An Alternate Approach for Measuring Poverty in Pakistan

Abstract

Poverty is a universal phenomenon and has been issue of concern throughout history. There are two main approaches that are used to measure poverty; unidimensional approach and the multidimensional approach. Although, poverty literature is profiled with multidimensional approach researches however, unidimensional approach is still being used officially in many countries. The most notable limitation of this approach is the dichotomization of population into poor and non-poor defined with reference to chosen poverty line. To some extent, it is unfair to consider someone non-poor even though individual is slightly above the poverty line. This study employs fuzzy regression technique that treats poverty as a matter of degree rather than an attribute that is simply present or absent in individuals. Fuzzy regression computes membership function that assigns values between zero and one to individuals around the given poverty line. Theoretical modelling of our modelling is based on Engel’s law, which states that, there is negative relationship between food share in budget and the real income. On balance, 30.34% people in Pakistan are living in poverty with significant contribution of Sindh and Baluchistan. Highest incidence of poverty is recorded in Rural Sindh where 63.95% population is living in poverty and lowest is in urban areas of Punjab and KPK. Among poor, 57% are illiterate which leads to 13.23% unemployment rate.