Dynamics of multidimensional child poverty and its triggers: Evidence from Ethiopia using Multilevel Mixed Effect Model

Empirical studies on the dynamics of multidimensional child poverty are very limited and still more research is required to understand its nature and triggering factors especially in the context of developing countries. In light of this, this paper tries to assess the dynamics of multidimensional child poverty and major factors associated with it using longitudinal data mainly collected to assess child poverty in Ethiopia. It uses multilevel mixed effect logit models that could possibly incorporates fixed and random effects to capture the effect of cluster level and time varying variables on multidimensional child poverty transition. Results of the multidimensional poverty analysis indicate that, although there were significant variations among regions, multidimensional child poverty has decreased during 2002-2009. The paper argues that multidimensional child poverty has dynamic nature that would possibly resulted from the interaction of multiple factors including household demographic, household capital (human, social and resources), household economic activities, geographic locations, and household shocks. Moreover, the study shows the relevance of considering cluster level differences during poverty analysis to generate information relevant for designing targeted policies and strategies that would help to distribute available development resources efficiently and achieve sustainable poverty reduction in developing countries.