We can consider the tradeoff devices reviewed in this article as dynamic devices or heuristics for allocation of jurisdiction: as components of constitutions. First, to which level of governance should responsibility be assigned: to which level of governance is this power more valuable? Second, is there a way in which this power can be fragmented, so that the portion that is more valuable at the local level is enjoyed there, while the portion that is more valuable at the central level is assigned to the central level? These questions are central to the economics of federalism, which addresses the utility of congruence between effects and governance, and would seek to establish governmental units based on such congruence, subject to the costs of fragmentation of authority. It is essential to recognize that each tradeoff device serves as a heuristic to allocate authority in particularistic, fact-specific ways over time, and thus may provide a more complex solution to the level of authority problem than a simple static and/or broad allocation, such as is expected to be found in constitutions.
In fact, there are two main methods for vertical allocation of power in a federation or quasi-federation. The first method is through specific rules listing areas of competence of the center and the constituents, respectively. The second method is to specify a standard, such as "subsidiarity," or such as NP, BT, CBA and CCBA. We have seen that, at least in the U.S. and in Europe, attempts to allocate powers vertically by reference to rules specifying discrete lists of powers have been eviscerated or rendered indeterminate by the complexity of public policy. The alternative route to vertical allocation of powers discussed in this article, by virtue of the use of standards, or those tradeoff devices that go beyond the simplest NTR and SMERT, embraces complexity and seeks to maximize social values in particular circumstances. These principles do not make it easier to allocate powers, or less political, but simply recognize the particularity of choices in this context, and embrace the difficulty of this task. The choice between lists, on the one hand, and on the other hand standards like "subsidiarity" and the tradeoff devices studied here, is a choice between rules and standards. It is also, importantly, a choice between courts and legislatures, with important moral and theoretical reasons for choosing one over the other.
The doctrine of subsidiarity (subsidium meaning to help) holds that action should not be taken at a higher level unless it helps: unless individuals get more of what they want. In this contractarian sense, the tradeoff devices catalogued herein are simply judicially-created (and by now, legislatively-accepted) tools of subsidiarity. "[F]or an economist, the principle of subsidiarity means that the production of public goods should be attributed to the level of government that has jurisdiction over the area in which that good is 'public.'" Of course, the definition of "public goods" for a particular society depends on its goals, and there are always spillovers of public goods. The question then is where to draw the line. The task in economic terms is to develop a heuristic that will achieve an optimal level of congruence between jurisdictional boundaries and public goods. Furthermore, it is for lawyers to remind economists that it is insufficient to consider only the size, or quantitative aspects, of jurisdictions, and that it is also necessary to consider the institutional design, or qualitative aspects, of jurisdictions. This leads us back to CCBA.
After all, what would a subsidiarity "impact analysis" look like other than a comparative cost-benefit analysis of central versus local action? The solution to the "trade and . . . problem" that maximizes the net sum of trade benefits and regulatory benefits is CCBA. This analysis is of course incomplete unless the possibility of centralized action is taken into account: while local action may be better than local inaction, central action may be better than both. The conventional formulation of subsidiarity analysis is a rebuttable presumption of the converse case, in which central action is presumed worse than either local action or local inaction.
Sadly, we cannot simply refer the subsidiarity/ level of governance decision to a brilliant economist, or even a team of brilliant economists, for resolution through CCBA. The first, and most important, is the problem of interpersonal comparison of utility. No economist can tell us the value of our preferences. "In systems of multi-tiered government, different views about what to regulate and what to leave to the market can be held among the constituent states and/or among the vertical layers of government." Economic analysis, at least formal economic analysis as it stands today, provides only limited tools in this context, due to the fact that these preferences are often not monetized, and are not revealed through market mechanisms. Thus, it is essential not simply to divine the solution to the CCBA problem, but to design institutional mechanisms that will allow either direct or indirect individual input to the CCBA problem: to reveal preferences. Often, the appropriate institutional mechanism will simply be the market. Under other circumstances, political institutions may better allow preferences to be revealed and integrated.
This article has argued that while CCBA can help us to choose institutions, and, as applied by courts may provide solutions to trade and . . . problems that maximize the net benefits of trade and regulation, CCBA has limitations. These limitations are intrinsic to CCBA, but also are dependent on the particular institutional structure in which the decisions are made. Various simplified or truncated devices, and various institutional fora, for making these tradeoffs, may be indicated in different factual contexts. "Social scientists have concluded from their studies that decision-making shortcuts are appropriate for relatively unimportant decisions, and fuller optimization is worth the time for major ones." In addition, it is clear that courts or dispute resolution tribunals may not be the best place to engage in CCBA; rather, the redistributive question always raised by potential Pareto efficiency is seen as the natural province of legislatures. Finally, legislatures overcome the problems of interpersonal comparison of utility insofar as they are places where preferences are revealed and collated directly.
The question of which preferences to express at a lower level and which to express at a higher level is the question at the core of subsidiarity analysis. Just as each of our decisions is made through cost-benefit analysis, this article has argued that the choice of level would be made most accurately by cost-benefit analysis. However, this point is only one input in the choice of tradeoff device, which includes as other real-world inputs costs of administration and error and distributive legitimacy, as well as problems of commensuration and interpersonal comparison of utility.
It is well to repeat that CCBA is inevitably political, and is never neutral. How could it be different, given that it seeks to bring together diverse preferences? Thus, it is important to approach CCBA with modesty, and to recognize that the best role for CCBA is in support of preference-revelation mechanisms like politics or markets.
Finally, it may be worthwhile to suggest some testable hypotheses for further research. A new institutional economics theoretical perspective might yield a hypothesis that where political transaction costs are low for the production of central legislation, the political system would be used to make judicial tradeoffs of the kind discussed here. Thus, in circumstances of low "legislative" transaction costs, a narrow NTR rule would be sufficient, referring the more difficult decisions to political decision-making. On the other hand, in circumstances of high political transaction costs, it may become more attractive to accept tradeoff decisions made by courts, suggesting a LTRAT, BT, CBA or CCBA rule. It is incumbent upon society to make tradeoffs: the only question is which institutional mechanism should be used. In order to test this hypothesis, it would be necessary to examine particular circumstances to understand the move from NTR to NP or BT and sometimes, back again. For instance, can we explain the Keck decision in these terms? The line of argument might point out that prior to the legislation of the Single European Act, schlerotic legislative capacity gave impetus to the ECJ's development of the Cassis de Dijon line of jurisprudence, in order judicially to make the tradeoff decisions that could not be made legislatively except at high transaction cost. The Single European Act facilitated central legislation, diminishing this motivation for ECJ action. On the other hand, the growth of at least LTRAT in the GATT/WTO system may be interpreted as a reaction to the inability to legislate easily in that context. Again, decisions must be made, and NTR alone seems to leave too much on the table. Anecdotes such as these could be reviewed with precision and grouped together, to determine whether this hypothesis could be used to predict the use of tradeoff devices in particular circumstances.
 See Eugene D. Cross, Pre-Emption of Member State Law in the European Economic Community: A Framework for Analysis, 29 COMM. MKT. L. REV. 447 (1992).
 See Jean-Paul Jacqué & Joseph H.H. Weiler, On the Road to European Union--A New Judicial Architecture: An Agenda for the Intergovernmental Conference, 27 COMMON MKT. L. REV. 185, 202-203 (1991). See also Tommaso Padoa- Schioppa, Economic Federalism and the European Union, in RETHINKING FEDERALISM: CITIZENS, MARKETS AND GOVERNMENTS IN A CHANGING WORLD 154, 155 (Karen Knop et al. eds. 1995).
 As in Canada and Australia. Jacqué and Weiler point out that systems that start out based on lists, on rules, often move quickly to principles of allocation of competence, or standards. Id. at 203.
 See Bermann, supra note 251, at 333-334, and sources cited therein. Indeed, the purpose of Bermann's article is to suggest how subsidiarity might be taken seriously.
 See Kommers & Waelbroeck, supra note 9, at 226: "In the United States, commerce clause jurisprudence has been a particularly cogent vehicle for interpreting the federal nature of the American political system."
 Komonchak, supra note 247, at 299.
 For other discussions of the relationship between subsidiarity and proportionality, see RENAUD DEHOUSSE, DOES SUBSIDIARITY REALLY MATTER? 14 (1993) (arguing that subsidiarity involves an efficiency assessment, while proportionality only considers means and ends); The Principle of Subsidiarity, Communication of the Commission to the Council and the European Parliament, Doc. SEC (92) 1990 final, October 27, 1992, at 5; Jacqué & Weiler, supra note 270, at 206.
 Padoa-Schioppa, supra note 270, at 155. Padoa-Schioppa adds the following:
While the general principle is clear cut, its application may not be so simple. . . . I shall not try to explore here why the body of economic analysis stemming from this central proposition is so small and only just beginning to develop, but a large part of the answer probably lies in two factors: first the analytical complexity of removing the customary 'one economy, one government' assumption; and second, the scant attention paid by modern economists to the institutional dimension of policies.
Id. at 155-56.
 Bermann, supra note 251, at 379.
 Jacques Pelkmans, Governing European Union: From Pre-Federal to Federal Economic Integration, in RETHINKING FEDERALISM: CITIZENS, MARKETS AND GOVERNMENTS IN A CHANGING WORLD 166, 169 (Karen Knop et al. eds. 1995).
 Gottlieb, supra note 22, at 855, citing Helmut Jungermann, The Two Camps on Rationality, in JUDGMENT AND DECISION MAKING: AN INTERDISCIPLINARY READER 627, 633 (Hal R. Arkes & Kenneth R. Hammond eds., 1986)
 Dehousse has pointed out the "dual subsidiarity" at work in the EU: "subsidiarity with respect to the [EU's] main raison d'être, namely market integration, and subsidiarity with respect to national regulatory policies." Dehousse, supra note 109, at 388.
 This modest role for CCBA is greater than that suggested by Kelman, who argues that "the notion that there even exists a separate procedural technique for evaluating the polity, grounded in simple responsive preference aggregation, is, while not invariably wrong or self-deluding, so frequently wrong that it is almost certainly a pernicious myth." Kelman, supra note 119, at 150.
 Similarly, and in accordance with the views of Kitch, Regan and Scalia, in circumstances of low political transaction costs for bargaining between constituent units, a narrow NTR rule might also be sufficient.
Top of the page