ABSTRACT
The review of previous analyses of electoral behaviour in England (Chapter 1) indicated that the conventional wisdom prefers the ‘class-equals-party’ model and that, despite evidence to the contrary from aggregate data analyses, geographical variations in this equation are ignored. Those who reject the ‘class-equals-party’ model replace it with others in which geography is also considered unimportant. The analyses in Chapter 2 suggest that these alternative models are also insufficient representations of the behaviour of the English electorate. A ‘one nation’ perspective hides substantial geo graphical variations. The ‘class-equals-party’ model is a relatively poor predictor of the partisan distribution of voters in individual constituencies, as is the national flow-of-the-vote matrix.
