Mumbai: India is facing reverse migration of labour from factories to farms, though the norm in the developing world is the opposite. The Centre for Monitoring Indian Economy (CMIE) has been monitoring these numbers and its Consumer Pyramids Household Survey shows that these numbers have been steadily rising in recent years. The government's own Periodic Labour Force Survey (PLFS) shows that employment in agriculture, as a percentage of total employment, has gone up from 42.5% in 2018-19 to 45.6% in 2019-20.

This raises twin questions: Is India unique because it is reporting this reverse trend? And what does this mean for the economy? To discuss this, we are joined by Mahesh Vyas, chief executive officer, CMIE.

Edited excerpts:

How large is India's total labour force and how has it been moving around, as per your own studies and the government's reporting?

When we talk about the labour force of a country, we usually talk in terms of people who are more than 15 years of age. That's an arbitrary cut-off but it makes sense that people who are less than 15 years of age are more likely to be in school or should be in school. From 15-19 years of age, there's a gradual increase [in the number] of these people coming into the labour force, where they offer themselves for employment. So if you take the employment of [those who are] more than 15 years of [age] in India, that's nearly a billion people.

The world over, a large proportion of those 15 years and older, to the order of 65%, come to the labour markets and say that 'we're willing to look for a job'. Most of them find work and therefore, a fairly large proportion--more than 55 or 57%--actually finds employment.

In India, the proportion of this large stock of people above 15 who offer themselves for work is very low. According to our estimates, that number is 42-43%; [and is] now going lower, at 41% or so. Of this, a certain number of people get jobs. According to us at the CMIE, that number is of the order of 400 million people. So of the billion people or so, approximately 400 million (40%) are actually employed in some way or the other. The government--the PLFS [Periodic Labour Force Survey], for example--does not give you the absolute numbers. They only give you the ratio and it is up to researchers to convert those ratios into actual numbers. The people who have done this in the past will give you a number that's anywhere between 450 million and 470 million or so. The CMIE's number is 400 million, the government's number would be 460-470 million.

If we were to use 400 million as the benchmark, what would be the sector-wise numbers?

According to the CMIE data, of the 400 million people who are employed, 150 million are in agriculture, 150 million are in services, about 90 million are in industry, of which only 30 million are in manufacturing and 60 million are in construction. So less than 10% of the total number of people who are employed in this country are in manufacturing. This number has fallen--it used to be around 40 million, it's come down because of the stress we've had in the last couple of years.

Tell us about this shift--of people migrating from factory to farm--which according to the government's figures, is even more pronounced than what you have reported. And what does this mean in the macroeconomic sense, looking at how trends usually evolve in other developing countries?

One of India's big achievements after the 1991-92 liberalisation was a big shift from farms to factories. The debate in those days used to be that we have got jobless growth, but then researchers found out that there was a big shift in favour of high productivity--from farms to factories. So we did not grow those jobs but we did move people out from farms to factories. In the country as a whole, there was a great improvement in productivity.

In recent times, we have been seeing that the non-farm sectors are not able to provide adequate jobs--and that makes sense because from 2011-12 or so, the investment ratio has been falling steadily. So if investments don't happen, no new jobs get created, what will happen to the new labour force that comes in? It will obviously try to come to cities to find jobs. If it doesn't find jobs, it will go back to farms and claim employment in farming. And that's very disturbing, actually.

More people are coming into the farm sector, which theoretically should have been going through its own evolution in terms of efficiency. Are people gainfully employed? Is there a way to measure that?

When it comes to people employed in agriculture, there are two kinds. To simplify the problem, let us say there are farmers and then there are agricultural labourers. There is a substantial increase in farmers employed, say, as clerks or security guards in a factory. Because that factory could be an MSME in the rural sector and it shut down, the farmer lost her job and went back to her village and started doing what she was doing before she came to the factory--which is being involved in carrying some firewood or shovelling some mud or something similar, very low productivity work. But when you go and ask her whether she is employed, she will say, "Yes, I'm employed as a farmer in my fields." This is actually disguised unemployment--this person actually should have been classified as unemployed, but because she went back home and there is a small farm, where you can always go and do something, she called herself employed.

In CMIE, we do not consider many of these people employed because they're hardly doing any work, whereas in PLFS, they go to a great extent to say, "Even if you're doing a little bit of work in your farm for family business, unpaid, we will still call you employed". This is actually not a good thing in our measurement system, we need to recognise that women do work a lot--and many men do work a lot--but they're not gainfully employed.

You said the biggest increase in employment was in agriculture but within that, the increase was largely among women--60% of all employed women were engaged in agriculture. What does this imply?

Women work the hardest in our country and I use the word 'work' very deliberately, where I distinguish work from employment. Women are most likely doing three shifts. They are working in the fields, they are working at home--taking care of children and old people--and running the house. So they're doing great work but they are not fruitfully employed. What happens is that they are mostly working in family enterprises, not adding much to the output. They're merely being involved. So they're spreading the available work very thin.

Ideally, we would have liked these women to be working in factories, where the productivity is much higher, or in the services sector where the productivity is also quite high. You can even shift from agriculture to construction--even there, the productivity is high. So the normal transition is, women who are not skilled--or even men who are not skilled--move from farms to construction sites, where they gain a few skills, where their children get better education, and then move to factories. And then they move to the services sector for even better productivity.

This transition is normal for every country in the world. In our case, because of lack of investments and stalling of the growth process, this transition has reversed from factories to farms. This is actually a structural problem that seems to be playing itself out. It will require us to work hard to cover the hole we are digging ourselves into.

Over the last two-three years, we have seen a lot of foreign direct investment coming in, mostly into the newer technology firms, as venture capital, private equity. All of that is also creating jobs of a certain kind. Is that not having any impact on these numbers?

Not an adequate impact. Obviously, all the investments that are going into all kinds of e-commerce companies or similar new-age ventures, even fintech to a small extent, are having an impact. But it's not adequate. It's not enough to employ lots of people for a long time and in good quality jobs.

This is the problem. Every time we say that unemployment is rising, or the employment numbers are not rising adequately, somebody says: "Look, these people are adding so many million jobs; IT companies are hiring so many more people, they are hiring lakhs of people right now." So obviously somewhere someone is getting employed all the time. The point is that it's not adding up adequately to absorb all of us sufficiently.

The government's estimate of 45.6% of the working population being employed in agriculture is higher than your estimate of 38% for 2019-20. Why is this the case and what does this tell us?

The difference is largely because they show more women working in agriculture. And I gave an example earlier. What happens is that, the number 45.6% that the PLFS puts out is on what they call the 'usual status', which is actually very, very lenient. The usual status effectively says that if a person is employed for approximately a month, in an entire 365-day period, that person is called employed for the year as a whole. Even the current weekly status that the PLFS uses says that if you're employed for half a day in a seven-day period, you're employed. Mind you, if the person was employed for 30 days, let us say in September 2019, and then was unemployed all the way to June 2020 and was desperately looking for a job, PLFS would still say 'preference to employment'. Although you are unemployed for 9-10 months of the year, you are still called employed because preference is given to employment. So a) it is too lenient, b) it has got too much of a preference.

We [at CMIE] don't do this. There is no preference in our system, and there is no leniency in our system. We ask, "Are you employed today?" And if you can't say about today because you've still got to go as a daily wage worker to find a job, we ask you "Were you employed yesterday?" The answer is 'yes' or 'no'. You were either employed or you were not employed. And there is no recall problem, there is no leniency, there is no preference. We are a lot more stringent and therefore our number turns out to be lower. We have a larger number of people who are out of the labour force.

How much of this is an impact of Covid-19? You have pointed out that we saw dramatic surges in both directions--unemployment shooting up by over 120 million and then dropping back dramatically to single digits in less than a year. So how much of this would be reversible?

Agricultural labourers will lose their jobs in agriculture as soon as the sowing period is over, which is delayed currently. And they will try to find jobs in daily wage situations. They will live in that very low-productivity area for a very long time till the economy picks up, [but] it's not going to be very easy for them to get out of that situation. But the farmers will find it very difficult to find gainful employment again because they've lost a salaried job mostly or they have lost a small enterprise they were running. They will find it very difficult to come back. So both agricultural labourers and farmers will find it difficult to come back to normalcy, and the normalcy [itself] wasn't good. So for us to not just reverse this migration but also go back to a higher growth trajectory which we require to get people into more gainful productive employment--it really is a tough long ride ahead.

We welcome feedback. Please write to We reserve the right to edit responses for language and grammar.