There is a thriving industry within Indian academia churning out poverty estimates. But there is still no unanimity on the sharp decline as claimed by the Planning Commission. Latest numbers indicate that the poverty level averaged 27.5 per cent in 2004-05 and dropped to 25.9 per cent in 2005-06. That means there are 260-odd million poor people in the country whose income is less than Rs 356.30 a month in the villages and Rs 538.60 a month in the cities. These numbers themselves are a damning indictment of six decades of development. What sense does it then make to redefine poverty on the basis of access to education, health, infrastructure, clean environment and benefits for women and children, in addition to the existing poverty line?
To be sure, the share of the population deprived of access to basic facilities is bound to be much larger than 25.9 per cent — only indicating a greater failure on the part of the State that is ostensibly committed to inclusive development.
The big question is also whether these poverty magnitudes can be reliably measured given the weaknesses of our statistical system. The infirmities of the wholesale price index that policymakers depend on in underestimating inflation are well-known. The basic data for estimating poverty comes from large-scale surveys of consumption expenditure conducted by the National Sample Survey
Organisation (NSSO) every five years, and the National Accounts Statistics (NAS) whose estimates of consumption are on the high side. Scaling up the NSSO’s data on the size distribution of household consumption to the NAS data pulls up a large number above the poverty line.
This was one reason for the scepticism surrounding official estimates of poverty reduction.
Redefining poverty also requires a beefing up of NSSO’s capabilities, given a severe shortage of investigators and high–levels of attrition at the field-staff level. Otherwise, it would only provide further grist to the poverty industry.