Is NSS right on consumption fall?

As we head towards the budget session of parliament, we are sailing without the compass economic data has traditionally provided. The once vaunted statistical system lies in disarray with the report of its flagship National Sample Survey (NSS) on consumption expenditure being held back by the National Statistical Office (NSO).

The leaked portions of the survey report suggest that consumption declined in real terms between 2011-12 and 2017-18. Many critics argue that the unfavourable findings have led to the suppression of the report while NSO claims that there are serious data quality issues with the 2017-18 survey.

Whether consumption spending has grown fast, grown slowly, or actually declined --- as the NSS findings suggest --- have huge implications for economic policy-making. And yet, in the absence of data, we are all flying blind.

A study by researchers from the National Council of Applied Economic Research (NCAER) and the University of Maryland provides an independent assessment of changes in living standards for a panel of 4828 households across Rajasthan (2,706 households), Bihar (1,643 households) and Uttarakhand (479 households). The study shows that contrary to the NSS findings, consumption spending has grown between 2011-12 and 2017. Yet, the pace of growth is significantly lower (in real per capita terms) compared to the growth between 2004-05 and 2011-12.

Per capita incomes grew by 3.5% per annum between 2011-12 and 2017 and per capita consumption expenditure grew by 2.7% per annum over the same period. The same households experienced per capita income growth of 7.2% per annum and per capita consumption growth of 4% per annum between 2004-5 and 2011-12.

The data also point to a consistent increase in ownership of vehicles over the two time-periods. While ownership of cars and motorcycles rose 11 percentage points to 22 percent between 2004-05 and 2011-12, the same figure went up 10 percentage points between 2011-12 and 2017.

One of the challenges facing the interpretation of consumption data over this period pertains to difficulties in disentangling long-term changes from the short-term shock caused by demonetization. The demonetization of high-value currency notes, implemented in November 2016, led to a tremendous cash crunch. It adversely affected the purchasing power of consumers and incomes of small businesses and informal workers.

However, the demonetization effect may have been temporary, the study suggests. While temporary cash crunch affected how much money households could spend, families made adjustments by reducing discretionary expenditure without affecting expenditure on food, health and education.

For 2,391 households interviewed between February-April 2017, their real per capita incomes registered only a small rise (2.1% per annum) compared to 2011-12, while consumption declined slightly over the same period.

For 2,347 households interviewed between May-July 2017, their real per capita consumption registered a sharper rise (5% per annum) between 2011-12 and 2017 compared to the period between 2004-05 and 2011-12, when their real per capita consumption grew at 3% per annum.

The spurt in consumption for the latter group may not indicate a boom though. This could well be due to catch up consumption for items families were unable to buy during the cash crunch created by demonetization earlier in the year or it could be purchases taking place to avert paying Goods and Services Tax (GST) that was implemented from July 2017. Per capita income growth between 2011-12 and 2017 for those interviewed in the latter period (May-July) at 5% per annum was lower than what they had experienced between 2004-05 and 2011-12, at 7.5% per annum.

As authors of the study note, these results deserve to be treated with caution. They are based on a relatively small sample of 4828 households across three states interviewed in 2004-5, 2011-12, and 2017 as part of the India Human Development Survey (IHDS). Only 81% of the households initially interviewed in 2004-5 could be contacted for subsequent interviews. While the sample was randomly drawn in 2004-5, more urban households were lost than rural households, making this a selective sample. Moreover, unlike NSS which collects detailed data on over 500 consumption items, the IHDS groups these items into 52 categories providing a somewhat rough estimate of consumption expenditure.

Nonetheless, if the trends reported in the study hold true for the entire country, it would indicate that India’s growth story in recent years may have lost some of its shine but has not collapsed altogether. The boom phase of 7-8% growth in household incomes was over by 2011-12 and has moderated since then. Consumption has also witnessed a similar slowdown but it has not declined in absolute terms, the study suggests.

If these figures paint a picture of modest growth, they also highlight the need to ensure credible survey data to guide public policies. The NSS consumption survey was initiated in July 2017 just around the time that the IHDS survey was concluding. Hence, if the results from the IHDS hold true, NSS should have documented a weak increase in real consumption rather than an absolute decline. This suggests that NSO claims about data quality issues must be taken seriously.

It is time to reinvigorate India’s data systems to restore trust in official statistics, and enable evidence-based policymaking.

Sonalde Desai is a professor of sociology at the University of Maryland and professor and centre director, NCAER-National Data Innovation Centre.