Kasper Vrolijk examines and offers causal evidence on the link between trade exposure and social cohesion using rich micro tax data and a natural experiment of exchange rate liberalisation in Uganda. Results show that exposure to exogenous exchange rate shocks has significant albeit economically small effects on social cohesion: it reduces trust, enhances participation, and has ambiguous effects on identity. These effects operate largely through the expenditure channel (or household exposure) and to a lesser extent through the earnings channel (captured by worker and firm exposure).