Elected officials differ from each other and from the voters in how they perceive the role of representatives. If these role perceptions and the behavior of MPs as representatives are shaped by the institutional constraints of political competition (i.e. electoral rules, see Carey and Shugart, 1995. Crisp et al 2004, Lancaster 1986), then institutional engineering can help in narrowing the gap between principals and agents. The proposed paper primarily investigates the impact of election systems on attitudes towards representation using a uniquely rich data-set. The data-set comprises surveys of politicians and citizens conducted in 2010 in two countries which use mixed electoral systems, Romania and Hungary. By focusing on countries which have multiple electoral channels the analysis can hold constant many potentially confounding variables. Furthermore, the study is able to go beyond the limits of works that investigated mixed systems, cf. Bawn and Thies 2003, Shugart and Wattenberg 2001, Zittel and Gschwend 2008 by exploiting some minor, but theoretically important differences between the two countries: while in Romania all politicians run as SMD candidates, and only the logic of seat allocation divides them into two groups, in Hungary some are SMD candidates, others are list candidates, and a third group is present in both channels. The quasi-experimental design and the detailed surveys allow us to investigate not only how the method of obtaining seats shapes attitudes towards representation, but also to isolate the impact of: the channel of nomination, local politics experience, perceived ideological distance from the party, campaign style, degree of competitiveness of nomination and election, party membership duration and party office, as well as social-demographic characteristics. The contrast with citizen surveys enables us to gauge the degree of discrepancy between elites and masses and to identify the factors which contribute to the difference between the two groups.
