International audienceThe aim of this paper is to propose a picture of French political supply in 2007 as perceived by voters, with no a priori ideas regarding the relevant political dimensions. Our study is based on experimental data collected during the French presidential elections on April 22nd, 2007. Two alternative voting methods were tested: the approval voting rule and the evaluation voting (2,1,0) rule. We provide an extensive analysis of the political supply in terms of simple political characteristics, derived from an MCA on the ballot data: groups of candidates are associated with different political positions with no a priori conjectures about political programs, societal trends, or candidates' valence