Exploiting VAA-Generated Data to Identify Policy Dimensions: The Case of England

Political scientists often talk about "ideological dimensions" that aggregate related policy issues into a single latent construct. Typical examples of ideological dimensions are (economic) Left versus Right and (socially) liberal versus conservative. In this paper I show how individual issues may be aggregated into three ideological dimensions in England by drawing from opinion data derived from the EUvox Voting Advice Application launched shortly before the May 2014 elections to the European Parliament. I go on to map the positions of party supporters with respect to the dimensions I identify. Unlike many other studies of dimensionality, the dimensions are not defined a priori, but are generated inductively from the opinion data using a combination of Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA) and Mokken Scale Analysis. The results suggest that a cultural dimension that measures the degree of openness to outside influences such as the EU and immigration better distinguishes the main political parties in England than the traditional (economic) Left-Right dimension.