After junking report on consumption, NSO now looks for reasons3 min read . Updated: 22 Nov 2019, 12:17 AM IST
- NSO has asked all regional offices to highlight issues raised by field inspection staff during the survey
- An international group of 201 economists put out a statement protesting the report’s suppression
NEW DELHI : India’s statistics ministry has sought critical field inspection reports pertaining to a junked consumption survey on a “most urgent" basis, a move seen by many statisticians as an attempt at digging up dirt on the report to justify its scrapping.
Statisticians including former top government officials said the move appears to be an attempt to find grounds for burying the report after already having rejected its findings over “data quality issues".
The quinquennial survey, conducted in 2017-18 to measure consumption of households, showed a decline in average real consumption spending from the previous survey round in 2011-12. The latest survey, which was supposed to be released in June, was delayed and eventually junked on 15 November, after leaked portions of the report were published by Business Standard.
On 18 November, the field operations division of the survey arm of National Statistical Office (NSO) sent a letter to all regional offices asking them to send a detailed report highlighting the “important/critical/alarming issues" that had been raised by field inspection staff during the consumption survey. Mint has reviewed a copy of the letter. All field offices have been asked to send such reports on or before Friday.
“This is a very unusual step, and seems to be an attempt to find fault in the survey post-facto, after having decided that the survey report has to be junked," said Pronab Sen, former chief statistician of India. “The usual process is that regional offices conduct routine inspections during a survey, and in case any data discrepancies are found, they re-survey those households. This is supposed to be taken care of in the survey period itself."
Agreeing with Sen, P.C. Mohanan, former chairman of National Sample Survey Office (NSSO, now clubbed under NSO) said the timing of the move was unusual. Mohanan had resigned from the National Statistical Commission last year over the suppression of yet another survey report, which had documented rising unemployment.
“They (NSO officials) may want to produce a technical paper on the survey listing the problems encountered during field work," said Mohanan. “Data collection problems in consumption surveys are nothing new. In any case, such inputs should have been sought while drafting the report, and not after a decision on the release of the report is taken."
Questions sent to the chief statistician on this issue on Wednesday evening remained unanswered till press time. The officer in charge of the field operations division, which sent out the letter, declined to comment on this issue.
The ministry’s decision to junk the consumption expenditure survey report has faced criticism from both within and outside India’s statistical establishment. A group of 201 economists from around the globe has issued a statement protesting the suppression of the report and seeking the release of the raw underlying data, so that it may be discussed and debated.
While the ministry has cited an adverse report from an expert committee that studied the consumption survey’s findings to justify the decision to junk the report, the expert committee did not actually recommend its suppression. This was first reported by Business Standard on Thursday.
“The committee was asked to investigate data quality issues, and it made some observations regarding that by checking the internal validity of the data, and based on comparisons with other datasets," said a person with direct knowledge of the matter, on condition of anonymity. “The committee did not ask for the report to be withheld."
“In any survey or any data-gathering exercise, there will be data quality issues," the person cited above said. “In this case, the committee highlighted that for a few richer decile classes, the survey findings may not have been entirely accurate, and suggested some refinements for future surveys. And the committee also noted the discrepancy between consumption survey findings and the national account statistics (consumption figures, as computed during the calculations of the gross domestic product). But this discrepancy has existed for long, and which of the two data sources (national accounts or survey) is more reliable remains a matter of debate."
The discrepancy between survey data and national accounts has been a matter of debate since at least the 1980s. In a landmark 1988 study, economist B.S. Minhas examined this issue, and argued that there is no reason to privilege the national account estimates of consumption over estimates based on NSSO consumption surveys. Minhas’s findings led the Planning Commission to abandon its practice of scaling-up consumption survey estimates to match the national account figures.
Examining this issue several years later in 2005, Nobel winning economist Angus Deaton noted that Minhas’ observations held true even then.
“India’s experience...demonstrates that when surveys and national accounts disagree, it should never be assumed that one or the other is correct," Deaton wrote in his widely cited research paper on India’s consumption and poverty numbers.