the poor, then the appropriate performance index would be not just th- head-count ratio but the Sen-index.
VII Conclusions It has been shown that the revised population estimates which incorporate the results of the 1981 census corrected for under-enumeration, would lead to an estimated 334 million poor in 1981-82 under the assumption of distributional neutrality of growth where growth in per capita expenditure (PCE) is inclusive of the additional consumption generated by IRDP. A netting out of additional consumption generated by IRDP pushes up this figure to 340 million (Table 5, line 4)'. If we subs tract from this, 11 million people who could plausibly have been pushed above the poverty line by IRDP the estimated incidence of poverty in 1981-82 would be 46.50 per cent or 329 million poor instead of 41.50 per cent or 282 million poor. It must be stressed that even this estimate of the poor assumes at the very minimum that the groups which are deemed to have been pushed up by growth did experi ence the average growth rate in their PCEs. If even this assumption does not hold, .is is likely in the urban case, the number of urban poor and hence the total ' (rural plus urban) poverty population would be higher.