World Bank suggests 100mn poor vs official 450mn
Does India have 350 million poor, 450 million, or under 100 million? While the Tendulkar formula put the number of poor at 354 million in FY10, Rangarajan put it at 454 million—in an interesting coincidence, for FY05, Tendulkar’s estimate was also 100 million more than the previous, Ladkawala, estimate—and estimates based on the World Bank’s new purchasing power parity (PPP) index could put the number at even below 100 million, unless the World Bank decides to dramatically hike its poverty line. It is important to keep in mind, though, this is just a number. It is no one’s case that poverty is uni-dimensional, measured only through head-count ratios of percentage of population below a poverty line. Accordingly, HDI (human development index) and multi-dimensional poverty measures have evolved. India’s poverty numbers are based on the National Sample Survey (NSS) which, as the name suggests, is a survey, not a census. The more sensible way to identify poor households is through, as the Planning Commission suggested since the 9th Plan, ownership of assets – the socio-economic surveys used for rural households are based on a similar idea. The Tendulkar vs Rangarajan vs someone else debate is really a red herring—logically, if the threshold level of consumption is raised, the poverty numbers will also rise, which is how the Tendulkar numbers were higher than Lakdawala and Rangarajan higher than Tendulkar.
Regardless of poverty line, there is no denying that over time, poverty ratios have declined. That does not negate the proposition that declines should have been more (a function primarily of the rural sector). Nor can there be an argument that the poverty line represents a minimal subsistence level of consumption and no more—as economies develop, poverty lines do increase. To muddy waters further, there is an international poverty line of $1.25 per person per day—there is an additional one of $2 per person per day. The original 1993 $1.08 PPP poverty line was roughly comparable to the Indian poverty line then. But it is no longer possible to be that categorical. That $1.08 was increased to $1.25 in 2005 (because prices in developing countries had been under-estimated) and is now being reportedly shifted from a 2005 to 2011 base. In the process, the number of poor people in India will decline from 400 million to 100 million, good for achieving poverty reduction targets of Millennium Development Goals (MDGs).
But beyond that, this number should mean nothing at all. Since income and expenditure distributions are log normal, sharp declines in poverty are indeed possible, especially when thick part of a distribution passes above the poverty line. PPP measures also have a role in GDP estimations. If apple-to-apple comparisons are made, no one can contest the temporal decline in poverty. But beyond that, this number has no implication for policy, apart from the obvious of growth being good for poverty reduction. Unfortunately, this is yet another number that will be debated, though the sampling for all PPP surveys remains somewhat suspect.