Spatial and Temporal Dynamics of Multidimensional Poverty in Khyber Pakhtunkhwa, Pakistan: A District-Level Analysis (2004 – 2020)

Abstract

This study analyzes spatial and temporal dynamics of multidimensional poverty across 32 districts of Khyber Pakhtunkhwa, Pakistan, using 2004–05 to 2019–20 PSLM surveys. Applying the Alkire-Foster methodology, the Multidimensional Poverty Index (MPI) is constructed using education, health, and living standards dimensions. Results show, that although multidimensional poverty has declined overall in Khyber Pakhtunkhwa, large differences between districts continue to exist over time. Newly merged and remote districts (e.g., Kohistan, Mohmand, and Shangla) exhibit the highest MPI scores. Fixed-effects panel regressions reveal that higher dependency ratios, poor road infrastructure, and greater casual labor occurrence significantly increase multidimensional poverty. Year effects indicate significant poverty reduction until 2014–15, followed by partial reversal in 2019–20. Robustness checks using alternative poverty cut-offs (k=1/4) confirm stable district rankings. The findings demonstrate that multidimensional poverty in KP is structurally and spatially rooted, requiring district-specific, place-based policy interventions, as uniform policy approaches failed in resolving such like deprivations.