This paper aims at assessing the role of national institutions in observed development differences between countries with heterogeneous geographical and anthropological backgrounds.Comparative economics most often rely on national averages as collected by international databases.We explore another promising path where the location of national borders is viewed as a historical“natural” experiment. Drawing from large sample household surveys undertaken around 1990, wecompare the development outcomes of neighbouring localities on both sides of the Côte d’Ivoire borders with Burkina-Faso, Ghana, Guinea and Mali. We examine a large span of variables including fertility, polygamy, child nutrition, education, agricultural productivity, utilities connexion, andmonetary welfare. In order to control as much as possible for geographical and anthropological factorsunderlying national differences, we compute simple mean differences between localities close to theborder, and also implement a matching nearest neighbour and a simple linear regression-discontinuity estimator. This identification strategy of the impact of national idiosyncrasies seems well suited tomost African borders which have been drawn rather arbitrarily during the colonial period, between two colonial powers or inside the same colonial empire. Results first show that borders matter. They reveal most striking national idiosyncrasies, some which date from the colonial era and some others which appeared later. In the end-1980s and mid-1990s, there was obviously a positive externality of living in Côte d’Ivoire. Results also show that national averages differences and border treatment effects do not necessarily coincide; in some cases, geography and anthropology also matter.