Only way was to compare job survey with 7-year-old census: S P Mukherjee

In a Q&A, the chairman of an expert group advising the govt, says data in the enterprise-based that which showed a 22% rise in factory jobs in April-June, was misrepresented

Many experts questioned the findings of the enterprise-based survey that reportedly showed that employment generation in the manufacturing sector rose by 22 per cent during April-June of the current financial year compared to the Economic Census of January, 2013-April, 2014. S P Mukherjee, chairman of an expert group that gave technical guidance to the government on enterprises-based job survey, tells Indivjal Dhasmana that there was misinterpretation of the data on this count. Twenty-two per cent more jobs in factories were not generated during April-June quarter, but those that existed as on April 1, 2021 were higher by 22 per cent compared to those during the time of economic census. Edited excerpts:  

The authenticity of the survey would have been enhanced had we compared it with the previous enterprises-based survey for the July-September quarter of 2017. Why did we not do it?

That comparison would have also meant going back a few years. Of course, the number of the years we'd have to go back would have been shorter had we compared it with the last round instead of revisiting the Economic Census of 2013-14. It was also not a year that was very close to 2021. Roughly four years have gone by. Secondly, that was a survey covering eight sectors, not nine, which was the case in the new survey. The earlier survey had slightly lower sample size with certain establishments not being there. However, the Economic Census had those establishments. Thirdly and most importantly, the method of estimation for the earlier survey was slightly deficient. I could not get hold of the records pertaining to the methods of estimation that time. Unless I am sure how those estimations were computed, I cannot compare the estimates given by the April 1 survey to the previous one. So far as the economic census is concerned, there was no question of estimates, as those were actual figures.

But how can one compare a survey with a census and that too when the latter was conducted 7-8 years ago?

A sample survey should be judged in terms of the estimates it provides for the total. The sample survey can't be considered well conducted, if its estimates are not close enough to the total. In this particular quarter, I can't make estimates for the total because I would have to conduct a census that way. So, what you say is not absolutely wrong, but not absolutely correct either. If I have to compare the findings of a sample and a census, these should theoretically be available for the same period. But they were not. If the complete count was available, who was foolish not to conduct a census? I agree with you that the comparison should have been with a closer period but there was none available. The census conducted last year is yet to be published.

When will we be able to make an ideal comparison of the establishments-based survey because the next survey that will come will cover the unorganised sector too?

No, it would be slightly later when the survey will cover the unorganised sector. The quarterly survey of April-June of the current financial year will be perfectly comparable with the survey that will come for the July-September of the year.  

Many economists are not able to understand the employment generated by the manufacturing sector because it was battered by the second Covid wave. But the survey showed 22 per cent job growth in factories this April-June period compared to the census of 2013-14. Your take?

There is no question of job generation. The figures reported were of total employment that existed in a particular sector and not generated during a particular period. These were the figures of employment in the manufacturing sector as on April 1, 2021. That figure was higher than reported in the economic census of 2013-14.

Is the skilled labour force responsible for vacancies that existed in various sectors during the period?

Not for all vacancies. About 187,000 vacancies were reported during the period. Some 39 per cent of vacancies reported were due to reasons other than retirement and resignation. In the category "others", the major reason was non-availability of skilled manpower.

How many more surveys on labour will now come?

Work on two others which are based on different themes are going on. One is on migrant workers, the other is on domestic workers. Field work has started for the survey on migrant workers and that for domestic workers will start soon. These are All-India Surveys and the field work will take some time to complete. The survey on migrant workers should come by June or July of the next financial year. That on domestic workers may come one or two quarters later.

When will other surveys come up?

The two other surveys--one on road transport workers, the other on workers engaged by professionals--were to be taken up. We did some work on the technical side. But since then, the government decided not to get this work done because the frame from which the sample was to be drawn was not available. For instance, the Bar Council of India was not obliging us with the list of all their members throughout the country. Then there was a problem whether this was to be taken up by the Labour Ministry or the Ministry of Statistics and Programme Implementation.