Data credibility at stake

Unfortunate fiasco over consumer spend survey

The credibility of the Indian statistical system has come under scrutiny yet again, after the government on Friday decided to scrap the consumer expenditure survey, conducted by the National Statistical Office (NSO) during 2017-18, because of “data quality issues”. The decision came after the findings of the withheld report were published in this newspaper. According to the survey, compared with 2011-12, monthly per capita consumer expenditure fell by 3.7 per cent in 2017-18 — the first decline in at least four decades. While spending in urban areas went up by just 2 per cent, it registered a massive 8.8 per cent decline in rural areas. The government’s decision to reject the report means India will not have a new poverty estimate until the next survey is undertaken. The revision in the gross domestic product (GDP) base year would also have to wait.

The findings of the survey and the way the government handled the issue have raised several concerns. First, since overall economic growth has slowed significantly since 2017-18, it is likely that consumption expenditure in rural areas might have slipped further. This is worrying and indicates a reversal in poverty reduction. Second, the government seems to have no faith in the data presented by the NSO. Simply rejecting the survey has raised questions over whether the decision was prompted by the government’s unwillingness to accept data that doesn’t suit its narrative. It is possible that the results were significantly affected by demonetisation and the introduction of the goods and services tax. The government had also withheld the periodic labour force survey, which showed the unemployment rate at 6.1 per cent, a 45-year high. The report was released with the same data after the general election results were announced.

Third, as this newspaper has reported, there is a significant difference between the Central Statistics Office and NSO estimates of consumption. While the difference itself is not new and unique, the widening divergence deserves attention. Since the CSO and NSO data on consumption presents a completely different picture, it is bound to raise questions about the efficiency of the statistical system. Therefore, it is time to revamp India’s statistical system and enable it to collect, process, and disseminate information with speed and accuracy. The GDP data, for instance, comes with a significant lag and is subject to large revisions, which affects decision making.

At a broader level, the importance of credible and timely data cannot be overstated. Decision making in both the government and the private sector depends on the state of the economy. It is important to note that the GDP data is already under a cloud because it doesn’t match high-frequency indicators.  Withholding data and rejecting official survey results will further damage the credibility of the system. Non-availability or the unwillingness of the government to accept survey findings will hurt its ability to make policy decisions. Further, the private sector would be reluctant to invest if it doesn’t have a clear picture of the economy and suspects official data. The government would be well advised to address the problems in both the economy and the statistical system. Being in denial mode is no solution.