It is a capital mistake to theorise before you have data, British author Arthur Conan Doyle one said.
Nowhere is this truer than in India. Attempts by the government to sell a sparkling story on the jobs front – when statistics point otherwise – is no less than a valiant effort at selling snake oil.
The stubborn effort to run down the government’s own data as inadequate is to deny the evidential value of national statistics.
Studies carried out by the National Sample Survey Office (NSSO) are based on methodologies that are robust, time-tested and in line with methods adopted internationally for job data collection. The current government, however, has frowned down upon this and refused to release the results of one periodic labour force survey (PLFS).
What is instead put forth as proxies are industry data, industry-sponsored surveys and anecdotes of how Mudra loans are creating a hunky-dory job picture. Perhaps there is a belief that if evidence can be trumpeted to be absent, it is possible to argue that the problem is absent – though that is unimpeachably untrue.
The NSSO survey is based on random sampling, which is an unbiased technique for estimating the behaviour or characteristic of an entire population from a chosen sample. Every individual/household is chosen randomly so that each unit has the same probability of being chosen. Average sample here will accurately represent the population. Hence, based on the sample, valid conclusions can be drawn for the entire population.
But the PLFS report emerging from this survey has not been released on the ground that it is a draft study – and will need further approval before it is made public. It is true that a laid-down process should be followed. But it appears unlikely that the survey will be replaced by a new study or that the results will change dramatically.
Also read: A (Failed) Quest to Obtain India’s Missing Jobs Data
Meanwhile, other figures are bandied about to create a fog of confusion. It has also not helped that participants in India’s nightly TV news debates are people who have little knowledge of statistics or the actual subject.
The current rebuttals to the leaked NSSO report are a CII survey of 1,00,00 MSME firms which shows 3.2 lakh net jobs have been created in the last four years.
Another employment narrative is that over two million jobs have been created by new tech firms like Uber and Ola and that 40 million Mudra loans have been given to first-timers. Simply put, there are many jobs that have been created, but they can’t quite be counted.
To understand the broader context, we need to take a step back. First of all, employment and unemployment rates are arrived at on the basis of people who are employed and people who are actively seeking employment, respectively. If someone is self-employed but still actively looking for a job, he will show up as unemployed even though he could nominally be self-employed. Labour Bureau statistics show that in two-thirds of the states, self-employed get less than Rs 7,500 per month and 42% of them get below Rs 5,000.
Most kinds of self-employment, therefore, are not real jobs but ragtag jigs and do not provide either a steady life or living wages or social security. Hence, many ostensibly self-employed show up when better employment is advertised.
In the best of situations, jobs are a part of the ‘creative destruction’ process. Some jobs get created and some get destroyed and the net accretion is the addition to jobs. Industry reports will not capture this. Moreover, there is a selection and survival bias as only those industries which have added jobs are taken up for a survey. Industries which have folded up don’t figure in the calculus.
It is only in random sampling that one can capture these details. One can also capture other vital information such as whether the Uber drivers were earlier driving a taxi, auto or doing some other job. This would indicate mobility in the value chain but not the net addition of jobs.
The other piece of the job puzzle is India’s payroll data. The EPFO database showing an ‘increase’ of 20 million jobs is perhaps because the definition of EPF being mandatory for firms having 20 workers changed to firms with ten workers and all units which were earlier not under the purview of the worker security net came under its fold. This, as other experts have noted, is the formalisation of jobs although some new units (with ten workers) may figure in this picture. As for 40 million Mudra loans, an average of Rs 60,000 per unit often do not create jobs but only enable enlargement of existing units to slightly bigger ones.
Also read: In India, Rising Joblessness is a Tinderbox Waiting to Catch Fire
Why is a survey based on random sampling better than payroll statistics in India? This is because we have a dualistic economy with a large informal component and a smaller formal component. Payroll stats are good barometers only when the economy is largely formal. Households and farming disguise unemployment and underemployment and they are best captured in the random sample survey.
The passionate intensity in the debate regarding employment comes from those spokespeople who despite their burnished CVs and high-decibel arguments do not understand the basics of statistics, surveying and the nuances between different reporting formats. They can be forgiven because they don’t have the knowledge nor the basic tool kit and their passionate intensity is the surround sound of talking from a pre-determined script.
But what about the more knowledgeable people, who have chosen to keep quiet? The reason is that they want to put up a show that their version is not divergent from the prepared script. There is a huge incentive to keep quiet or mumble here and there rather than take on the issue squarely. Much worse, there is a disincentive for telling the truth beyond the diktats of apparatchiks of the establishment. The latter feels that any figure which does not support the narrative of relentless growth will go against the government.
The blame for low job creation, in a country as diverse as India, cannot solely be placed on the Central government. However, collecting and releasing truthful statistics is certainly an important duty. This is the starting point for course correction. But as long as you do not recognise the data, nothing can be measured and no correctives can be launched.
Satya Mohanty is a Former Secretary to GOI and Adjunct Professor, JMI.