The debate on institutional engineering offers options to manage ethnic and other conflicts. This contribution systematically assesses the logic of these institutional designs and the empirical evidence on their functioning. Generally, institutions can work on ethnic conflict by either accommodating (consociationalists) or denying (integrationists) ethnicity in politics. Looking at individual and combined institutions (e.g. state structure, electoral system, forms of government), the literature review finds that most designs are theoretically ambivalent and that empirical evidence on their effectiveness is mostly inconclusive. The following questions remain open: a) Is politicized ethnicity really a conflict risk? b) What impact does the whole menu (not just single institutions) have? and c) How are effects conditioned by the exact nature of conflict risks?