Poverty Mapping at District Level in Pakistan: An Application of ELL Method of Small Area Estimation

Abstract

This study utilizes the Elbers, Lanjouw, and Lanjouw (ELL) method of small area estimation to generate poverty data at the district level in Pakistan, addressing a critical gap in localized information. By employing the unit-level ELL model, we estimated poverty headcount ratios based on the proportion of poor households. The analysis integrates data from the Household Integrated Economic Survey (HIES) 2018-19 and the Pakistan Social and Living Standards Measurement Survey (PSLM) 2019-20. A comparison of Mean Squared Errors (MSE) between direct and model-based estimates reveals that the ELL model-based estimates are more reliable than direct estimates, particularly for districts with limited sample sizes. Our findings unveil significant spatial disparities in poverty across districts, with twelve districts exhibiting extreme levels of poverty and sixteen districts facing severe poverty conditions. This granular assessment of poverty status at the district level provides policymakers with invaluable insights for targeted interventions to alleviate poverty in these regions.