Paper by Peter John and Fredrik M. Sjoberg: “Citizens respond to information about democracy according to whether they are electoral winners or losers. This difference occurs both at the national and constituency level. Democratic interventions that seek to promote accountability and transparency might therefore impact citizens differentially depending on the political party that people support. In a placebo-controlled experimental design, carried out in Kenya, we find that democracy promotion boosts the external efficacy and political participation of ruling party partisans, but leaves those from the opposition unaffected. These responses—based on national incumbency—are further conditioned by the partisanship of the MP of the constituency where the voter resides. These findings throw new light on the impact of civic interventions, such as Get Out the Vote (GOTV) and civic education, common in Africa as well as elsewhere, as we show their benefits accrue to the electoral winners rather than to the losers…(More)”.