Payroll data raises serious questions4 min read . Updated: 20 Jun 2018, 03:40 PM IST
It will become more meaningful and useful if its reporting format is extended to reconcile the stock and flow measures
In April, the ministry of statistics and programme implementation released payroll data based on administrative records of the Employees’ Provident Fund Organisation (EPFO), Employees’ State Insurance Corporation (ESIC) and National Pension Scheme (NPS) for the first time. Following this, the NITI Aayog proclaimed quite emphatically in a press release that “the numbers from these three organizations are an eye opener and put an end to all speculations and conjectures regarding job creation in the economy". Unfortunately, analysis of the two monthly releases so far does not give one such conviction.
This new data series on India’s formal employment is, of course, path-breaking. Even though nearly 70% of the non-farm jobs in India are still in the informal sector (Economic Survey, 2018) and, therefore, survey-based employment measures can never be done away with, a definitive and high-frequency picture of formal employment would be of great value to economy watchers and policymakers. The data series also comes after a lot of initiatives by EPFO—like universal identifier and Aadhaar seeding of accounts—that have helped in de-duplicating records and cleaning up its database.
Much of the criticism so far against the use of administrative records for estimating generation of jobs has been regarding the mix-up of the effects of formalization (resulting in new establishment registrations) and jobs creation in such data. It was pointed out that when an establishment employing 19 persons earlier employed just one additional person, it becomes liable to register with EPFO and generates 20 new member registrations for EPFO. In its second monthly data release, however, EPFO has resolved this issue—by including another column on the number of new establishments during the month and, thereby, enabling analysts to adjust for the effects of new registrations. After this release, a State Bank of India research report estimated that about 15% of the increase seen in EPFO payroll data during September 2017 to March 2018 was due to formalization.
However, another issue with the payroll data—pertaining to the difficulty in reconciling the stock and flow numbers—is probably more serious. During the seven months for which the data has been released, EPFO added, on an average, 0.56 million subscribers every month. This is the number of new members enrolled with non-zero contributions, net of those who ceased to be members. When annualized, this number would mean a 13-15% rise in the stock of active EPFO members—something that is proving hard to rationalize with the other economic trends.
But the “stock" data, pertaining to the contributory members in EPFO in any month, has hardly moved in step. As per the labour ministry’s website, EPFO had 46.1 million contributory members in April vis-à-vis 45.5 million in November. There have been months when the number of contributory members has even fallen.
The “active" subscriber base of EPFO itself seems to be quite fluid. While the organization has more than 190 million membership accounts, less than a third are active. EPFO itself terms 60 million accounts as active, counting the ones with at least one month’s contribution during the year. The numbers seen in recent months are much lower than that liberal measure. An analysis in the recent Economic Survey that worked on a joint dataset of EPFO and ESIC had reported average active subscribers in both these schemes during April-November 2017 as 60 million with the peak in any month being 71 million.
These varied estimates and the fluctuating stock of EPFO membership, which cast some doubt over the quality of the formal jobs that are being analysed through these data, could possibly be on account of two factors: irregularity in remitting of contributions by establishments and the temporary nature of certain jobs. A footnote in EPFO’s data release states that “the estimates may include temporary employees whose contributions may not be continuous for the year". The data quoted above suggests that the proportion of such temporary employees may indeed be quite significant.
Discrepancy between stock and flow measures is more stark when it comes to the ESIC data. For ESIC, the statistics ministry’s data release itself gives data on the number of existing employees paying contributions during the month, in addition to the newly registered employees. The stock figure of existing ESIC contributors has actually fallen from 29.4 million in September to 26.3 million in March, although there were 7.1 million new registrations during the interim six months.
Any measure of employment generation, for being treated as a meaningful economic indicator and input for policymaking, cannot be seen simply in terms of inflow. What is important is the net addition to jobs after taking into account the job losses happening in the economy. The non-farm payroll indicator for the US, which is followed very closely by global financial markets, is also a measure of net change in employment stock every month. Incidentally, that number comes out of an establishment survey.
The payroll data for India based on administrative records will become more meaningful and useful if its reporting format is extended to reconcile the stock and flow measures. Some measure of persistency in the EPFO membership base—such as the number of members who have received contribution at least once in last three months and in last six months—would be useful and should be included by EPFO in its data release. The net change in a more core set of EPFO’s active membership could become a more useful indicator of formal employment, rather than simply adding up new members and declaring end to the jobs debate.
Mangesh Soman is a Mumbai-based corporate economist.
Comments are welcome at email@example.com