The report says that these companies are not traceable. But they have been used in GDP calculations?
These companies would have submitted their balance sheet data to MCA. But when the NSSO
carried out the survey and took sampling frame from the list of active companies, these were found to be shell companies. Now, if a legitimate company is using a benami company to hide a part of its transactions, those transactions don't just go away, nor are they illegitimate. They are part of GDP. If we don't measure that, we would be systematically underestimating GDP.
Is the Central Statistics Office (CSO) then blowing up the data of these shell companies to arrive at its estimate of GDP?
As far as the MCA database is concerned, there were 1.3 million companies initially. When we started there were around 900,000 active companies. Active companies are those that have filed their balance sheets in the three past years. Now, for the purpose of GDP calculations, a cut-off date is taken (around the beginning of December), so all companies that have submitted data prior to that are taken into consideration. Some companies submit their data after the cut-off date, but it would not be proper to assume that they are not in business. So the estimates arrived at are then blown up by a factor that takes into account the paid-up capital of all active firms. As some of these companies are shell companies, their data are also taken into account and blown up. But if these shell companies do not submit their balance sheets for three years continuously, they are automatically removed from the list of active companies.
How much does the MCA21 database account for in the services gross value added? What are the other sources of estimation?
If my memory serves me right, it varies from service to service. MCA21 covers practically the entire financial services sector but a fairly small part of wholesale and retail trade. On an average, it would be around 45 per cent of all services. The rest of the sector — the non-corporate — is estimated from the enterprise surveys carried out by NSSO every 5 years. In the intervening years, the estimates are based either on proxy indicators or on extrapolation.
Given the controversy surrounding the MCA database, should the ministry in your view consider giving access to researchers to unit level data?
This is not in the hands of the Ministry of Statistics and Programme Implementation (MoSPI). The custodian of MCA21 is the Ministry of Corporate Affairs (MoCA), and they have refused to make the data public on confidentiality grounds. In fact, MoSPI itself does not have access to the unit-level data and relies on MoCA for some of the data in a structured format.