Operationalizing Multidimensional Representation

The study of representation has been a major research field in quantitative empirical political science but conceptual work in political theory is limited to a few landmark studies. While political theorists have developed sophisticated and multidimensional models of representation, empiricists have almost exclusively relied on a simplistic and unidimensional modelling of representative processes. We contend that although this ‘theory-empirics gap’ can never be closed entirely, it is much wider than it would have to be, due to several incompatibilities between the extant theoretical and empirical research on representation. We discuss these incompatibilities and argue that a ‘synthesis model’ of representation that is both faithful to political theorists’ conceptual impulses and operationalizable for empiricists must achieve four things: 1) explicitly account for the existence of parties as well as SMD versus PR electoral systems; 2) overcome the focus on unobservable cognitive processes within the representative; 3) abstain from the construction of Weberian ideal type-representatives; and 4) limit itself to institutionalized political representation. We develop the central tenets of such a model by discussing four theoretical dimensions of representation that have empirical correlates relating to representatives as well as voters: (a) surrogation (claiming and choosing constituents), (b) justification (providing reasons for actions), (c) personalization (relying on judgement), and (d) responsiveness (reacting to constituents). Interestingly, three of these four dimensions (i.e. a-c) have received very limited attention in empirical work. We sketch an empirical research programme on multidimensional representation focusing on these ‘understudied’ dimensions.