Amid many criticisms against India’s statistical system and the call for a revamp to enhance the credibility and utility of data gathered and disseminated, Rajeeva Laxman Karandikar, chairperson, National Statistical Commission, tells Priyansh Verma and KG Narendranath where the fault-lines lie and suggest ways to fix the problems. The NSC has recently been empowered to function as the governing council for the National Sample Survey Organisation (NSSO).

Q. How do you respond to the criticism that the National Statistical Commission (NSC) hasn’t lived up to its mandate? Your responsibility includes evolving policies, priorities and standards in statistical matters, but this role is largely being performed by the ministry of statistics and programme implementation (MoSPI) , while other central departments seem to act on their own behalf…

The NSC was formed in July, 2005 as per the Rangarajan Committee recommendations, which reviewed the Indian statistical system. It was done by a gazette notification and it was planned that in a year’s time it would be made statutory by an Act of Parliament. But this has not happened. So the answer to your question is yes, NSC has not lived up to its mandate as it appears in the gazette notification, but the reason is clear, it has not been empowered to do so.

Many issues that depend crucially on statistical matters are dealt by relevant government entities, including judiciary, without any reference to NSC. Let me mention one – several examinations conduced by governmental bodies for admission to institutions or for jobs are conducted in multiple shifts and this requires a process known as “normalisation”. This is purely a statistical issue and each agency ends up creating its own “formula”, perhaps unaware of NSC. We have urged the government to either empower NSC to fulfil its mandate or disband it.

Q. We are witnessing multiple issues about India’s statistical system – regarding data/sample collection, authenticity and credibility of data, (infrequent) periodicity of surveys, alleged suppression of data etc…

Let me take up one by one. Data/sample collection methodology needs to improve as NSSO- the wing under MoSPI that does this task – has been using methodology that was envisaged ages ago and they are reluctant to accepting any change. NSSO must recognise that in current times, several sources of data are available and they must change their functioning to use the same. NSC has set up expert committees to examine these issues and we hope this will lead to improvement.

As for authenticity and credibility of data, I have not seen anything specific and so I cannot comment. What I have seen are questions raised about methodology – and it is clear that poor methodology can lead to poor data quality. So once the methodology is reworked this will be addressed.

As for infrequent periodicity of surveys, all I can say is that it is not the task of a statistician to decide periodicity. That is for the end user to decide and to make available to MoSPI the resources – manpower and money – to do so. As for alleged suppression of data – I will not comment. Let me also add that the NSC was also acting as Governing Council of NSSO. While initially there was a steering committee that was empowered to examine issues in depth, seek corrections etc. and then the final report was approved and presented to NSC for formal approval. Over the years, the steering committee was disbanded and the final report was being presented to NSC for formal approval. This issue has been under discussion for last one year and now the steering committee has been constituted and NSC can take up the task of overseeing the NSSO activities, ensuring changes that are necessary for overall improvement.

Q. What are the issues with methodology of data collection?

Within the real of methodology, there are issues pertaining to sampling which I have mentioned above. Then there are issues with the questionnaire, which are so long and tenuous. The NSSO enumerators take more than two hours to finish an interaction with a respondent in many of the surveys conducted by NSSO. This is “meaningless” as respondents would find it tiring and it could affect the verity of information gathered. Having informally talked to various ISS (Indian Statistical Service (ISS) officers across the country, it is clear that this is more of a problem in urban areas. Long and lengthy questionnaire raises doubts on the reliability of the data collected. The questionnaire has to be shortened drastically.

Q. What changes should MoSPI make with regards to conducting surveys?

Currently, the MoSPI seems to decide on its own which surveys to undertake. This is not a good practice. They do take into account various wishlists they receive from ministries. But I feel that the “end user” (government ministry) should be involved at all stages in deciding the questions which need to be asked and the frequency or periodicity. I will also add another matter – not really related to conducting survey but equally important. In view of the volume and complexity of data collected, it is clear that adequate up-to-date IT (Information Technology) infrastructure is necessary for the all the activities that MoSPI is supposed to carry out, and not having the same is adding to delays in processing the data. We’ve alerted the government on the issue.

Q. Is the recently released fact sheet on household consumption expenditure reflecting the true picture of consumption?

I have not seen the said fact sheet and so I will not comment. I will add the following: issues that I have talked about – namely sampling methodology and length of questionnaire together do raise the question – is the conclusion reflecting the true picture?

Q. Should private agencies be involved in collecting data for the government?

I feel there is no problem with that provided the process is monitored at each stage by the end user. The tender document for award of the contract (of conducting the survey) should be formulated by an in-house team of experts (of the government), and not through a third party, as may be happening currently.