Our statistical machinery needs serious repair, says Meghnad Desai

Desai told Business Standard that India must get the data issue sorted out on a priority as it is giving the country a bad name

Economist Lord Meghnad Desai’s first article in 1962 was to question the data calculation of the National Sample Survey Office (NSSO). On Thursday, speaking at the launch of a data science course at the Meghnad Desai Academy of Economics, where he is the chairman, Desai gave his own take on the current statistical controversies around growth, unemployment and told Sachin P Mampatta that India must get the data issue sorted out on a priority as it is giving the country a bad name.

Edited excerpts:

There’s been some debate on the quality of data that’s been coming out. Most recently on GDP (Gross Domestic Product). What’s your view? 

All countries’ GDP data gets revised ex-post. But by and large, neither the International Monetary Fund nor the World Bank, nor any of the rating agencies have suspected the GDP data coming out of India. I think the GDP data and the inflation data are fairly sound. 

What about the jobs data? 

We have to actually take a serious look at our labour market data. Fairly respectable people have come to completely opposite conclusions about the state of the labour market. The Centre for Monitoring Indian Economy says one thing. (Economist) Surjit Bhalla and his co-author — they have a paper on their website — have a completely different view. And then suddenly you have the NSSO data coming out. For anyone to argue that unemployment is higher today than it was ever, more or less, is very hard to believe.

Five years of GDP growth of 7 per cent, and a declining participation rate of males, which nobody questions… labour force participation among males is declining (when) the GDP growth is 7 per cent. You have to have productivity growth at a super-fast speed to generate unemployment. 

What needs to change then?

I think what’s going on is that the way we define employment is very narrow.  

There are livelihoods and there are jobs. By jobs we mean nine-to-five jobs. We are largely a service economy. And very few service jobs are nine-to-five, straight-down-the-middle jobs. Maybe in the organised services sector, but not in the unorganised services sector. 

At some stage after the election is over, there should be a serious effort by the government to say that we really have to put the labour market data on a firm footing, because we cannot be ignorant about what is happening in the economy. There are a variety of ways of making livelihoods. Some of which involve a proper, salaried job. We have to find out how to track those different kinds of economies where people earn livings with irregular labour market participation but they make a living.

What else does the government need to do? 

I think our statistical machinery is in need of serious repair. It really is a pity for a country which was a pioneer in these things. When the United Nations appointed the national income commission way back in the 1950s, India was ahead of most other countries. I think we have to get back there.

What do you think has contributed to the decline? 

It’s because different private sector and public sector (entities) come out with such different estimates. And I think that they are all honest but I have never seen such contrast across any modern economy in my career. Whoever comes to power, and I’m presuming it will be Modi... let us get this thing sorted out as a national problem because it is giving India a bad name.

The NSSO does seem to capture the informal sector that you spoke about… 

It must obviously have been doing that for a long time. But it must obviously have broken down somewhere. The NSSO is a very respectable organisation. 

But I should say that my first published article in 1962 was to question NSSO data calculation. It has a very good international reputation, no doubt about that. The CSO (Central Statistics Office) has a good international reputation, no doubt about that. The CMIE has been around for many years. 

Either our definitions are seriously different or the economy is doing things that we are not catching up with.