‘Regular audit of core statistics essential’

Interview with R.B. Barman, former Chairperson, National Statistical Commission.

R.B. BARMAN: “The NSC should have real independence and be empowered with the authority to audit all core statistics transparently.”

R.B. Barman, former Chairperson ofthe National Statistical Commission (NSC), was member of the Indian Statistical Service before joining the Reserve Bank of India in 1979. He was president of the Indian Econometric Society in 2006-07, Vice Chairman, Irving Fisher Committee on Central Bank Statistics, Bank for International Settlement, Basel, Switzerland, and Member, International Data Forum. He sent a paper to Prime Minister Narendra Modi in June 2018 on how to improve official statistics but received no response. The gross domestic product (GDP) figures were contested in 2015, and the NSC asked the Central Statistics Office (CSO) to make a presentation, but it was ignored. Barman spoke to Frontline about the concerns involving transparency in data dissemination and the role of statistical agencies in the controversy surrounding the release of job data. Excerpts:

Why was there a problem about the back series GDP data? There were two back series, the second one conducted by NITI Aayog. Considering that it is not a statistical body, is it the correct thing to do?

There is a major change in the methodology for compilation of GDP in the new series with the base year 2011-12. The Ministry of Corporate Affairs 21 [MCA21] data on the corporate sector are a major improvement, but such data were not available for the period before 2009. There were other changes as reported in “Changes in Methodology and Data Sources in the New Series of National Accounts Base Year 2011-12” published by the CSO. However, with the help of a systematic approach to maintenance of data, which could relate Annual Survey of Industry Data with Corporate Sector Data, the problem of absence of MCA21 data for the earlier period could have been sorted out to a large extent for manufacturing sector estimates for back series. There are possibilities of many other data quality checking for other sectors, as mentioned in my article titled “Indian Official Statistics: Digital Transformation to Honour Citizens”, published in Economic & Political Weekly (June 30, 2018). If such a system had existed, it would not have taken such a long time to release back series, raising a controversy. We need to strengthen our system for collection and maintenance of data, improve the database on modern lines for their collation, relate these data using spatial big data warehouse, and improve timeliness, quality, consistency and coherence to inject greater robustness.

The first one [back series] was a part of a report of a committee headed by Sudipto Mundle and the CSO provided support in the form of Member Secretary to the committee. It mentioned how the exercise was undertaken and spelt out the limitations. One should not go beyond the candid admission made in the report, which was for public comment. Both Sudipto Mundle and N.R. Bhanumurthy have explained their position. The second was an official release. It is subject to public scrutiny. The role of NITI Aayog has no precedence and has been questioned. GDP figures were contested right from 2015 when the new series was released. The NSC asked the CSO to make a presentation before the Commission. The CSO ignored the NSC. As the NSC does not have administrative powers, no action could be taken.

Is the NSC’s role only to guide the National Sample Survey Office (NSSO)? What kind of challenges and problems did you encounter as NSC Chairman?

Official statistics is a public good that informs, supports and sustains democracy and advances socio-economic development. The NSC’s role is to see that the system provides good quality data on an impartial basis to “honour citizens”. The Rangarajan Commission recommendation, based on which the NSC was created in 2006, was of the view that “an independent statistical authority free from political interference having power to set priorities with respect to core statistics is needed to ensure quality standards of statistical process”. Data on GDP, inflation, employment, trade, etc., are part of core statistics. We tried to work within the ambit of our responsibility as spelt out in the gazette notification of 2006 setting up the NSC. The public can draw their inference about the effectiveness of the NSC from the reaction of the government to criticism on GDP and employment data. The NSC should have real independence and be empowered with the authority to audit all core statistics transparently.

The Minister informed Parliament that the NSSO was still processing data from July to December 2018 and that the report of unemployment rate at 6.1 per cent in 2017-18 cent was fake. He also said that it was the periodic labour force survey (PLFS) of the NSSO and that the surveys were still being done. He said the job of the NSC was only to guide the government. Is that so?

It is not true. The first fundamental principle for official statistics, accepted by the government, states that “official statistics provide an indispensable element of the information system of a democratic society, serving the government, the economy, and the public with data about the economic, demographic, social, and environment situation”. As governments are judged by their performance reflected in official statistics, among others, they cannot interfere with the independence of the NSC.

While there is a need for the big data warehouse as you write, will that address the problem of confidence faced over data authenticity?

Yes, if we can relate micro data with macro aggregates at different levels of aggregation, it will allow for imposing a code of practice to audit the entire system to check how the estimates meet the stipulated quality standards. Not only this, the ability to understand and analyse the nexus relating real sector, financial sector and fiscal sector data will help the economy grow at a faster pace.

Has there been a situation similar to the present one with the government withholding employment data?

I am not aware of any previous situation resulting in the withholding of the results of National Sample Survey conducted by the NSSO, after approval by the NSC or the Governing Council. To maintain integrity and credibility of statistics, the system should be independent of the government, as democracy expects. As for the interference with NSS data, I am not aware of any such incident in the past. The NSSO is backed by the expertise of the working group/standing committee consisting of top experts from the fields of social sciences and statistics and a robust system.

What then should governments look forward to?

My only observation is that there should be a system of regular audit of core statistics, including National Accounts and State Accounts, to maintain integrity and comparability of such data.

The NSC needs to be insulated from political interference. Is there any way the data that was meant to be released can be done by Parliament?

If an independent statistical authority is vested with production of core statistics by an Act, giving it legal backing as followed in many countries of the West, Parliament could ask them to explain anything as provided under such an Act. But Parliament should not approve finalisation of data. That will go too far and will be against the fundamental principles laid out by the United Nations in 2014 and accepted by the government by a gazette notification on June 15, 2016.

What plagues the Indian statistical system? Why is it facing a crisis of credibility today?

Annual Reports of the NSC for 2016-17 and 2017-18 were supposed to be placed on the table of Parliament. We are not aware of what action the government has taken on issues raised in these two reports. Even the report of the Rangarajan Commission, which was set up by the A.B. Vajpayee government, has been implemented in a half-hearted manner. The proposed act to empower the NSC was drafted but not enacted by the United Progressive Alliance government. As a result, the NSC does not have the legal backing necessary for it to derive the authority. Hence, the Indian statistical system suffers from many inadequacies. Look at the statistical system of the United Kingdom and the authority the U.K. Statistics Authority enjoys from its 2007 Act. In the U.K., statistical offices are gradually getting detached from Ministries for being independent in the real sense. Here, the government, through the Minister, controls officers of the Indian statistical system and expects every report to be cleared by the government.

Several private entities do their surveys on employment and industries, the Centre for Monitoring the Indian Economy (CMIE) for instance. How robust are these agencies, especially on information on employment.

I have seen a presentation by CMIE on their system for collection of employment data. They cover a large sample size regularly and the methodology is in the public domain. If one has to criticise such a method, it has to be specific. I have not seen such a criticism so far. In the absence of such an exercise, the criticism of the CMIE data does not get the backing of objective assessment. In the U.K., under the provisions ofthe code of practice followed by the U.K. Statistics Authority, there is a system for accepting any data as part of national statistics. We do not have such a system.