PremiumThe Indian statistical system is passing through a challenging phase and facing some serious questions on its credibility. Three important issues generating turbulence in the system relate to data quality, non availability of some important data and inappropriate delay in the release of data that has already been collected and compiled, besides the lack of transparency and communication with data users. Quality assurance of the Indian data system was historically provided by its adherence to internationally accepted data definitions, unbiased data collection mechanism, standard errors within permissible limits and data disclosures in terms of standards of data dissemination system. While at times data may have not been to the users’ specifications and scarce, data quality issues were invariably not raised.
With the release of the new series of national accounts with 2011-12 as the base year in 2015, for the first time issues of data quality and data being biased were raised. Though the series at aggregate level was nearly aligned, its association was weak at the sector level. Overall gross domestic product (GDP) growth and the growth at sector level shows significant divergence, often at odds with the other indicators and user perceptions. Greater data validation and its disaggregated dissemination could have removed the apprehensions that seemed to persist. When the back series was finally uploaded, instead of creating consensus, it further raised the issue of being biased as disclosures were never appropriate. In 2019, when the National Sample Survey Office (NSSO) reported that a significant number of companies under MCA 21 were shell companies, GDP data quality issues were again raised, as growth was deliberately shown higher.
A second controversy, this time not related to data quality, surfaced when an NSSO report based on its periodic labour force survey was leaked. Clarifications were issued by the ministry of statistics and NITI Aayog that this was not an approved report. On 31 May 2019, ministry of statistics released the findings of the NSSO labour force survey report. It was a case of withholding information and timing its release for non-statistical reasons. Employment data are important macro data. Earlier reports by the NSSO and labour bureau on employment were questioned, as these were based on partial information and did not fully reflect the structure. The new survey was, therefore, launched to overcome this limitation. However, its leak in the first place, released later with a rider of incomparability with earlier surveys, created more confusion and a stigma of purposive use with vested interests.
The need for adequate, quality and timely data has been well recognized. The setting up of the National Statistical Commission as an independent think tank and expert body to examine data quality and provide third-party certification was a step in the right direction for ensuring data quality and data credibility. The conversion of survey and statistical organization of NSSO and CSO into offices and formations within the ministry of statistics is stated to develop synergy and internalize the inherent strengths of these two formation for data collection and its validation. However, this step has ensured that the commission has become another bureaucratic tier and the two offices of the NSSO and CSO two independent drifting outfits. In the process, while we continue to suffer from delayed and inadequate data (no new consumption survey after 2011-12, service price and service production data awaiting release for over a decade), data credibility has emerged as a new threat. Institutional strengthening, better coordination, quality involvement of stakeholders and a better disclosure hold the key to end this turbulence.
R. Gopalan and M.C. Singhi are, respectively, former secretary and senior adviser, department of economic affairs.
