TOWARDS A POSITIVE THEORY OF FEDERALISM By Albert Breton It was in 1964, in Charlottetown, at the 36th Annual Meeting of the Canadian Political Science Association, that as a member of a panel commemorating the hundredth anniversary of the first meetings that led to the creation of Canada, I read a paper titled: "A Theory of Government Grants". After the panel discussions, I was approached by Tony Scott, who asked me if I could be interested in applying, with him, to the Canada Council for a grant to do research on "federalism". I did not know Tony Scott, but I knew of him and of his work. I immediately said "yes". We applied to the Council and we were awarded a Killam Grant. Upon reception of the grant, we began meeting and met over several years on a regular basis in Vancouver and Toronto. When in Vancouver, I stayed with Barbara and Tony; when in Toronto, Tony stayed with Margot and I. In the early 1950's, Tony had published at least three papers on "federal grants" and was, when we began working together, well aware of the many problems they appeared to pose in practice and also that they were exercises in standard "Public Finance Federalism" – a body of doctrine that is primarily concerned with public sector revenues and expenditures. In 1960, Pierre Trudeau was appointed professor at the University of Montreal where I also taught. During the academic year, between 1960 and 1965, we lunched together once a week. We covered a lot of ground; eventually, we began discussing the Federalist Papers which Pierre, it seemed to me, had memorized. This is when I began learning about federalism. The Federalist Papers is not cast in economic terms, but it is not that difficult for an economist to understand. However, it is only when Tony Scott and I began working together that I came to fully appreciate the significance of the Papers. We began by returning to the general concept of the 'span' of private and public goods, which, without calling it that way was implicit in my Charlottetown paper. We surveyed much of the literature on federalism, but it was clear that something basic was missing. Indirectly inspired by the Federalist, we became aware that governmental systems are structures – organizations. We had both read Ronald Coase on "The Nature of the Firm" and understood that basic to the role and operations of firms as organizations were "transaction costs". We decided that it would be wise to leave that expression to Coase's world of firms and markets; we opted for "organization costs" in dealing with governments and governmental systems. Identifying the relevant organizational activities and their costs, and specifying how these are related to the structure of governmental systems was, on the whole, fairly easy as these had been examined by a number of scholars. For example, on the demand side, the signalling of their preferences by citizens had been studied and their movement from one jurisdiction to another had received much attention. On the supply side, the coordination of governmental activities as well as the central role of administration had also received much attention. We then formulated assumptions to relate organizational costs to the degree of centralization of the governmental system. We would come to a shared view on these issues and immediately write something down. Except for Chapter 10 on "The assignment of the redistribution functions", which was written by Tony alone, all other chapters in the book are true joint products. A most difficult problem was to identify a mechanism that would generate an equilibrium degree of decentralization. We invented an institution – we called it "a constituent assembly" – whose function was to generate equilibrium assignments of powers that would minimize organizational costs. We were immediately criticized for that creation, as it was claimed, rightly we acknowledged, that there are not many of these assemblies around. Two years after the publication of The Economic Constitution of Federal States, (in 1978), we published a small book on constituent assemblies: we were both aware that the problem was still unresolved. The problem was redefined and successfully addressed by Pierre Salmon (in 1984), who argued that the mechanism was embedded in "intergovernmental competition". Tony was, I believe, never sure that Salmon's solution was "the" solution. However, that did not prevent him from using the basic model – short of an equilibrating mechanism – to shed light on real world problems of public sector decentralization. In 1984, he examined tax harmonization in federal and other states; in 1991 he analysed "the market for characteristics of property rights"; and in 2000, he analysed the assignment of powers over the environment".