A good beginning: Payroll data can benefit from more transparency

It is high time the government moved away from a survey-based approach of assessing employment growth in the country, and opted for payroll data

The Union government last week released the first-ever estimate of payroll based on the Employees’ Provident Fund Organisation (EPFO) subscription database. The data showed that 3.11 million new jobs were added in the formal economy between September 2017 and February 2018. The EPFO manages social security funds of workers in the organised and the semi-organised sectors and has, at present, 60 million members with active contribution. The payroll count is essentially the difference between the number of workers who joined and exited from the EPFO's fold, and as such is seen as the net addition to jobs. The Ministry of Labour and Employment has said it will from now publish age-band-wise estimates of all new subscribers as declared by their subscribers.

There are many reasons to welcome this data release. For one, it is high time the government moved away from a survey-based approach of assessing employment growth in the country, and opted for the payroll data. India’s efforts to devise appropriate policies for improving employment have for long been undermined by its inability to get a clear understanding of the ground reality about the number of jobs being created. That’s because surveys conducted for this purpose were not only poorly timed, they also suffered from inherent flaws, ranging from definitional issues to problems of duplication. The other big reason for relief is that, on a pro-rata basis, this data suggests that 6.22 million new jobs had been created in the last financial year (2017-18). This is pretty close to an estimate given by a study done by State Bank of India (SBI) Chief Economic Advisor Soumya Kanti Ghosh and Indian Institute of Management (IIM) Bengaluru professor Pulak Ghosh in January. The study had pegged new jobs at 7 million for the full year 2017-18. 

However, there are still some niggling doubts about the quality of data. For instance, many have talked about data duplication. In any case, new 18- to 25-year-old EPFO members do not automatically mean new jobs in the economy. For example, when a firm employing 19 workers adds the 20th and gets registered on the EPFO database, the economy gains only one job, even though the payroll will show 20 additions. Besides, when an employee loses her job or stops working, her membership from the EPFO database is not removed automatically. The Ghosh & Ghosh study had also spoken about the shoddy maintenance of the EPFO data. For example, a vast majority of the accounts had only names and no other details about the subscribers. Another cause for worry about the payroll data is that during the time period concerned — September 2017 to February 2018 — the Centre for Monitoring Indian Economy’s (CMIE’s) survey showed that the unemployment rate went up from 4.7 per cent to 6 per cent. Though the two are not strictly comparable, the fact is that the wide variance in estimates does not inspire confidence. 

For building more confidence in the data, many experts say that the government should provide the data on the number of new establishments that registered on the EPFO database during this time period and the size distribution of these establishments. This will help observers understand whether the numbers reflect new jobs or simply new additions to the database.