A METHOD FOR CROSS-NATIONAL COMPARISON OF POLITICAL CONSENSUS.

A number of writers on political behavior have recently focused their attention on the degree of consensus necessary to the maintenance of a stable democratic system. As an indicator of consensus, professor Seymour Martin Lipset has suggested the extent to which all sections of the population react in the same direction to major stimuli. However, there are grounds for arguing that the notion of consensus is not quite so straightforward, or unidimensional, as this approach would suggest. The purpose of the article to clarify two different implications underlying the notion of consensus and to suggest a way of presenting cross-national comparisons based on voting statistics so as to bring out both of these implications. Data from Great Britain, Canada and Australia are used in order to illustrate the method. The graphical method which seems the clearest is put forward. The method is to plot a graph showing the proportion of manual and non-manual voters voted left or right in successive elections in different countries. Each election is represented by a single point on the graph, so that it becomes possible both to compare one country with another and to trace a trend in the voting behavior of a single country. The result of any given election cannot, of course, be deduced from the graph, but it is possible to trace both uniformity of swing to the left or right and also any trend towards or away from a situation of minimal class conflict.