Jobs debate: Ingenuity or insinuation?

With the changing nature of labour market, we need to change the definition of employment in the NSSO survey

The recent jobs debate reminds us of the seminal essay by Isaiah Berlin in 1953 on classifying people into foxes and hedgehogs. In the essay, the fox chases many incoherent ideas in its pursuit of too many ends but the hedgehog has only one goal to pursue. We would like to put ourselves in the hedgehog category while decoding the jobsdebate.

The World Bank (WB) estimates that the total labour force in the country (the percentage of both employed and unemployed population as of September 2018) was 527.5 million. With an unemployment rate of 3.5 per cent, this translates into an employed workforce of 509 million. Contrast this with CMIE estimates of 397 million workforce employed in December 2018. On a  7 per cent unemployment rate, this implies 424.5 million is the total stock of labour force (as against 527.5 million the World Bank estimates).

Against this total stock, what could be the minimum number of jobs that India could create in the non-farm sector (industry & services) on a yearly basis? Netting out the share of agriculture at 43 per cent (WB estimates) and taking a minimal 2 per cent as the flow from the stock of labour force (China at 2 per cent flow on a 686 million base is growing at 6 per cent per annum and producing 15 million jobs in a year, whereas the US at 1.5 per cent flow on a 163 million base is growing at 2.6 per cent and producing 2.5 million jobs in a year) even the CMIE estimates at 424.5 million gives us a number of at least 5.3 million (as per WB this is at 6.6 million). Clearly, with an economy that is growing at 7 per cent on a yearly basis even on the least favourable base, India is creating no less than 5 million jobs a year. And please note, these are jobs that could be counted.

Illustration: Ajay Mohanty

Now, let us address the current controversy regarding the leaked PLFS-NSSO survey. It may be noted that to overcome issues of existing EUS-NSSO surveys like representation, periodicity and timeliness, the Ministry of Statistics and Programme Implementation (MoSPI) decided to undertake the Periodic Labour Force Surveys (PLFS) from 2017-18.  According to this leaked report, the unemployment rate in the country was 6.1 per cent in 2017-18, at its highest level since 1972-73. The report also provides unemployment rates for rural and urban areas with male and female categories in the age group of 15-29 years, which are almost two to three times more than the 2011-12 rates (based on EUS survey)

We believe the survey results must be released in the public domain without immediate delay for a fair and just debate about the numbers. Irrespective of when these are released, we would however want to raise the following points about the survey.  

First, the methodology adopted for the PLFS. It is a two-year pilot started in July 2017 based on education level of households where larger weights are assigned to households having higher number of 10th pass members above 15 years, whereas the EUS is based on expenditure (urban) or livelihood (rural) of households. Thus, PLFS and EUS are not strictly comparable.

Second, publishing the results for the age group of 15-29 years might result in erroneous unemployment numbers bucket-wise. Take this example: A typical urban youth reaches the age of 21/22 years at the time of graduation and 23/24 years at post-graduation. If he/she pursues further studies, add two to four extra years. Thus, computing the unemployment rate for the age group of 15-29 will give erroneous results as many are likely to remain unemployed till the age of 22/23. It will be, therefore, interesting to see the unemployment rates in differing age buckets between 15 and 29 years. We believe that there might be a significant decline in unemployment rates as the age band increases beyond, say 30, reflecting the aspiration of today’s youth.

Third, the meaning of employment is changing fast. We are in a gig economy where jobs and earning are no longer same — at least to the millennials. This change of perception in earnings and job is critical and any survey needs to reflect this new phenomenon. Hence, if a surveyor simply asks “are you employed”, the immediate answer will be an emphatic “no”, since employment means permanent salary every month. If one asks the same person, if his/her income is “zero”, again the answer will be an emphatic “no”, implying he/she has income.  This is the real fallacy of any employment survey and hence the need to move to payroll rather than employment survey (like the latest CII payroll survey). Ideally, such payroll survey and the payroll reporting of formal jobs through EPFO need to talk to each other to develop the methodology and survey questions. For example, payroll reporting from EPFO gives insight on which sector is creating the maximum or minimum jobs. Payroll survey should focus on these sectors when they do the stratified sampling and give weightage accordingly.

With the changing nature of labour market and need for specialised skilled people, we need to change the very definition of employment in the NSSO survey to reflect the reality. 

The writers are Group Chief Economic Advisor, State Bank of India and Professor, IIMB. Views are personal