Defence date: 9 November 2007 Examining board: Prof. Jaap Dronkers, European University Institute, Supervisor ; Prof. Gøsta Esping-Andersen, Pompeu Fabra University, External Co-Supervisor ; Prof. Andrea Ichino, University of Bologna ; Prof. Wout Ultee, Radboud University, Nijmegen First made available online: 16 July 2021 A widespread theme pervades both political theory and the social sciences, in which participation in certain types of democratic institutions could create a more competent, active and public-spirited citizenry. While the school of democracy hypothesis has seen a recent renewal, little empirical research has been carried out in order to evaluate it rigorously. I tried to answer this crucial democratic question by leading an ethnographic study in three cases of municipal participatory budgeting in France, Italy and Spain. They indeed appeared as good training grounds for individuals, as they are empowered institutions that aim at including lay citizens in the discussion and production of local public policies. After almost two years of precise micro-sociological research, I saw people change, sometimes radically. Some, disappointed by their experience, became increasingly cynical about participatory democracy and politics in general. Many others however acquired new civic skills and competences, a wider knowledge of their environment and of the political system, and became increasingly involved in associations, social movements and political parties. Some even became professional and were integrated in municipal electoral teams. Participatory democracy has therefore the potential to empower citizens and create new local elites - be they critical ones, re-boosting civil society, or institutionalised ones, regenerating representative government. Some specific factors appeared crucial in the pattern of self-change I observed. Firstly, the biographical availability and previous political experiences of actors had a decisive impact. Secondly, self-change required mastering the discursive rules of competent behaviour of the institution - what I call their grammatical rules - to be integrated and thus experience intensive participation. Mastering the norms of good behaviour in public - learned through trial and error, and from sanctions and rewards mechanisms, i.e. from the power of the emotions felt in public and especially under the eyes of unknown strangers - people had then the chance to see their personal and political trajectories more radically affected. The individual impact of the involvement in participatory democracy institutions seems therefore to be both the increased civic competence of actors, but also the exclusion of the unskilled and incompetent citizens. The process of self-change can therefore appear as both fostering emancipation and exclusion.