City size distribution of African countries: A spatial analysis using geospatial data

Abstract

This paper examines city size distributions across 43 African countries from 2013 to 2021, addressing the applicability of Zipf's law, the evolution of urban hierarchies, and spatial interdependencies. Using a novel dataset combining population figures with nighttime light intensity, we test for adherence to Zipf's law, analyze urban dynamics through rank-size distributions and Markov chain models, and employ spatial autoregressive estimation to capture cross-border effects. We find that while population distributions generally follow Zipf's law, economic activity (proxied by nighttime lights) shows deviations, indicating inequality in resource allocation. Regional variations are pronounced, with Southern Africa exhibiting the closest adherence to Zipf's law for population. Urban hierarchy analysis reveals stability at distribution extremes, with medium-sized cities most likely to transition downward. Our spatial analysis uncovers complex interactions between economic, socio-cultural, and political factors in shaping urban structures, highlighting significant indirect effects and regional interdependencies. These findings have important implications for African urban policy, supporting arguments for developing secondary cities to promote balanced economic growth and reduce urban poverty. Our results contribute to understanding complex urbanization patterns in developing economies and inform context-specific, regionally aware urban development strategies.