Estimation of equivalence scale and assessment of its impact on poverty measurement in Bangladesh

Abstract

Anti-poverty policies for sustainable development require efficient targeting, for which appropriate poverty lines play a crucial role. In Bangladesh, official poverty lines are estimated with the implicit assumption that there are no economies of scale in household consumption with respect to household size or composition, which raises the question of the accuracy and reliability of the measurement of poverty line. We test the existence of economies of scale, estimate their size, and assess the impact of applying equivalence scale to poverty measurement, using the 2010 Household Income and Expenditure Survey data of Bangladesh. The results confirm the existence of economies of scale in household consumption. Following the model developed by Kakwani and Son, the overall index of economies of scale in household consumption is estimated around 0.85. Modified poverty lines show that under official poverty lines, the probability of being poor is high with respect to household size. The result implies that the poverty head-count ratio(HCR) for households with large number of members might be overestimated in Bangladesh, and that there may be an incentive for low income families to enlarge family size to avail of anti-poverty public transfers.