The analysis of the origins of order in decentralized societies is perhaps the most famous contribution of African studies to the study of politics. Yet recent scholarship has argued that too much emphasis has been placed upon decentralized systems. On the one hand, their occurrence appears to be relatively infrequent; on the other, even in so far as decentralized societies do exist, they can arguably be regarded as transitory – as societies which once were centralized or which are in the early stages of a movement toward more centralized political forms.

This essay examines various hypotheses concerning the economic basis for political centralization. For among the most frequently posited motives for the formation of states is the desire to achieve economic objectives – ones that presumably could not be achieved under decentralized political systems, such as those described by Evans-Pritchard.

By political centralization is meant the surrender of voluntarism as a basic principle of social action. Under a centralized system, individual members of society are no longer decisive; they cannot veto collective decisions and unanimity is not required. Instead, socially binding actions are taken by a sub-set of society's members, and individuals can be compelled to comply with decisions taken by these agents. As empirical marks of centralized systems, we can note the existence of a bureaucracy, an army, or a central political figure, such as a chief or monarch.