Once a globally renowned institution, India’s Central Statistics Office (CSO) today faces the gravest credibility crisis in its history. An investigation by Mint has found that a key database of companies picked up in 2015 as an input source to calculate its new series of gross domestic product (GDP) is riddled with gaps, large enough to distort final output figures. It was bad enough that serious infirmities had been pointed out in other aspects of the revised calculation exercise. The GDP deflator now in use to reduce the nominal size of the economy to its real figure may have misrepresented overall inflation, for example; secrecy over this deflator had cast some doubt over our new GDP growth numbers, drawn now from inflation-sensitive measures of value-addition. The latest revelations seem to confirm other suspicions of the new methodology. As Mint reported on Wednesday, a significant share of companies in the aforementioned database could not be traced by field staff of the National Sample Survey Office (NSSO) involved in a year-long hunt that ended in mid-2017. Other firms had been misclassified; since groups of registered firms in specific sectors are used as proxies for commercial activity in the informal sector, this must have had distortive effects of its own. The field staff also noted that several firms in the database had dodgy accounts.
This is not the first time the authenticity of such input data has been challenged. Indeed, many statisticians were iffy about the MCA-21 database in question. But it has not been put to such close scrutiny before. The CSO has so far been unable to provide a credible response to criticism of the poor quality of corporate information being used. What is crucial at this juncture for the CSO to win back credibility is to throw open everything that has gone into its new calculations for independent statisticians to examine. The office must put out far more detailed documents to explain exactly how it has gone about its growth estimation process, as was suggested by NITI Aayog vice-chairman Rajiv Kumar in a 2015 research paper. It must also release the MCA-21 database in its raw form so that others can analyse those numbers. A statistical audit of our GDP numbers by experts who are not involved in the original process would go a long way toward quelling doubts. Without such openness, India’s official statistics will remain under a cloud, which would do the country no good.
Apart from assuring the country a more robust oversight mechanism that would ensure the reliability of our official statistics, other ideas of transparency could be adopted as well. As suggested by some economists and statisticians in the past, the CSO could make it a standard practice to issue detailed explanations for any revisions of GDP data, especially if these revisions are large. Perhaps other structural changes are also required that would reduce the scope for arbitrary tweaks of formulae. The entire episode has exposed a weakness that goes to the heart of economic management. As a fast-emerging economy that must sustain its rise under ever more complex circumstances, India cannot afford to formulate policies and action plans on the basis of indicators that risk turning out to be false. Our numbers need a thorough clean-up.