The disappointing results of the NSSO survey meant that two detailed reports based on it had to be junked. (Pradeep Gaur/Mint)
The disappointing results of the NSSO survey meant that two detailed reports based on it had to be junked. (Pradeep Gaur/Mint)

NSSO’s latest report a setback in effort to improve state GDP data

  • The NSSO survey was supposed to fill gaps, but given the large number of out-of-survey units, the findings won’t be used
  • The design for the Annual Survey of Services Sector survey, due to be rolled out early next year, is yet to be finalized

NEW DELHI/MUMBAI: The disappointing results of the National Sample Survey Office (NSSO) 74th round field survey of services firms have impaired India’s chances of improving the accuracy of state-level estimates of gross domestic product, or GSDP, according to economists and statisticians.

The data issues in GSDP estimation matter because the changes in the methodology have led to wide discrepancies across states. This has impacted the borrowing limits for states, as well as the likely share of federal funds that they could receive (see Plain Facts).

The advisory council of the 15th Finance Commission is set to discuss these issues on Thursday, two people with direct knowledge of the matter said on condition of anonymity.

Mint reported last week that a key database introduced in India’s new GDP series, MCA-21, has been found to be full of gaps, citing a “technical report" prepared by NSSO on the basis of the 74th round survey.

The report showed that 16.4% of companies in the MCA-21 database were either non-traceable or closed, and another 21.4% were “out of coverage" or misclassified. The disappointing results of the survey meant that two detailed reports based on the survey had to be junked.

Documents accessed by Mint show that official statisticians were banking on the survey results to remedy the problems with gross state domestic product figures.

Queries regarding the accuracy of GSDP numbers mailed to chief statistician Pravin Srivastava and senior Central Statistics Office (CSO) officials on Saturday evening have not elicited a response so far.

As reported by Mint earlier (“How India’s statistical system was crippled", 8 May), Ahmedabad-based economist and member of the monetary policy committee of the Reserve Bank of India Ravindra H. Dholakia had first raised questions about the new GSDP figures.

In a paper co-authored with Manish Pandya of the Gujarat directorate of economics and statistics and published in the journal of the Pune-based Indian School of Political Economy (ISPE) in 2017, Dholakia pointed out that the flawed use of the MCA-21 database had distorted GSDP figures for Gujarat.

At a 2018 conference held by ISPE in Pune to discuss these issues, statisticians from a few other states also raised questions about the new GDP series, and the manner in which the MCA-21 database was used to “allocate" state-level estimates from national estimates in a top-down manner.

In her response to these criticisms, CSO’s deputy director general at the time, T. Rajeshwari, acknowledged the limitations of the MCA-21 database and the lack of state-level information, and said the results of the 74th round survey of services firms would help bring about a significant improvement in the quality of the GSDP estimates by allowing states to generate their own bottom-up estimates of the corporate sector.

“The NSS 74th round (prelude to the Annual Survey of Services Sector)... is a first survey of its kind, covering large business establishments of services sector using list frame," said a concept paper presented by Rajeshwari at the conference. “In future, this survey would enable states to compile estimates of private corporate sector."

The discouraging results of the 74th round survey meant that hope had been belied, said J. Dennis Rajakumar, director of Economic and Political Weekly Research Foundation, who was also a participant at the Pune conference.

“Given the limited state-level information in the MCA-21 database, CSO uses proxies to allocate the national figures to states," Rajakumar said. “For the industrial sector, state-level estimates from the Annual Survey of Industries data are used to allocate valued added in output. But for services, where there is no such survey available, there is a problem."

The 74th round services survey was supposed to fill this gap but given the large number of out-of-survey units, the survey findings won’t be used for estimation.

The results of the survey, apart from raising serious questions about the MCA-21 database, have raised doubts over the proposed Annual Survey of Services Sector survey, which was supposed to follow this survey and use the MCA-21 database as its sampling frame.

The design for the Annual Survey of Services Sector survey, due to be rolled out early next year, is yet to be finalized.

There is little clarity on the way forward considering the limitations of the MCA-21 database highlighted by the 74th round technical report, a person with direct knowledge of the matter said on condition of anonymity.

“The way forward lies in discussions," said former chief statistician Pronab Sen, who heads the technical committee for Annual Survey of Services Sector. “We are still discussing."

Close