Recently, we had expressed concerns that with the GDP growth rate falling in the post 2011-12 period, the decline in the poverty ratio would be slow. During 2011-12 to 2018-19, both GDP and agriculture growth were lower than in the earlier period. The terms of trade were not in favour of agriculture and rural wage growth has declined. The economy has grown by 5 per cent and 4.5 per cent in the first and second quarters of FY20 respectively, and growth is expected to be around 5 per cent for this fiscal year. Thus, we have reasons to believe that the rate of decline in the poverty ratio will be low.
The leaked NSS consumer expenditure data for the year 2017-18, however, shows that the real monthly per capita expenditure (MPCE) has actually fallen between 2011-12 and 2017-18. It shows that real MPCE declined from Rs 1,501 in 2011-12 to Rs 1,446 in 2017-18 — a fall of 3.7 per cent at the all-India level. In rural India, it declined by 8.8 per cent, while in urban India it increased by 2 per cent over the same period. These findings do raise some questions. Most importantly, can MPCE fall when the income growth rate has been positive, both in aggregate and in per capita terms? There is no corroborative evidence to support the decline of MPCE. Even the report of Oxford University on the Global multidimensional poverty index shows that the incidence of multidimensional poverty had almost halved between 2005-06 and 2015-16, climbing down to 27.5 per cent. We have to note that the terminal year of this study was 2015-16.
Is there any period in the past where MPCE has actually fallen? Typically, in recent years, the consumption expenditure surveys have been conducted after a five-year gap. Table 1 gives the trends in MPCE since 1993-94. In every survey, it has risen over the previous survey. The only exception is 2017-18. Even in 2009-10, which was a drought year, MPCE went up.
In an article, S Subramanian has carried out a full analysis of the 2018 survey data. The survey shows that while real mean consumption declined for all the deciles between 2011-12 and 2017-18, for the bottom decile, it declined by 1.3 per cent, while for the top decile, it declined by 17.4 per cent. It is interesting to see that the decline in mean consumption was much higher for the top decile. This study estimates poverty ratios for 2011-12 and 2017-18 using the Rangarajan Committee’s poverty line. The estimates show that the poverty head count ratio increased from 31.2 per cent in 2011-12 to 35.1 per cent in 2017-18. According to this study, inequality declined from 28.7 to 25.8 in the same period.
What is disturbing in the survey data for 2017-18 is the alarming difference between the aggregate private consumption expenditure given by it and the figure provided by the national accounts statistics (NAS). These two estimates of consumption (NSS and NAS) do not match in any country; India is no exception. What is perplexing is that the difference in India between the NSS and the NAS consumption has been widening over time. From a difference of less than 10 per cent in the late 1970s, it has widened to 68 per cent in 2017-18, that is, the survey estimate is only 32.3 per cent of NAS estimates (Table 2).
The expert group (led by C Rangarajan) had carried out an analysis of the possible reasons for the difference between the estimates. It made some headway, but could not fully explain the difference. Therefore, it continued with the practice, initiated by the expert group, led by D T Lakdawala, and continued by the expert group led by S Tendulkar, of estimating poverty in India solely by reference to the size-distribution of private consumer expenditure based on the NSSO methodology. However, with the difference rising to 68 per cent, the time has come for a deeper analysis of the factors contributing to the difference.
According to NAS, private consumer expenditure increased from Rs 49,10,447 crore in 2011-12 to Rs 74,17,489 crore in 2017-18. This is an increase of 51 per cent. The per capita consumer expenditure increased by 41 per cent during this period. But, according to the survey, per capita consumption expenditure decreased by 3.7 per cent. The difference is too big to be pushed under the carpet. The NSSO Advisory Group or the National Statistical Commission must study the problem and come out with possible suggestions for improving the collection of data through both routes.
It may be recalled that government had difficulties in accepting the 2009-10 consumer expenditure data put out by the NSSO because it was a drought year. The then government recommended carrying out another survey, and it was undertaken in 2011-12. We think that the present government should also release the report and the unit level data of the 2017-18 survey, which can be analysed by researchers. The National Statistical Office can also give a note on the limitations of data in the report. The government can order another survey after analysing the reasons for difference.
This article first appeared in the print edition on December 13, 2019 under the title ‘Mind the statistics gap’. Rangarajan is former chairman of the Economic Advisory Council to the Prime Minister and former Governor, RBI. Dev is director and vice chancellor, Indira Gandhi Institute of Development Research.
📣 The Indian Express is now on Telegram. Click here to join our channel (@indianexpress) and stay updated with the latest headlines