its best and next year it'll do something. The census will also come, data from the census may also come after a year or two. So the government's standpoint will be that we are doing our best and the IMF has also recorded in annexure 7 that the government is trying to change these things. 

Karan Thapar: My last question, you say the government claims it's doing its best. Is the government's best good enough? 

Arun Kumar: Well, till now it has not been good enough. But we hope that it'll be good enough. The government depends a lot on data manipulation to show a good image. And that's the real worry. Will the government leave the agencies neutral? Like the GDP data, when they asked the NITI Aayog to do it. The NITI Aayog should not have done it. It should have been a statistical organisation that does it regularly. So the fear is whether the manipulation of data will continue in spite of all the good intentions about changing the base year and incorporating better unorganised sector data etc. But will the manipulation decrease? That is the real worry.

Karan Thapar: Professor Kumar, you've made that point very clearly. The real danger lies in the government's desire to manipulate data and present it in a way that looks much better than reality. I thank you for the time you've given me. Take care. 

Arun Kumar: Thank you.

§

Karan Thapar: I'm now joined by India's former chief statistician Pronab Sen. Professor Sen, the IMF has given India's national account statistics a C-grade. Let me start by asking how much of a concern is this?

Pronab Sen: Well, it depends on whose perspective you're looking at it from. If you look at it from the IMF's perspective, what they are saying is that this particular set of data is not adequate for the kind of surveillance that the IMF does on member countries. The other side is us. who are users of the data on our day-to-day decision making. For us, the accuracy is more important

Now the report of the IMF is very interesting because it gives us a B-grade on all the data sources that are going into the the GDP but gives the GDP a C grade, which means that they are looking at – and they haven't articulated it clearly enough for me to pass comment – what the nature of the problem is.

Karan Thapar: Nonetheless, a C grade is the second lowest grade. Should that be a matter of concern for the world's biggest democracy?

Pronab Sen: Well, this is the thing – they are saying for surveillance. This means that the IMF monitors how our policies are playing through the economic system. That's what the data is used for by the IMF. And I think what they are saying, in effect, is that the data that's been provided doesn't give them enough confidence that the changes that are happening in the economy are being picked up at adequate intervals. 

Karan Thapar: And how serious is that? 

Pronab Sen: For them, it is serious. Because they are interested not in the long-term trends of the economy, but in the more shortrun movements that come as a response to policy changes. So that may be a source of concern. But I have not been able to understand what the heart of the concern is. 

Karan Thapar: Let's first focus on how India calculates its GDP because there, as you pointed out, the IMF has given India a C-grade. We use the formal organised sector as a proxy for calculating growth in the informal unorganised sector. But given that the informal sector altogether, including agriculture, accounts for pretty close to 45% of the GDP, do we really know how the unorganised sector is fairing or is it just an intelligent guesstimate? 

Pronab Sen: When you don't have data available, there is no option but to use proxies. The question is are the proxies reasonable. And remember, these estimates are changed when the actual data becomes available. Now this is something that we've been doing for years, for decades even.

There's been some concerns, but when you look at a country like India, where, as you rightly said, 45% of the economy is informal and about 28% is non-agricultural informal sector, getting regular data from these entities is extremely difficult. But in the defense of the national statistical system, they've been trying to make efforts at doing so. Remember, all this data has to be collected through surveys and surveys are both expensive and extremely time consuming.

And this is not taking things off the books of account. Because of that, there has been some delay in incorporating the more frequent surveys that we've been having lately and I think this is a game plan. They plan to kick it off with the new revision that they're doing to the GDP series. 

Karan Thapar: I'll come to that new revision in a moment's time. But let me first ask you at the moment, the way we calculate our GDP, do we have a reliable and credible understanding of how the informal unorganised sector is doing? 

Pronab Sen: We have some understanding. But the point is that unless you have hard data you can only go by past relationships. And so what has happened in the Indian system is that we've looked at the data points that we have from the unorganised sector – that is the hard data points – and we've tried to correlate those to the formal sector data. And if we feel that there is a close enough correlation, then it makes sense to use the formal sector data which we are getting with quarterly frequency. 

Karan Thapar: But is there that close relationship? 

Pronab Sen: It's not that close, but it's close enough. So the way they do it is not by taking large aggregates. They do it sectorally by product groups. 

Karan Thapar: So are you saying we do have a reliable and credible idea of the unorganised sector or are you saying it's less than reliable, less than credible? 

Pronab Sen: Well, no, no. It's less than reliable. I'm not sure how less than credible it is. It is less than reliable for the very simple reason. There's a strong underlying assumption that both the formal and the informal sector are essentially moving in the same direction, not perhaps by the same degree, but the same direction. That's the key assumption.

Karan Thapar: And does the assumption hold?

Pronab Sen: Now that is where the problem crops up. If we have reason to believe that they've moved in opposite directions, then what would end up happening is that you would be either overestimating or underestimating the unorganised sector.

Karan Thapar: And after demonetization, GST and the pandemic – just to take those three – are the organised and the unorganised sector moving in the same direction or are they experiencing very different results?

Pronab Sen: No. In fact, you're absolutely right. Those three shocks led to a situation where they were moving in opposite directions. The unorganised sector was diminishing and the space that was being vacated, the market space that was being vacated, was being taken over by the corporates. So the corporates were expanding rapidly and the unorganised sector was shrinking.

This has reversed over the last couple of years. Once the effects of the lockdown were gone. Now there is evidence that the unorganised sector is starting to bounce back and grow. So we have had a blip in the middle. Now whether that's enough to damn the entire exercise is moot. 

Karan Thapar: But you're saying two very interesting things. There were several years when in fact we were probably overestimating the unorganised sector because we use the organised sector as a proxy. And we were doing so at a time when the unorganised sector was diminishing while the organised sector and the corporate sector was growing. 

So there were years when we were overestimating the unorganised sector. Now you believe now you believe they are back in kilter. 

Pronab Sen: Well they I'm not sure they're back in kilter but they're moving back towards kilter. 

Karan Thapar: So even now we are overestimating the unorganised sector?

Pronab Sen: We could be overestimating the unorganised. Yes. 

Karan Thapar: So the IMF's concern about the overall GDP figure therefore must be a correct concern. We're overestimating it. Well, yes, but the solutions are not obvious. Because the only way you can correct it is to do away with the assumption that the informal sector behaves roughly like the formal sector. 

This means you have to be able to get independent data.

Karan Thapar: Which we don't have. 

Pronab Sen: Which we do have now. Now we are doing annual surveys. So we still won't get it on a quarterly measure. Perhaps we could do that as well. But we'd certainly have annual data. 

Karan Thapar: Let's come to some of the other concerns the IMF has with our GDP calculation. The first concern the IMF has, it says, is there are sizable discrepancies between the production and expenditure approaches of measuring GDP. As a result of which, the two often differ. How much of a concern is that? 

Pronab Sen: That is a concern. Let me try and give you a sense of where the problem lies. The problem essentially lies in prices. So if you think of the production approach measure, those are supposed to be measured from the point of view of the producer. How much is the producer getting?

The expenditure side is how much are those who are spending money, how much are they paying. Now, the difference between these two is that the prices are very different. For instance, if you're looking at consumers, you would use the consumer price index because that's what people are paying. But the consumer price index is not what the producer is getting. So for that you need what's called a producers's price index, which we do not have. So the concern over prices is a legitimate concern and this is a concern we've been aware of for decades. 

We've made several attempts to collect data on producer prices. But here's the problem – no producer, particularly in the formal sector, is willing to give you what they get. They treat that as highly confidential and we've been banging our heads against a brick wall trying to get that information. Apparently the NSO is trying, yet another time, to get hold of that data and I wish them all luck. 

Karan Thapar: You're suggesting that it's an unsolvable or irresolvable problem but nonetheless the problem of…

Pronab Sen: It is not, Karan. It is not an irresolvable problem. In the National Statistics Act, we have a provision where we can coerce the corporations into giving us this information. 

Karan Thapar: Why isn't it used?

Pronab Sen: We prefer not to. We prefer that it be voluntary, 

Karan Thapar: But it's not coming to us. 

Pronab Sen: In developed countries, it is not voluntary at all. It is mandatory and there are severe penalties.

We do have penalties but to the best of my knowledge, this particular provision of the statistics act has never been invoked in India.

Karan Thapar: Which means that the discrepancies between the production and expenditure approaches of measuring in GDP remain and those are serious discrepancies. 

Pronab Sen: Yes, there are serious discrepancies. But they've gone up and down. 

Karan Thapar: Now the IMF also highlights the fact 

that India's GDP is based on an outdated base year 2011-12 and secondly that India uses the wholesale price index as the GDP deflator because we don't have good producer price indices. These I take it as serious concerns. 

Pronab Sen: The IMF is absolutely right. We have delayed changing the base year for far too long and as I already mentioned, we haven't been able to get the producers prices at all. 

Karan Thapar: So these are serious concerns.

Pronab Sen: It is a serious concern. But, in a sense, what I think the IMF is hinting at, is please invoke the penal provisions of the Statistics Act. That's what they're saying. Now can you imagine the reaction that we will have from the corporations, if we tell them to either give us this data or pay a daily fine of Rs 1,000. 

Karan Thapar: It may make the government temporarily unpopular but it will mean that the data that we have is accurate and the figures we produce as part of the national accounts are accurate as well. At the moment they're not accurate. They're not even reliable.

Pronab Sen: Supposing they just don't respond to you, what do you do?

Karan Thapar: You mean the government would be helpless? 

Pronab Sen: The government would have to take them to court. Can you imagine taking practically each and every corporation to court to gouge the data out of them? 

Karan Thapar: So then this is a problem. 

Pronab Sen: Well, it's a difficult problem. I mean that is why we try to keep it as voluntary as possible so that the respondents give us the information because they see the utility of it for everybody in the country. 

Karan Thapar: In other words, we have to persuade them to accept that they have to reveal the data and we're not able to persuade them at the moment. But we have to just keep trying in the hope that we succeed one day.

Pronab Sen: One day. That is correct. 

Karan Thapar: In which case, let me ask you this. At the moment, as things stand, how credible, how reliable is the GDP figure we regularly produce? >

Pronab Sen: Well, in my opinion, it's no better and no worse than what it was ten years ago. 

Karan Thapar: So, it hasn't improved?

Pronab Sen: Data sources haven't improved. We are trying to improve them. The NSO is trying to improve them, right? But we will get the results of those efforts only when the next GDP base change happens, which is probably next year. 

Karan Thapar: So at the moment, what we have is, as you said at the beginning, less than reliable. 

Pronab Sen: Yes. 

Karan Thapar: Did India's GDP really grow by 8.2% in quarter 2, which is the figure we announced yesterday? Because nominal GDP is only 8.7%, which suggests that inflation is just 0.5%. So is that figure of 8.2% reliable? Is it credible? 

Pronab Sen: This is the million dollar question. And you know what surprised me about the IMF report is that they actually give us a clean chit on the price data.

Karan Thapar: Which you don't accept? 

Pronab Sen: No, my personal view is that the whole problem is in the price data. The IMF has mentioned that our price data also have outdated bases. That is correct. But an outdated base has a very serious implication, which is, that as people's consumption patterns change, certain goods and services go out of the market and new products come in. 

The waning products will have diminishing and lower prices than the new products coming in. So we are measuring a lot of products whose prices are probably on the way down because those products are probably being phased out. 

Karan Thapar: So what's the answer to my question then professor Sen?

Pronab Sen: The answer to your question is simple. We, in principle,have agreed to revise our price indices every five years, which we haven't done. 

Karan Thapar: So is that 8.2% growth figure for quarter 2 credible and reliable or do you have serious doubts about it?

Pronab Sen: It depends on the prices right now. If you think about it, we know what the consumer price index has been saying. The consumer price index has essentially been saying that the prices are extremely low. And 0.5%, no, it's not as low as that. But what seems to be missed out here is the fact that the low inflation that we've experienced in the recent past has essentially been driven by falling agricultural prices. 

What's called the core sector, there the inflation is about 4%. It's not 0.5%. That 0.5% is coming essentially from agriculture, where you've had negative inflation.

Karan Thapar: What you're suggesting to me therefore is that you have serious reservations about accepting the 8.2% figure because that's based upon a deflator of 0.5%. 

Pronab Sen: Yes. So here's the interesting thing. Agriculture prices have been falling but look at the growth rate of agriculture in GDP. It's only 3.5% 3.6%. We are used to a growth rate of about 3%. But a negative price and debt should have pushed that up to seven or eight? It hasn’t. Manufacturing has gone up. 

Karan Thapar: Okay. So I will then, for the sake of the audience, underline that you do have serious reservations about accepting the 8.2% growth figure for quarter 2 announced yesterday, because it's based upon a deflator of 0.5%, which may be true of agriculture, but you don't believe it's true for the rest of the economy. 

Pronab Sen: That's right.

Karan Thapar: Now, let's come to the consumer price index because the IMF has also commented upon India's CPI, which is India's main measure of inflation. 

The IMF in this instance has given the CPI a B-grade and said that it means broadly adequate for surveillance. In this instance, should we be pleased or should we be disappointed that we haven't qualified for an A grade? 

Pronab Sen: It would be a little difficult to qualify for an A grade until and unless you met your own specified time limits of revising it every five years. The last time we did it was 10 years ago. 

Karan Thapar: So the B-grade is justified.

Pronab Sen: The B is certainly justified. 

Karan Thapar: The IMF has also commented that the CPI basket fails to accurately represent current spending habits. How badly does it fail in this regard? 

Pronab Sen: The two questions you asked, Karan, are related. The reason why I think the CPI is not up to the mark is because you haven't revised the basket. So you are capturing old products, products which are on the way [out]. You're not capturing those products which are in demand today. 

Karan Thapar: Which also suggests that the CPI is not credibly reliable. 

Pronab Sen: It's reliable insofar as it goes.

Karan Thapar: Which is not very far. 

Pronab Sen: No, it's 10 years old. I mean it should not have been more than 5 years old. 

Karan Thapar: Now the MoSPI is presently working on updating both the GDP and the CPI bases as well as the methodology for calculating both of them. And the new series will be released sometime next year, probably at the end of February. 

How hopeful are you that the concerns that the IMF has raised – both about the CPI and the calculation of GDP – will be adequately responded to?

Pronab Sen: I think it would address it to quite a substantial extent. But my source of confusion is that the IMF hasn't commented on the data that the GDP is based on. And the corrections that are going on are on the data side. The IMF has, on the data side, said we are pretty much in B but this GDP is in C. Now that confuses me. Are we doing something wrong in the compilation of the GDP?

Karan Thapar: We are. 

Pronab Sen: We may be. What we do know is that for the quarterly GDP estimates, we make a lot of assumptions. We simply don't have quarterly data for most things. Now when you don't have the data, you have to go by assumptions. You look at past relationships, past trends and try to do the best you can. But until we get to a situation where most of the data that we need for quarterly estimation are actually collected physically, this problem is not going to get solved. 

Karan Thapar: So you're suggesting to me that even though we will make our best effort to tackle the concerns the IMF has pointed out, you don't think they will be fully or even perhaps adequately resolved? 

Pronab Sen: Yes. I don't think we can for the very simple reason that you had brought up right in the beginning, which is the informal sector. The only way we are going to get informal sector data is through surveys and surveys are time consuming. We cannot do it in a quarter. 

Karan Thapar: Which means that the problems the IMF is talking about, despite our best effort to correct them, will linger and continue for many years to come. 

Pronab Sen: Yes. But they will go into the rear view mirror once the economy has totally stabilised because then everything's going to move together. 

Karan Thapar: But we don't know when that will happen.

Pronab Sen: The whole problem has cropped up at a time when there were a lot of ups and downs in the economy. 

Karan Thapar: Quite right. GST, demonetization and a pandemic to name just three. Two of which were man-made.

Pronab Sen: Yes, indeed. So give us another year or two. And once things come back on even keel, I think a lot of these problems will simply get overlooked. 

Karan Thapar: Well let's hope you're right professor Sen. I thank you for the time you've given me. I thank you for your understanding. Take care.

The Wire is now on WhatsApp. Follow our channel for sharp analysis and opinions on the latest developments.