Davis and Moore's theory of stratification (1945), though frequently discussed, has stimulated remarkably few studies. Perhaps this is due to the lack of derivations of empirical propositions in the original article. I would like in this note to outline some empirical implications of the theory.

Davis and Moore's basic argument is that unequal rewards tend to accrue to positions of great importance to society, provided that the talents needed for such positions are scarce. ‘Society’ (i.e. people strongly identified with the collective fate) insures that these functions are properly performed by rewarding the talented people for undertaking these tasks. This imples that the greater the importance of positions, the less likely they are to be filled by ascriptive recruitment.

It is quite difficult to rank tasks or rôles according to their relative importance. But certain tasks are unquestionably more important at one time than at another, or more important in one group than another. For instance, generals are more important in wartime than in peacetime. Changes in importance, or different importance in different groups, have clear consequences according to the theory. If the importance of a role increases, its rewards should become relatively greater and recruitment should be more open.

The following empirical consequences of the theory are ‘derivations’ in a restricted sense. We identify supposed changes in the importance of roles, or identify groups in which certain roles are more important.