A central argument for trade liberalization is that when the `gains from trade' are shared, countries see large gains in economic development. In this paper, I empirically evaluate this argument and assess the impact of elite capture on regional development. Africa provides a unique study ground because the arbitrary placement of country borders during the colonial period partitioned hundreds of ethnic groups across borders. This partitioning is a source of variation in population heterogeneity and cross-country connectedness that is independent of economic considerations. Thus, African borders provide both a credible instrument for bilateral trade flows and enable the assignment of trade flows ---and their impacts--- to individuals. I find that while ethnic networks increase trade flows, increased trade activity decreases subnational economic development when measured by satellite data or individual wealth. I show that this counter-intuitive result comes from elite groups capturing the gains from trade, with detrimental impacts on trust and democratic progress in society.