Forcing conclusions upon the EPFO data PDF Print E-mail
Friday, 27 April 2018 03:56
AddThis Social Bookmark Button

Shobhana edit

The data is too fraught with problems and one-offs to be used to make conclusions about India's robust jobs growth


With the EPFO just releasing data to show 3.1 million subscribers were added in the last six months, government economists like NITI Aayog chairman Rajiv Kumar are extrapolating this to say 6.2 million formal-sector jobs were created in FY18; “Cassandras should please give up”, he tweeted. Earlier this year, Pulak Ghosh of IIM Bangalore and Soumya Kanti Ghosh of SBI came to some very similar conclusions—4.5 million new jobs in FY17 at least—that came in handy for a poll-bound government accused of jobless growth. The numbers, though, could be far from reality.

While this newspaper has also believed that the poor jobs creation data reported by the labour bureau—4.2 lakh jobs in FY17—didn’t reveal the true picture, drawing conclusions on employment from the EPFO database is fraught with risk. For one, the 6.2 million new jobs fly in the face of both investment and consumption trends. The rise in private final consumption expenditure has decelerated over the past several quarters and there has been virtually no pick-up in investments for close to two years now. The performance of the jobs-intensive export sector has also been very modest; if there is an uptick in FY18, it comes off a weak base in previous years.

The conclusions from the EPFO data appear forced for a variety of reasons. If you assume 6.2 million new subscribers on a base of 60 million EPFO “active subscribers” (those who made at least one EPF contribution during the year), this translates into an incredible 10% of the base being added in just the last one year. If you choose to use the “average subscriber” base—the EPFO gets this by dividing the total number of contributions in a year by 12—of 47.5 million in FY18, this means as much as 12.5% of subscribers came in over the past year. The numbers look even more curious if you consider the fact that 80-85% of jobs in the country are typically in the informal sector; if 6.2 million formal jobs, are we saying 35-40 million jobs were created last year? Or are we saying there has been a dramatic formalisation, post-GST and DeMo, and that, say, around 10-15 million jobs were created? To add to the mystery of how many subscribers the EPFO has, apart from the “active” and “average”, there are 33 million accounts that are “Aadhaar-seeded”.

There is also the impact of the 2016-17 amnesty as a result of which around 4.9 million people joined the EPFO. If one considers the “average” number of EPFO subscribers of 37.6 million in FY16 and 41.2 million in FY17 and adjusts for the addition through the amnesty, the effective subscriber base falls to 36.2 million in FY17 or lower than that in FY16! And, to use Radhicka Kapoor of ICRIER’s example, if the number of employees in an enterprise increases from 18 to 22, the EPFO system will record this as 22 new jobs since units with less than 20 employees do not pay EPFO.

With Aadhaar seeding, the EPFO data will get cleaner over time, but given the disparities in older data, meaningful comparisons are difficult. Serious economists, within and outside the government, also need to explain the inconsistencies across not just various jobs datasets but also GDP and other growth data. The KLEMs data released last month, for instance, showed the workforce had contracted between FY14 and FY16 to 482.7 million with about 1.2 million jobs being lost at a time when the economy was growing at a fairly fast clip. No meaningful policy debate or action is possible till such time there is reasonable clarity on the actual situation.

Last Updated ( Friday, 27 April 2018 04:05 )

You are here  : Home Miscellaneous Forcing conclusions upon the EPFO data