What’s falling: Poverty or quality of analysis?

Dodgy Data

Surjit Bhalla, India’s Executive Director (IMF), Arvind Virmani, former Chief Economic Advisor under UPA, and K Bhasin, in an IMF Working Paper, state that to estimate poverty, when no survey has been undertaken, is to take the most recent survey (2011-12) data and update individual consumption (or personal) income by the corresponding growth rate observed in the national accounts (NAS).

However, there are problems with estimating poverty based on the NAS estimate of private final consumption expenditure (PFCE). Practically no one in the world uses NAS estimates of consumption. It overestimates consumption.

Using NAS to estimate poverty

First, Karshenas (2003) has argued that for national accounts, the basic problem is that private consumption is calculated as a residual, and errors in the estimation of other components of the GDP all impinge on the consumption estimate. Nobel economist Deaton also shows that from the estimates of production, imports and exports, the statistical office computes net domestic consumption. The final total is a residual; there is no direct measurement of consumption. As a result, errors in measuring exports, production, inventories, or animal feed are all cumulated into the estimate of consumption.

A second problem with the national accounts estimates of private consumption is that they implicitly include spending by, a) unincorporated businesses, and b) non-profit organisations (such as religious groups, trade unions, political parties, hospitals). To the extent that these items do not affect the consumption of the poor, the use of NAS consumption averages leads to an underestimation of poverty.

Third, these items together with the imputed residential rents, which are included in the national accounts’ private consumption, are not always recorded in household surveys, and can add up to large differences between the two series. This is one main reason for the difference between average income and consumption in the national accounts from the survey means, and the growing wedge between the two series with increasing per capita income.

Deaton has shown that average consumption from household surveys is often lower than average consumption from NAS, something that is as true in India as it is in the US, and this is interpreted (by people like Bhalla) as evidence that household surveys systematically understate consumption.

Fourth, other problems with the NAS are documented by scholars, who note that in many countries, less than half of the national income estimate is derived from primary sources. More importantly, they point to the difficulty of capturing “informal” (including, but not confined to, illegal) income generating activities. This is especially true in India: unregistered firms are 85% of all firms and unregistered informal workers are 91% of the workforce.

Work within the Indian government has revealed serious problems with the accuracy of the NAS measures of consumption. K Sundaram and Suresh Tendulkar (2002), reviewing this work, show that the NAS estimate for consumption of fruits and vegetables in 1993–94 in nominal rupees more than doubled between the 1998 and 1999 versions of the national accounts. The estimate for clothing fell by about a half, and that for rent, fuel, and power rose by more than 40%.

Even with some cancelling out of pluses and minuses, total consumption was revised upwards by 14%. (This revision, if it were carried through to poverty construction using the NAS methodology, would cut the Indian poverty rate by a little less than a half.) So much for the Bhalla et al methodology.

Turning to their findings, they claim that the mechanism for achieving this reduction in poverty is as follows. Real per capita consumption growth grew, they say, at 5.9% per annum (2014-19), higher than the 4.0% growth during the highest GDP growth period (2004-11).

However, how is it that consumption growth is higher from 2014-19 although GDP growth is lower? This question becomes pertinent especially as it was officially established, based on NSS household consumption surveys of 2004-5 and 2011-12, that the highest ever poverty reduction was observed in the history of independent India during this very period? This already should be giving us a hint that NAS consumption is overstating personal consumption in India, as globally.

Secondly, poverty decline over 2014-19, Bhalla et al claim, during a lower GDP growth period (2014-19 vs. 2004-1) happened due to the steep decline in inflation – from a CAGR of 8.4% in the high GDP growth period to just 3.6% during 2014-19.

Do these mechanisms hold? First, they ignore all data on inequality, rising unemployment, falling employment rates, falling wage rates and rising food inflation. Inequality had fallen between 2012 and 2019, primarily because all consumption was getting compressed and trending downwards. However, the recent PRICE, PEW and Oxfam inequality reports (2021) showed that income inequality increased during the pandemic, as the income of 84% households fell.

Second, while the unemployment rate increased from 2.2% to 5.8% during 2012 and 2019, the youth unemployment rate increased from 6% to 17% during the same period (PLFS). Worse still, CMIE data show that the share of the working age population with work (or employment rate) declined from 43% in 2016 to 37% in 2021.

Third, we (Mehrotra and Parida, 2021) have shown that real wages of casual workers as well regular workers, both in rural and urban areas, either stagnated or fell from 2012 to 2019. That is what accounts for the sharp fall in Tendulkar poverty by 140 mn (207 to 268 million) over that same period.

So, even if inflation from 2014-19 was lower than 2004-11, the real point is that real wages, after taking inflation into account, were rising fast between 2004-5 and 2011-12. The falling wage rates since 2012 are consistent with falling GDP growth rates, especially since 2016.

There is concentration of India’s workers between 2017-18 and 2019-20 at nominal wages below Rs 200 a day: 52% of self-employed, 30% of casual wage workers, and 21% of regular workers earn less than Rs 200, which is well below both the rural and urban (Tendulkar) poverty line (projecting forward from the 2011-12 to 2020). An additional 16%, 36% and 19% of each earn Rs 200-300 a day.

Moreover, a food inflation rate of 31% between July 2020 and July 2021 (RBI Bulletin August 2021), would wipe out the impact of the 5 kg of free grain; add to that the fuel inflation, thanks to wilful taxes.

The claim that India’s poverty rate (by $1.9 by PPP) had fallen from 7.4 % in 2014 to 2.5% in 2020 before these grain transfers in 2019, and to 0.8% in 2021, does not stand up to scrutiny.

(The writer is Visiting Professor, Centre for Development Studies, University of Bath, UK)