Dynamics and Determinants of Household Poverty in Ethiopia: A Correlated Random Effect Approach

Abstract

Despite the implementation of the numerous programs, strategies and policies in developing countries, poverty has remained a persistent challenge. This study investigates the dynamics and determinants of household poverty in Ethiopia using two waves of panel data from the Ethiopian Socioeconomic Survey (2019 and 2022) and a correlated random effect logistic regression model. The results reveal that the incidence, gap, and severity of poverty have increased over time in Ethiopia. Rural and female-headed households are disproportionately affected, reflecting persistent spatial and gender-based disparities. Transition analysis further indicates that 31.64% of households experienced chronic poverty, 43.92% of households were in transitory poverty and 24.44% households remained sustainably non-poor. The regression results  indicate that family size (β = 0.318, p = 0.000),  age of household  heads (β = 0.006, p = 0.010), participation in productive safety net program (β = 0.361, p = 0.000), being from Amhara region (β = 0.982, p = 0.000), being from rural areas (1.115, p=0.000), year dummy and being from SNNP region (β = 0.732 p = 0.000) are positively and significantly related with poverty. But, male-headed households (β = −0.317, p = 0.000), and education level of household heads (β = −0.615, p = 0.000) are negatively and significantly related with poverty in Ethiopia. This study is limited by its focus on a unidimensional welfare indicator, and by the exclusion of the Tigray region from the analysis. These findings highlight the need for multidimensional, region-specific, and gender-sensitive interventions to curb poverty, including expanding rural infrastructure, improving agricultural productivity, enhancing educational opportunities, and redesigning PSNP to focus not only on short-term consumption smoothing but also on long-term investment in agricultural and non-agricultural activities. 

Keywords: Correlated random effect Model, Endogeneity, Foster-Greer-Thorbecke, Ethiopia, Poverty