This book provides a new way of looking at the old problem of the assignment of powers in federal structures. A federal state is, by definition, one in which there exists two or more jurisdictional levels between which authority over domains of public policies has to be assigned. In Canada, for example, the provinces have been given exclusive jurisdiction over education; currency and international trade are assigned to the federal government; and both levels have concurrent authority in agriculture. Furthermore, in Canada, as in all federal states, the assignment of powers changes over time in an almost continuous way.The theory developed in this book suggests that the total amount of resources -defined in the broadest possible way -used up in running the public sector varies with the way powers are assigned to different jurisdictional levels; or, to put it differently, varies with the degree of centralization in the public sector. The absorption of resources for the purpose of running the public sector -to be distinguished from resources absorbed in the supply of the public policies themselves -takes four forms; resources used up by citizens to signal their preferences to governments, or to move from one jurisdiction to another; and those used up by governments to administer themselves, and to co-ordinate their activities.Two basic models are examined. In one, the assignment of powers which uses up the smallest amount of resources is analysed. In the other the assignment which is produced by politicians and bureaucrats operating within the framework of representative governments is studied. The two models are applied to the particular problems posed by redistribution and stabilization powers. A new We dedicate this book to the memory of fernand cadieux and tO HENRY ANGUS remembering their friendship and intellectual influence A considerable amount of our research time over the last four years has been devoted to the preparation of this book. The successive drafts -four in all -as well as papers and seminars based on them, which we have delivered during these years, attest to the unmistakable fact that there is but limited resemblance be tween what we are now presenting to the public and what we started out doing.The evolution of our ideas was governed partly by the dynamics of our ap proach to federalism, which was itself conditioned by our desire to produce a uni fied framework to analyse the layered structure of government, but it was also determined in no small measure by the discussions and the written or verbal com ments and criticisms which we have received, together or separately, from num erous friends and colleagues.We cannot hope to name them all. However, beyond a blanket 'thank you' to all who have helped us, we would like to thank in a special way all those who, at one time or another, have provided us with detailed comments, generally in writ ten form, which have had a profound influence on our work. They include Thomas Borcherding,
