Class locations, consumption locations, and the geography of voting in England

Much ecological analysis of voting patterns in England has been concerned to show (a) the importance of occupational class as a determinant of electoral behavior, and (b) that deviations from a predicted pattern based on class variables are consistent with the neighborhood effect hypothesis. Recently, Dunleavy has criticized both these orientations, replacing them with the concept of consumption locations. The present paper shows, using a new procedure to estimate voting at the constituency level, that there still remain spatial variations to be accounted for.