It is assumed that once employed, we will have work related to employment to do. But, there is no definitional link that necessitates work to be done to qualify as an employed person. You could be employed and not have any work to do.
Mohanan and Kar point out that a lot of what was counted as employment during the lockdown was not associated with any work and, therefore, the employment estimates overstate the level of production activity during the lockdown.
We agree. Deploying time-use data in the preparation of national accounts as suggested by the authors and recommended by the System of National Accounts (SNA) 2008 in estimating production levels would be a great, even revolutionary, move. Consumer Pyramids Household Survey (CPHS) now provides time-use data along with employment data. And, the point being made by Mohanan and Kar is borne out by this data. The proportion of employed persons who worked eight hours or more for pay or profit dropped from 80 per cent in Wave 18 of September-December 2019 to 55 per cent in Wave 20. Wave 20 was conducted entirely during the lockdown months of May-August 2020.
Interestingly, there have been people even before the lockdown who claimed to have done no work although they were employed. In Wave 18, 0.5 per cent of the employed respondents stated so. During the lockdown the proportion of employed persons who said that they spent zero hours doing any work for pay or profit shot up to a significant 8 per cent.
Given the sharp drop in work among employed people seen above, deploying time-use data in estimating the level of production would be much better than assuming that all employment is the same during normal times and lockdown times.
But, constructing national accounts statistics is not the only objective of conducting labour surveys. In this context, there is nothing erroneous in the employment data collected through traditional means even during the lockdown. Avoidable misinterpretations do not obviate the need to generate the data because these are useful in applications other than national accounts. Used in conjunction with other data including time-use and wages they yield exceptionally good insights into the conditions of the labour markets.
Policy implications for an economy where labour has employment but no work are very different from an economy where labour has neither employment nor work. The former is a better position to be in till it can last. It gives policy makers a window of opportunity to arrange work before enterprises boot out labour because there was no work for a prolonged period of time.
The lockdown has broken many traditional assumptions in data collection and interpretation. The now broken assumption that employment implies work is just one. Another broken assumption is that employment implies wages. People have jobs but no work and wages. Others have jobs that provide work but no wages. The lockdown has challenged the link between jobs, employment, work and wages. But, national accounts is not the only casualty.
Mohanan and Kar clarify that jobs refer to positions. A person is employed to occupy a position offered by a job and perform the work associated with it. Jobs are, therefore, not persons. Data on jobs are best collected through an enterprise survey and not a household survey. A household survey can provide estimates of employment. These are persons in jobs. But there could be jobs that are vacant. A household survey cannot tell us about job vacancies. An enterprise survey can. And, an enterprise survey cannot inform us about unemployed labour. Only a household survey can. The two are, therefore, complementary.
A household survey provides insights into the supply of labour and an enterprise survey provides insights into the demand for labour. The two together help create an environment to take informed decisions for labour markets to clear and also help create better national accounts.