The first task of the study consists in the establishment of the attribute which is to be explained, i.e., the degree of democracy in the countries under discussion. In purely practical terms this is a matter of compiling a wealth of data on the situations in these countries. But before this can be achieved, we must, of course, decide on which information we shall seek, and how this in turn is to be weighed and interpreted. We need a number of empirical indicators which to a reasonable extent reflect the degree of democracy in the different states. The question is, what should these measures be?

The simplest and, in view of the cumulative nature of science, most fruitful approach is to relate to a firm, well-founded tradition within the field of research – that is, if such a tradition exists. As was mentioned earlier, however, this is hardly the case. When we survey, the fairly extensive research hitherto pursued we are immediately struck by the variation which prevails regarding the indicators of democracy which have come into use. In the face of this motley assortment we can only state that whoever undertakes this task must make his or her own choice of indicators and give reasons for his or her stance. Such, in brief, is our starting-point.

How are we to proceed? It is clear that if the choice of empirical indicators is to be convincing it must relate to, and reflect the fundamental criterion of the theory of democracy, namely the general principles which characterize democratic government.