How India Can Make Official Statistics Credible: Former Data Chief

Bengaluru: “There is no such report,” said India’s energy minister Dharmendra Pradhan on April 12, 2019, about a leaked National Sample Survey Office (NSSO) report that said unemployment in India was at a 45-year high.

Five million men lost their jobs between 2016 and 2018, according to the 2019 State of Working India report by Azim Premji University.

P C Mohanan, former acting chairman of the organisation that approved the report, has now told us: “The National Statistical Commission (NSC) approved this report, it is reflected in its minutes of the meeting. The minister is possibly not aware of this.”

There was much controversy when the Business Standard reported at four-decade-high unemployment rate of 6.1%, based on a leaked report of the Periodic Labour Force Survey (PLFS) conducted by the NSSO between July 2017-June 2018.

A few days before the leak, on January 28, 2019, P C Mohanan, acting chairman of the NSC and another member, J V Meenakshi, resigned, protesting the government’s suppression of the PLFS. The NITI Aayog, a government think tank, and Arun Jaitley, finance minister, classified the report as a “draft” while Dharmendra Pradhan, central energy minister denied the report even existed.

There was a feeling the commission was “being side-lined in key statistical matters”, Mohanan told IndiaSpend. This may be the “first known instance” where an NSSO report was not released because “its findings are not matching the narrative articulated by the government”, he added. The NSC must be reconstituted at the earliest and steps must be taken to empower it, said Mohanan.

Mohanan, 63, has been a government statistician for over 30 years and was a member of the NSC since June 2017 till his resignation. He was involved in unemployment, migration and healthcare surveys at the NSSO. Mohanan was a member of the expert committee on agricultural statistics, estimating housing shortage, and an evaluation committee appointed after the 2006 Sachar report, which studied the condition of Muslims.

In an email interview, he talked about the controversy surrounding the report, government interference in the functioning of the commission and the need to reinforce India’s data-collection system.

Multiple surveys have concluded that jobs is the most important election issue in 2019. Although the issue of unemployment may not be new, do you see anything different in terms of the numbers unemployed or the kind of unemployment reported during this government’s tenure compared to previous governments?

Unemployment has always been a concern of the central government since the beginning of the five-year plans, each of which accepted employment generation as one of the key objectives. The issue took a back seat in public discourse after the economic liberalisation [in 1991] and the ensuing high growth of the economy.

The current debate on unemployment has happened in a slightly different context. The last year for which official survey data are available is 2011-12. We did not have any issue with the survey reports in the past. It was well accepted that the unemployment rates did not show any spurt in the past and the employment growth was not in tune with the economic growth resulting in the coining of the term ‘jobless growth’.

Meanwhile, the NSSO, the main government survey agency for data collection, had switched to collecting data on employment on an annual basis through the PLFS. The first of the annual report for the year 2017-18 was slated to be released in December 2018.

We also had a couple of other large scale surveys, one by the ministry of labour and the other by centre for monitoring Indian economy [a private data collection organization]. These two surveys had been showing an increasing level of unemployment.

The PLFS report also indicated a rising unemployment rate and lower work participation rates in its 2017-18 report, to an extent giving a different picture of employment-unemployment than what the NSSO has been reporting in its previous surveys.

Union energy minister Dharmendra Pradhan said that “there is no such report”, when recently asked about the leaked NSSO report. Your comments?

The NSSO surveys are done openly, though it is possible that given its low profile, many are not familiar with it. The work program and the calendar for release of reports are decided by the NSC, which has complete technical oversight of NSSO. The NSC has approved the report [and] is reflected in its meeting minutes. The union energy minister is possibly not aware of this.

What transpired over the few months before you decided to quit in protest over delay in releasing the jobs report?

The job report was not the only issue that led to our resignation from the NSC. The commission was set up as the apex body for all official statistics. However, it was the feeling in the commission that it was being side-lined in key statistical matters, including the launching of major statistical initiatives. The non-releasing of the employment report was the last key issue where the NSC found its role diluted.

Have you faced similar interference in the release of data during your time with the government?

Possibly this is the first known instance where an NSSO report that is part of a series of similar reports was not released because its findings are not matching the narrative articulated by the government.

Certain sensitive information like religious or mother tongue data etcetera are not usually released to public by census authorities, but never a survey report on a common topic like employment.

The NITI Aayog jumped in defence of the government. Vice chairman of NITI Aayog, Rajiv Kumar, said that the four-decade high in unemployment was “not verified”, while CEO Amitabh Kant wrote that comparing the previous employment-unemployment surveys and the periodic labour force survey (PLFS) “is erroneous” and that the “report does not have my approval”. How do you assess the role of NITI Aayog in this matter? Does this undermine institutional credibility of organisations like the National Statistics Commission?

In the scheme of things, NITI Aayog has no role in the production or dissemination of official statistics, except those relating to its functioning. The approval of NSSO reports is within the mandate of NSC and no other agency has any role in it. The NITI Ayog commenting on the NSSO report as erroneous or draft was very unusual. Once released all users including government agencies have a right to comment on it. Bringing out these deviations in the case of employment report have perhaps only enhanced the credibility of the NSC.

During the press briefing on the leaked report, NITI Aayog said that Ola and Uber alone have added more 2 million jobs in India. One of the criticisms of the gig--or temporary jobs--economy has been that it may not be generating fresh employment and that it has a problem of “rotating attrition”. Your comments?

In the NSSO only the current employment details are collected and not the previous employment history. So one cannot really say how much of the employment is fresh or how many people have lost their jobs etc. The survey measures the level of employment and unemployment.

Rotating attrition mentioned by you is quite possible since these types of employment is seen mostly in large towns, which have higher job mobility than small towns or villages. Given the nature of the jobs offered by the gig economy one does expect large scale job rotation.

What trends do you see in the employment scenario in India based on the NSSO report? What role did demonetisation have in exacerbating rural distress and do data reveal a causal link? Are there positive trends that you observed in the data?

The numbers published by some newspapers from the latest survey report show major changes in the scenario. Specially because all the previous surveys have been showing only a very low open unemployment [people willing to work have no jobs] and reported very gradual changes in the employment pattern (It was only in the early 1970s that NSSO had reported high unemployment rates). However some trends like declining work participation of women have been observed in earlier surveys also.

As for the effect of demonetisation on employment, please remember that NSSO estimates give only the level parameters from which one can assess the distress. But the survey cannot pinpoint this to any event like demonetisation as the survey was not designed to assess its impact.

There are certainly positive trends like the increasing years spent by youth in education leading to a qualitative improvement of the workforce. The report indicates a much more dynamic employment situation. More analysis can only be done once the report and the unit level data are available to researchers for detailed analysis.

Recently 108 economists and social scientists working across the world issued a letter on government suppression of data which said that “economic statistics are a public good” and that “the Indian statistics and the institutions associated with it have however come under a cloud for being influenced and indeed even controlled by political considerations”. What, in your view, must be done immediately to stop such interference, and to strengthen institutional credibility?

Basically the Indian statistical system is built on sound principles with a specialised common cadre of officers managing the central agencies. It is considered as a highly competent body among national statistical systems by international agencies. A strong and professional statistical system can defend itself, though unfortunately it has not come forward to do so amidst the raging controversy over job data.

On the immediate issues raised by the groups of eminent economists, [former Reserve Bank of India governor] Raghuram Rajan recently suggested that a group of independent experts should take a look at the gross domestic product (GDP) compilation and database and methodology, if required. This is a good suggestion to re-establish the credibility of the GDP figures and make the process more transparent.

Presently the NSC does not exist. The NSC should be reconstituted at the earliest and steps taken to empower it with adequate resources as an independent watchdog so that the questions raised by us in our resignation letters are fully addressed. NSC should be the sole agency responsible for official statistics and reporting only to the parliament. In the long term NSC should be a statutory body.

However, the human and organisational resources of the statistical agencies have not improved since the 1980s. Considering that the Indian economy is now the [world’s] fifth largest and the population is second largest, the statistical agencies require strengthening, if they are to measure the economy and population characteristics with improved timeliness and accuracy.

Lastly, it has become necessary not only to enhance the human resources of the central statistical office and the national sample survey organization but revamp its internal architecture to adapt to the changing data needs and data handling procedures, and most importantly, the challenging field conditions faced by investigators.

(Paliath is an analyst with IndiaSpend.)

We welcome feedback. Please write to respond@indiaspend.org. We reserve the right to edit responses for language and grammar.