DisFact #19: The politics of India's jobs data

Happy Sunday, readers!

Welcome to DisFact, a weekly newsletter about Indian politics, policy and the economy. I am Samarth Bansal. If you enjoy this newsletter, please forward it to a friend. If you’ve been forwarded the newsletter, here is the signup link. Here is the list of all previous issues.

There is a cliché in political commentary that goes by: “good economics is bad politics” and vice-versa. But creating millions of jobs, as Modi had promised in the 2014 campaign, is both good economics and good politics.

But Modi government was unable to do so. New data from a “leaked” official report shows that the unemployment rate in India is at a four-decade high. The government will continue to spin the numbers, but the figures should settle the debate.

Before diving into data politics, the focus of today’s newsletter, let’s be clear: The jobs crisis in India is not because of Modi—it is a deeper structural problem confronting the Indian political economy. That he unleashed the needless demonetisation scheme—which most economists now agree was a disaster—may have aggravated the problem. And now, far from being honest and transparent, Modi government is obfuscating facts about the state of the Indian labour market, and in the process, degrading the credibility of India’s statistical institutions. That is problematic.

Where are the jobs? Where is the data?

What: The Business Standard published key findings from the official jobs report that the government did not want you to read: at 6.1%, the unemployment rate in India in 2017-18 stands at a four-decade high. In 2011-12, the unemployment rate 2.2 per cent.

The report paints a dismal picture of India’s labour market and shows that the Modi government was unable to keep up its promise of creating millions of jobs. Expectedly, the government tried to bury this report for around two months and withheld its publication.

That was one of the reasons why the last two independent members of the National Statistical Commission (NSC) resigned on Monday, who felt they were not “taken seriously and being sidelined by the government.” NSC is an autonomous body which was set up in 2006 to add credibility to India’s official statistics.

Thank to Business Standard’s Somesh Jha, my former colleague at The Hindu, who scooped out this important report.

Other key takeaways: The joblessness rate among youth was at a significantly higher level compared to the previous years and “much higher compared to that in the overall population.”

The rate of joblessness among rural males in the age group of 15-29 years jumped more than three times to 17.4 per cent in 2017-18 compared to 5 per cent in 2011-12. Similarly, the unemployment rate for female youth in rural areas stood at 13.6 per cent in 2017-18 compared to 4.8 per cent in 2011-12.

Read the full story here

What data is this: The figures come from the first Periodic Labour Force Survey, or PLFS 2017-18. It was conducted by the National Sample Survey Office (NSSO) which functions under the aegis of Ministry of Statistics and Programme Implementation. Earlier, the NSSO conducted an employment survey every five years. For more regular and effective monitoring, the government decided to conduct the survey annually from April 2017—and it was named PLFS.

What is NSSO? Data from NSSO is central to policymaking in India. It is the official source of key socio-economic indicators collected via large-scale representative sample surveys. Yes, there are general concerns about NSSO data collection apparatus—and India’s statistics in general—but bringing that up to discredit jobs data is dishonest. If that’s the argument, we can’t even measure our GDP, as a lot of the national income calculation is based on NSSO survey data.

What is the government saying: That the “leaked” report was just a draft report and further work was needed to be done before its release. Some are saying that the new data is not comparable to the previous one.

I love to get nerdy about methodological details, and it is crucial to so while looking at any dataset. But right now, we can’t as the public does not have access to the full report. What we do know is that the report itself stacked the unemployment rates of the new data with previous data going all the way back to 1972-73.

What other data tells us: Other evidence is in line with the survey findings. Data from the Centre for Monitoring Indian Economy (CMIE) shows that the “count of unemployed has been increasing steadily” and 11 million jobs were lost in 2018.

Data politics

Burying the NSSO report was the latest of the many attempts by the Modi government to engage in politics around jobs data. Here are others.

Attempt 1: Discredit CMIE data

Two weeks ago, Railway Minister Piyush Goyal questioned the credibility of the CMIE data mentioned above. Obviously, Mr Goyal doesn’t read this newsletter. But I still have some unsolicited advice: he should consult economist Bibek Debroy, the Chairman of Prime Minister’s Economic Advisory Council. In November 2016, almost a year before he joined the council, Debroy called CMIE’s unemployment product “respectable” and “a treasure trove of data” in an opinion piece for the Indian Express.

Attempt 2: Create hype by pointing to misleading data — EPFO

In October 2017, Modi claimed—and every government spokesperson continues to do so—that the rise in the monthly payroll subscribers of Employees’ Provident Fund Organisation’s (EPFO) is proof of new employment.

What is EPFO data?

An understanding of formal employment in the Indian economy can be gleaned from the data which rests with the Employee Provident Fund Organisation, the Employee State Insurance Corp. and the National Pension System. These jobs are considered ‘formal’ because employers provide some kind of social security to their employees. The EPFO data captures companies with more than 20 employees. (BloombergQuint)

EPFO data is not a good metric to gauge overall employment: Increase in EPFO numbers doesn’t necessarily mean that the total employment has increased. In July 2017, for instance, the jump in EPFO subscribers was attributed to a government amnesty scheme which allowed firms to come clean on their actual staff strength without being penalised. Experts say that the jump in EPFO data is not new employment—merely enrolment of employed persons into EPFO.

The mysterious “independent study”: In a televised interview, Modi boasted about findings from an “independent study” that analysed EPFO data and claimed that 7 million people would be added to the payroll in 2017-18. That can’t happen without new jobs being created, Modi claimed.

Again, journalist Somesh Jha found out that this so-called “independent” study was actually triggered by a request from the Prime Minister’s Office (PMO). In fact, the NITI Aayog helped the study’s authors to get access to EPFO data—the “first time that the EPFO data on employees was made available to a select group of non-government researchers”.

Attempt 3: Point to everything but comprehensive data

Here is how you spin the data. Some arguments from government officials:

  • Cab aggregators Uber and Ola are creating jobs that are not captured by stats. (Okay, sure: But the whole point of NSSO surveys is to capture the entire spectrum and all kinds of jobs. That’s why it is a household survey.)

  • How can GDP grow at 7-8% and not create enough jobs? (Boss, that’s exactly what economists call the “jobless growth” phenomenon)

  • A huge amount of loans disbursed under the Mudra scheme shows that the economy created jobs (Fact: 90% of the loans under Mudra are up to Rs 50k—if that create jobs, it should have done under previous regimes as well. Read here)

Attempt 4 (prediction): Bring out new data

The government will publish a new report that will show a massive boost in employment under Modi. In fact, Bibek Debroy, the chairman of the Prime Minister’s Economic Advisory Council, said that new round of the NSSO survey will be announced soon and that he his “sure that that particular survey will show that there has been substantial employment and substantial job creation”. Excellent.

Same old story: The politics around jobs data is similar to the latest controversy surrounding the GDP data. That’s another tangent—and statistically more complicated—but the relevant bit for this discussion is that we now have three back-dated GDP time series. We need back series data to compare GDP growth trends over time. Of the three series, two show an upward revision of GDP growth during the Congress-led UPA government; one shows downward revision for UPA. No need to guess: the government accepted the latter, which shows that the growth rate under the Modi government was higher than the UPA.

Ideally, this is supposed to be a purely technical discussion. Trained statisticians are best equipped to tell us which data is most reliable. But it is difficult to ignore the political context in which the numbers are operating.

Why data politics is problematic: Obviously, politics around jobs data is happening because the publication of the report won’t fare well on the government’s resume. But it would be problematic if the government genuinely believes that the labour market is doing fine. Moreover, delaying the publication of stats for political gains sets a bad precedent and raises questions about the credibility of India’s economic data, which has long-term effects. We need good data for informed policy making as well as to infuse confidence in investors to take India seriously.

Talk to me

Comments? Feedback? Suggestions? Write to me at samarthbansal42@gmail.com or hit reply to this email. And if you find this helpful, please spread the word. Thank you!