There have been a few issues raised in the public domain, including in this paper (“A misleading story of job creation”, Jan. 22), regarding our report, ‘Towards a Payroll Reporting in India’. We believe they must be addressed. This is the first study to dig out raw data and use big data analytics to construct a payroll report in India. Previous studies used the survey-based method and extrapolation, which is arguably flawed.
Disproving claims
The first issue relates to the fact that new registrations in the age group of 18-25 years with the Employees’ Provident Fund Organisation (EPFO) are not those who have got new jobs but are primarily a byproduct of formalisation initiatives by the forces of demonetisation (FY-2017) and the Goods and Services Tax (FY-2018). However, this is merely a claim or a hypothesis that is not substantiated by any data, but has instead been inferred from many surveys during these periods.
First, we considered only those who specifically joined jobs during that year in the age group of 22-25 years (most jobs in the 18-25 age group are clustered around 22). They are hence strictly first-time employees. Also, we only considered those first-time employees who had been making continuous non-zero contributions since their date of joining.
Second, even for the sake of argument, if we assume that these jobs were already there, it means that only people in the age group of 18-25 years lost jobs during demonetisation and thereby got formalised and no one else in other age groups lost their jobs. And if such a hypothesis existed, we were definitely under-reporting payroll in India for both the formal and informal sectors, and that is precisely our bone of contention for creating a better reporting of payroll.
Third, it may be possible that during any year, existing employees who are below 19 years would come newly into the EPF and add to the payroll. However, this is a normal event every year and we believe that this figure is not material on net basis as we have excluded first payments from those who are 25 years and above and who could have also joined the payroll for the first time (the number of such outstanding payrolls is as much as 54.8 lakh).
In a similar vein, it has been pointed out that job losses as a part of GST cannot be captured by EPFO. Again, even if we assume that there have been job losses, we need to mention that our objective is to capture first-time payroll reports through EPFO. By definition, EPFO is not a platform to estimate hypothetical job losses, but new payroll capture through real-time data analytics. We would be happy if our celebrated labour economists can construct a measure of such thorough use of big data analytics and not specifically rely on market surveys.
Fourth, it has been also argued that when an employee loses her job, her membership from the EPFO account is not removed automatically and hence the payroll number could be an overestimation. This is a misconstrued argument as we have worked strictly with only active EPFO accounts and those who are making continuous contributions since the date of joining till the time we did our analysis. In fact, we adjusted 4.2 crore individuals data from our database: Even a single detail missing for an account meant that the account details were incomplete and hence the account was not considered for our purpose. All those who joined under the Amnesty scheme were not considered. Nor were those who made no contributions. We also divided our data into two subsets of new employees with continuous contributions and old employees for arriving at the FY-17 March stock, and then we worked on the new employees only. All this has been elaborated in our report and it is surprising that this was not read.
Our stock of payroll for March 2017 is 9.2 crore, which is much less than the National Sample Survey Organisation estimates. This means that our estimates were purely conservative. Also, the argument that EPFO contributions have jumped many times in FY-2017 and hence create a false narrative is not correct, as this is exactly the point that we are making but for a different reason. There were 10 million EPFO additions under the Amnesty scheme during January and June of 2017 and this has resulted in a surge in EPFO contributions, but we have not taken this into consideration for our analysis.
Also, we must remember that about 1.5-1.8 million people retire every year, creating vacancies for new hires, and hence 7 million is the gross figure. But the fact that there were this number of new employees on payroll stands. Young people are getting jobs and contributing to Employees Provident Fund and Employees’ State Insurance and that is what the data shows. Even though we appreciate the enormity of India’s job challenge, we believe that the debate about jobs should move to adequate reporting, skilling and compensating our labour force. Until our well known labour economists understand this simple logic, they will continue to say that data analytics has gone berserk!
Soumya Kanti Ghosh is Group Chief Economic Advisor, State Bank of India, and Pulak Ghosh is a Professor at IIM Bangalore. Views are personal
