Payroll Data: Chidambaram’s Carping Has Some Validity, But Ultimate Joke Is On Him

Ever since two economists, Pulak Ghosh, professor at IIM Bangalore, and Soumya Kanti Ghosh, group chief economic adviser at the State Bank of India, came up with an estimate of new jobs created using payroll data, serious efforts are on to debunk their research. Their “mistake” was to say that some seven million jobs may get created this year (2017-18), based on additions to Employees' Provident Fund Organisation (EPFO), Employees' State Insurance Corporation (ESIC) and the National Pension Scheme (NPS) subscriptions during the year so far. If they had said that seven million jobs had been lost, they would have been hailed as heroes worthy of a Padma Shri by non-National Democratic Alliance (NDA) parties.

The research study (see here) by Ghosh and Ghosh was essentially making a fervent plea for collecting better payroll data. They were not claiming a net increase in economy-wide jobs, and their limited extrapolations on payrolls were also extremely conservative. For example, they took only first-time payroll joinees in the 18-25 age cohort, and all of them should have made positive contributions to the fund since joining. They also excluded the numbers that came into EPFO accounts as a result of the amnesty scheme given to employers in 2017.

However, it is also not fair to dismiss all criticism as unworthy of discussion, especially if it gives us a direction on how jobs data should be collected and collated in future.

Former finance minister P Chidambaram joined the carping when he criticised the Prime Minister for including the self-employed, including a pakoda seller who may have taken a Mudra loan, in the definition of jobs. “Even selling pakodas is a 'job', said PM. By that logic, even begging is a job. Let's count poor or disabled persons who are forced to beg for a living as 'employed' people,” Chidambaram tweeted. By making this elitist statement, he may only have disclosed his disdain for the humble “pakoda seller”, but the point to note is that we do indeed need a proper definition of jobs.

The problem is clearly one of defining the word carefully. Broadly, it should mean a contractual relationship between a formal employer and employee, and which involves someone getting regular pay, and some social security benefits in lieu of work. Then we essentially need to get better data from the formal sector, from governments to public sector and private sector businesses, and professionals like chartered accountants and lawyers, among others. We also need to define whether someone working part-time, or on a contract, has a job. In my view, we should view this as jobs, even though we could label them as vulnerable jobs, jobs that are real for the moment, but not secure.

This data will need to be “cleaned” through a process of Aadhaar or other forms of de-duplication, so that someone shifting jobs and opening another PF account is not counted as holding two “jobs”. Professionals who work as CAs, or doctors or lawyers also need de-duplication, since many of them could be working with big firms or hospitals, which would already be part of the EPFO network. We can only add those professionals who work independently, and who may not be part of EPFO. As an aside, one must point out that the “independent” doctor or CA is no different from the self-employed pakoda seller except in terms of professional qualifications. The pakoda seller’s expertise may be self-taught, but if she is any good at her pakodas, she is a limited professional too. If what she does is not a “job”, and she is classified as self-employed, or owner as a micro business, we may need to do the same with CAs and lawyers working on their own.

So, the first point is simple: we need to gather data for payrolls (which constitute jobs), for partial payrolls (contract workers, part-time workers), and for capturing various forms of self-employment, including single-person businesses, and self-employment in agriculture.

The second major criticism of the Ghosh-and-Ghosh data is that it may not be a net addition to jobs. All that may have happened is that people who were anyway employed but not part of EPFO or ESIC are now becoming a part of it.

This criticism is valid, and needs further investigation. However, the joke is really on the critics. If, as they say, this increase in payrolls over the last two years is merely the result of non-formal jobs going formal, then the government deserves its share of credit for it.

Two reasons why.

One, if 9-10 crore is the total number of jobs which form part of recognisable payrolls (Ghosh and Ghosh say the stock of existing EPFO, ESIC and NPS subscribers plus the new joinees estimated by them add up to 10 crore), it follows that 90 per cent of Indian jobs being in the informal sector is a myth.

Chidambaram sought to punch holes in the research, by writing in The Indian Express: “In the article, the authors reported that the total ‘payroll’ stock as on March 31, 2017, was 919 lakh. So, it has taken the country 70 years to create a ‘payroll’ stock of 919 lakh jobs but, miraculously, in just 12 months, the country will generate 70 lakh new ‘payroll’ jobs — that is nearly 7.5 per cent of the current stock!”

While Chidambaram can question whether change of this magnitude is possible in such a short time, the simple answer to his query is that when a tipping point is reached, change can be exponential. We saw the spike in growth rates under UPA-1 which had almost nothing to do with what the Manmohan Singh government did in terms of reforms. It was the result of years of reforms done earlier, and an uptick in coordinated global growth.

Chidambaram could well have asked how poverty rates, which fell so slowly in the first 60 years after Independence, suddenly fell sharply during two particular years of the United Progressive Alliance (UPA) regime, despite the lack of job creation. According to National Sample Survey Office reports, in just two years - 2009-10 and 2010-11 - India’s poverty rate declined by 7.8 per cent, a feat never achieved before. It could have been a statistical blip, but could also have been the result of strong growth and a steady pumping of resources into the rural sector that started with Atal Behari Vajpayee and accelerated during the UPA.

The question to consider is whether some tipping point was reached during Modi’s fourth year in government, which tipped the scales in favour of more formal jobs. Ghosh and Ghosh estimate that of the total employment figure of around 504 million, 108 million are formal sector jobs. This means more than 20 per cent of jobs are now formal, where there is some amount of social security. The only thing we need to check is if their salary levels are enough to provide a reasonable livelihood.

Two, if even after excluding the EPFO amnesty scheme, formal sector payrolls are swelling by seven million in 2017-18, we need to ask ourselves two additional questions: is this because more informal sector workers are joining the formal sector, or is it because of exogenous shocks to the informal sector. Like, for example, demonetisation and the launch of the goods and services tax from July 2017.

It is possible to speculate – but it needs empirical validation – that the pressure to go formal began with demonetisation, and the pressure continues with GST. GST is levied on a firm’s value-added, and value-added is lower (and hence the taxes lower) if your costs (including labour costs) are higher, and captured in the formal set of books. It is reasonable to assume that those who want to be part of the GST network and benefit from input tax credits would want to also formalise their real cost structures inside their books rather than keep them outside.

It is interesting to note that Rajeev Gowda, one of the Congress party’s articulate Rajya Sabha MPs from Karnataka, argued some time back that the focus must be on organised sector jobs. Writing in The Hindu in December 2016, he sought to debunk the argument that NDA-1 under Vajpayee created more jobs than UPA when growth was higher. Gowda wrote: “If one needs to look at empirical data for jobs numbers, one indicator is provident fund (PF) registrations. Most permanent salary-paying jobs in the formal sector entail PF payments to employees. The UPA-1 period had an increase of nearly two million PF registrations compared to less than a million during the NDA-1 period.”

Maybe, Chidambaram, and the critics of the Ghosh and Ghosh method of payroll job accounting, need to settle their arguments about EPFO data with an intellectual wrestling match where all parties agree not to change the goal-posts in case the data goes against their wishes.

Whether you think jobs grew under the Modi government or not, one thing is clear: formal sector jobs are rising significantly, whether due to what the government did or some other unknown variable. And that is a big gain for the economy.

For now, despite some valid criticisms of the Ghosh and Ghosh data, the joke is really on Chidambaram.