P α Poverty Measure: An Estimable Approach

The focus of this paper is on deriving a formula for estimating the P_α measure of poverty when the relevant consumer expenditure/income distribution data are available in grouped form. The formula is derived assuming consumption/income distribution to follow the three parameter lognormal distribution as a function of the moments of the income distribution. The moments are estimated using the method of Maximum Likelihood and hence the estimated poverty measure is consistent and efficient under certain very general conditions. The proposed formula is Illustrated with reference to the Indian National Sample Survey (NSS) data on consumption distribution for the period 1961-62 to 1990-91. The results show that the poor benefited (suffered) more during periods of general improvement (decline) in economic conditions and levels of living.