ADVERTISEMENT

Former National Statistical Commission Chief Questions Scrapping Of Consumption Survey

There is no technical justification for scrapping the survey, says former NSC chief PC Mohanan.

Customers and pedestrians pass clothes hanging for sale at a market in New Delhi, India, on Sunday, July 7, 2019. Photographer: Anindito Mukherjee/Bloomberg
Customers and pedestrians pass clothes hanging for sale at a market in New Delhi, India, on Sunday, July 7, 2019. Photographer: Anindito Mukherjee/Bloomberg

The government’s decision to withhold the household consumer expenditure survey on technical grounds is unconvincing, according to PC Mohanan, former acting chairperson of the National Statistical Commission.

The survey, according to a Business Standard report, showed that monthly per capita consumption expenditure fell 3.7 percent between 2011-12 and 2017-18. The government, subsequently, said that it was withholding the survey due to data quality issues.

Mohanan, in an email interview to BloombergQuint, said that even in the unlikely event of data quality issues in the survey, a report should still have been brought out stating the findings. Mohanan added that there is no technical justification for scrapping the survey, while adding that even an appearance of government intervention in economic data is a matter of concern for data users. “Economic statistics is the basis for business decisions and the lack of it or losing faith in it will have serious consequences for investment decisions,” he said.

Edited excerpts below:

Is the government’s decision to scrap the last round of consumption survey justified?

The consumption survey, like any other National Sample Survey, is planned meticulously under the guidance of a working group with external and internal experts. NSSO has been doing the consumption surveys since its inception in the 1950s.

Scrapping the survey after the report has been drafted is very puzzling because there are well established procedures for data checking and cleaning during and after the survey. If there were data quality issues it would have been discovered long before the report is drafted. If there were any major data related issues a report still should have been brought out stating all the findings. Thus no technical justification for scrapping the survey after all operations are complete can be thought of.

Is the government intervention in economic statistics a cause of worry?

It is accepted by all and also adopted by UN in the fundamental principles of official statistics that official statistics should be produced independently strictly according to technical considerations. Even an appearance of government intervention would be a matter of concern for data users.

Economic statistics is the basis for business decisions and the lack of it or losing faith in it will have serious consequences for investment decisions.

If a survey is inconsistent with other available data points, how should the government / NSSO deal with it?

The survey methodology for gathering data cannot be faulted as these are based on sound statistical theory. The problems can arise when the respondents fail to provide the correct information or the survey personnel fail to ask the right questions or when the survey questionnaire does not have all the relevant questions. All these are very important issues. Solution is not junking the survey but analysing the reasons and reporting the same. Data users can then decide on the extent to which they can use the data. Having a professional organisation with permanent survey staff is intended to obviate such possibilities.

Very clearly the Statistical Office has gone for the simple way out viz. junking the survey and refusing to come out with any clarifications on the problems. This is a highly unprofessional way of dealing with the problem. A technical report on the survey ought to be brought out. In fact the enterprise survey done during the previous year by NSSO using the MCA-21 list of companies had problems due to most companies not existing at the given address. No meaningful estimate could be compiled from the data. When this was brought to the notice of the National Statistical Commission, which I chaired at that time, it was decided to publish a technical report highlighting all the problems. This was done by NSSO, and resulted in a lot of debates on the use of MCA data for GDP estimation. It clearly showed the inadequacy of using administrative lists for data collection.

I am extremely worried about the credibility loss of Indian government statistics and the morale it will have on the large number of statistical staff working in the government when their efforts are negated.

Will it now be tough to rebase GDP and update the CPI index if we don’t have the survey?

Consumption survey results have only a small role in GDP rebasing as most of the required information comes from elsewhere. Those items for which this survey results are used can be approximated from earlier surveys or other indicators.

The rebasing of CPI depends to a large extent on the changes in the consumption basket of households, which comes from this survey. It is not clear how and when the CPI rebasing will be taken up.