What are the linguistic effects of political institutions? Are consensus-building, power-sharing democracies more likely to recognize minority languages? In this article, I argue (1) power-sharing institutions—proportional electoral rules, parliamentary systems, and federalism—are less likely to recognize minority languages than their moderation-inducing, power-concentrating counterparts; but (2) if there is recognition, the level of recognition is actually greater in the former than in the latter. By testing this argument using a newly constructed language-in-education barometer, I find a significant and robust relationship between political institutions and minority language recognition.