NSSO’s Services Sector Study: What It Means (And Does Not Mean) For India’s GDP Data

A technical study on India’s services sector conducted by the National Sample Survey Organisation has sparked a debate on whether a large number of ‘shell’ companies detected are skewing India’s gross domestic product data.

The NSSO’s ‘Technical Study On India’s Services Sector’ threw up a large proportion of ‘out-of-survey’ services companies in the MCA-21 database used for the computation of GDP data. The finding was first highlighted by Mint newspaper on Wednesday morning.

In the course of the study, the NSSO found the following:

• 38.7 percent of a sample of 35,456 units in the services sector were found to be 'out-of-survey' when the MCA-21 database was used. This includes closed, out-of-coverage, non-traceable units.
• Another 6.8 percent of the firms were tagged as ‘casualty cases’, which includes firms that did not respond.
• This proportion of ‘out-of survey’ firms was lower in earlier databases such as the economic census (12 percent) and the business register (22.1 percent).

Economists and statisticians that BloombergQuint spoke to are divided about what this finding means for India’s GDP data, which uses the MCA-21 database.

Pronab Sen, former chief statistician of India said the finding is unlikely to have led to any anomalies in calculation of GDP. PC Mohanan, former acting chairperson of the National Statistical Commission also felt that the government’s statistical department may have accounted for the existence of shell companies. Professor R Nagaraj of the Indira Gandhi Institute of Development and Research, however, believed the finding could suggest overestimation of GDP.

Calls to Pravin Srivastava, chief statistician of the country were not unanswered. TCA Anant, former chief statistician who led the effort to create a new GDP series, declined to comment.

An edited summary of their views is below:

Pronab Sen, Former Chief Statistician of India

The study was intended to get operating data of services companies. The frame was taken from the active companies on MCA-21 and a sample was drawn. Of the sample, about 36 percent or so are non-traceable.

What this means is that there are a large number of shell companies. This is not news to anybody. The government has been actively weeding out shell companies over the last few years. So what this suggests is that the number of shell companies, particularly in the services sector, are large.

The question is does this lead to an over-estimation of GDP? The answer is - it does not. The reason for that is that shell companies do not create value on their own but they do store value which is created by some other company, which is carrying out benami transactions. But that value is created. It is just not shown in the books of some other company but on the books of the shell company.

If I don’t capture that, I will actually be under-estimating GDP.

What might happen due to shell companies reflecting value which is being created by other firms? It could, for example, lead to value being attributed to the services sector when it is actually being created in the manufacturing sector.

Watch the conversation here:

PC Mohanan, Former Acting Chairperson, National Statistical Commission

Anomalies like this can be found in many databases.

What is important is how you adjust for this anomaly when calculating the GDP data. Even in the ASI (Annual Survey of Industries) we found factories which were closed or some which were registered but were not producing anything, so these issues are accounted for.

R Nagaraj, Indira Gandhi Institute of Development Research

When an official organisation could not find a response from 45 percent of this sample [MCA-21], it is a serious issue. Just because there are one million active companies, it does not mean that they are all working. We know the reality of the corporate sector where there are thousands of shell companies of various kinds, which are untraceable, or engaged in little productive activity (as reported by the media).

When the universe [of data] is poor and uncertain, you are purely attributing some value-addition to them just because they are on the MCA list.

We went over this issue in official committees stating that the MCA and CSO have to give us more information and conduct further cross-checking before using these numbers for calculating GDP.