The European Union Commission has proposed using consolidated base taxation and formulary apportionment to tax the EU-source income of multinational companies. This paper examines US state experience with a similar approach. Despite some positive lessons, especially the need to consolidate income of affiliated companies, lessons are mostly negative, especially regarding the choice of apportionment formula, the use of economic criteria to define the group whose income is to be consolidated, and complexity caused by lack of uniformity. US experience says nothing about using value added to apportion income-an approach that is conceptually attractive, but subject to transfer pricing problems.

This paper discusses how the theory and the practice of tax assignment-which level of government should tax what, and how-depend on history. It describes the meaning and methods of tax assignment, reviews implications of Musgrave's three-branch view of public finance, notes the importance of accretions to knowledge-of the technology of taxation and of the economic effects of taxation, speculates about how economic evolution affects the conventional wisdom on tax assignment, identifies questionable tax assignments found in various federations that are legacies of history, and emphasizes the danger of assuming "one size fits all" in tax assignment. 1 This wording of the question is adapted from the title of Musgrave (1983). The other pieces of the puzzle of intergovernmental fiscal relations are a) the assignment of expenditure functions; b) the design of intergovernmental grants, which may be needed to compensate for spillovers of costs and benefits between jurisdictions, to offset horizontal fiscal disparities, and to overcome vertical fiscal imbalance; c) the borrowing powers of subnational governments; and d) the institutional framework within which all these issues are addressed. These topics are beyond the scope of this paper. 2 I draw on my earlier work on tax assignment, much of which is not cited; see McLure (1998) and (2000b) and works cited there.

A s an advocate of decentralization, I find it difficult to respond to Remy Prud'homme's arguments against decentralization. It is undoubtedly true j L that decentralization, done badly, can cause problems. But the article is not devoted primarily to demonstrating that proposition; for the most part it argues that decentralization, per se, is wrongheaded. Prud'homme (p. 201) sets up a straw man that he calls the "pure decentralization of fiscal federalism." He then argues that relying on this theory would lead to a variety of problems, including a tolerance of interjurisdictional disparities, an inability to conduct macroeconomic stabilization policy, and a loss of productive efficiency in the public sector. In a section entitled "Beyond the Centralization-Decentralization Dichotomy," he presents his antidote to this approach. The "centralization-decentralization dichotomy" is a false one. Prud'homme cites no literature advocating the "pure" theory of decentralization, and the antidote that he prescribes is little more than the conventional wisdom in this area. (The reference list contains only one item that pre-dates 1985; thus it omits many classics in the literature, which present a more balanced theory.) First, Prud'homme states that according to the "pure" theory of decentralization, interjurisdictional disparities are "abnormal phenomena. .. that will be reduced and eliminated automatically by the movement of goods, capital, and labor" (p. 202). The implication that there are no transfers in the decentralized theory can be true only if one ignores literature that does not conform to the author's Procrustean bed of "pure" theory. Academic literature on fiscal disparities and intergovernmental grants goes back at least to Buchanan (1950), which evoked the response by Musgrave (1961). Oates (1972) devotes one of six chapters to "The Theory and Use of Intergovernmental Grants," and Break (1980) contains a fifty-page chapter on "The Economics of Intergovernmental Grants." The study of such grants has not been confined to academic research. Although the use of grants is relatively limited in the United States, that is not the case in other federations. In Australia the very title of the official history of the

No abstract