A nice post thought of posting..
Firstly, There is already a nice Reuters' spin on the document posted regarding absolute poverty which has been proved wrong. Secondly, When they say people living on X amount, they mean consumption per capita, not income. Third, all data on income, wages, consumption, etc. is based on NSSO's 1999-2000 round, i.e. the reported data are 8 years old, given in FY1999-2000 prices, and have all of the limitations of a CSO-run data collection process (viz. competence, discounting black money, etc.). I would posit that while 77% of the population is poorly off by Industrialized standards ($2/day reported to the government by my hisab is roughly equivalent at PPP to the income poverty lines in the EU-15), they are still 4x as well off as Nigerians (where 77%+ of the population lives in below 1985PPP$1/day income poverty), i.e. the poverty of this group is relative not absolute.