Trade and ... Problems, Cost-Benefit Analysis and Subsidiarity

1 The Trade-off Problem and the Structure of the Argument

A The Trade and ... Problem

With the reduction of tariffs and quotas since the inception of the General Agreement on Tariffs and Trade (`GATT') in 1947, conflicts between trade values and other social values have arisen. While these conflicts are now apparent, their contours and ramifications are largely uncharted. Furthermore, the explicit formulas or approaches provided to resolve these conflicts in treaties, constitutions and precedent seem incomplete and incoherent.

It is obvious, but not always accepted, that neither trade values nor other social values are, by themselves, preeminent.² Rather, we are forced to choose the extent to which each value is to be implemented: i.e., we must make trade-offs among these values. We do so through legislative and adjudicative processes. The main way that adjudicative bodies have made trade-offs in the trade arena is through rules that relate the burden on international or interstate trade to the local, or at least locally-determined, regulatory benefit, including anti-discrimination rules,³ simple means-ends rationality tests, least trade restrictive alternative tests (including the `necessity' test under Article XX of GATT), proportionality tests stricto sensu, balancing tests and perhaps cost-benefit analysis. While diverse, these doctrines share a common feature - the willingness to juxtapose, and in many cases to commensurate between, trade values on the one hand and non-trade values on the other.

This article engages in a comparative institutional analysis of formulaic and other institutional means to address the conflict between trade values and other social values.⁴ The means examined are used by adjudicative bodies in the European Union, the World Trade Organization and the United States.

B The `Trade and ... Problem' as a Multi-level Institutional Choice Problem

The `trade and ... problem' manifests itself of course in particular circumstances, and each circumstance must be addressed separately, except to the extent that benefits arise from analysing similar problems together. This problem is one of synthesizing and maximizing complex preferences in the context of multiple overlapping communities.

If we had a discrete community to govern, the problem would be much more manageable. This `island' community would simply leave it to its political process to determine how much of each `good' it wants or, alternatively, to determine when the market should do so. While both the political process and the market are no doubt imperfect, we have no other forum. Politics is often the default means for choosing among incommensurables, such as allocative efficiency (which can be measured in currency, albeit with costs and limitations) and environmental protection (which is considerably more difficult, although perhaps not impossible, to monetize). By the same token, the reason we do not have a single global community to govern is because there are important benefits in diversity, which allows smaller groups of people to express and satisfy different preferences more effectively.⁵ However, there are horizontal spillovers that cause one group to be affected by the legal rules and policy decisions of another group. Finally on this point, the `trade and ... problem' may also be interpreted as one of overlapping vertical allocation of power, wherein the discrete organs of vertical power, including the state, regional organizations and global organizations, must share power in particularized ways. That is, broad allocations of power, over `trade' or `health' or `environment', are found to overlap, and these overlaps must be reconciled. For example, in the United States, the Constitution assigns potentially complete power over interstate trade to the federal government, while the states are generally competent in relation to health regulation issues. Given the overlap between these issues, the `trade and ... problem' amounts to a question of federalism.

Therefore, the `trade and ... problem' is one of intersecting jurisdictions, on both horizontal and vertical axes, each with varying interests. If the conflict between the market and regulation were a chess game, then the conflict between trade values, or more generally international values, and other social values would be a three-dimensional game, with geometrically increased complexity. In addition to choosing between laissez-faire and intervention, the level of intervention must also be selected.

Institutional choice has multiple parameters. The first parameter to be addressed is the vertical level of society at which choice takes place. Second is the type of institution - for example, legislative versus adjudicative - to be assigned the task of choice. Third is the rule that the selected institution will follow. This article focuses on this last parameter as a device, applied at a central adjudicative level, to select between the assignment or denial of power to local legislatures. However, the rule applied at the central adjudicative level may also determine the choice between central adjudication and legislation as the institutional setting for decision.

C A Taxonomy of Trade-off Devices

In this section, we examine the trade-off devices used by the European Court of Justice (ECJ), GATT or WTO dispute resolution panels, or the WTO Standing Appellate Body, and the US Supreme Court in connection with `trade and ... problems' within the EU, GATT/WTO and US systems, respectively. The major categories of trade-off devices are listed and briefly defined below. In each of the jurisdictions studied, these trade-off devices appear in combination, rather than alone, and each category of device conceals considerable latitude for heterogeneity. Thus, despite the list of only six categories, far more combinations and variations are possible.

(i) National treatment rules. This is a type of anti-discrimination rule that examines whether different legal standards are applied to comparable cases, as between the domestic and the foreign. National treatment rules entail surprising complexity. In order to deal with more difficult cases, they sometimes incorporate some of the tests set forth below in this list.

(ii) Simple means-ends rationality tests. These tests consider whether the means chosen is indeed a rational means to a purported end. Simple means-ends rationality testing is often combined with limitations on ends. Analytically, this form of testing is included in all of the tests described below, and is sometimes used as a proxy to detect discrimination. As it imposes little real discipline, and is often included in other tests, this test is not analysed in detail below.

(iii) Necessity or least trade restrictive alternative tests. This type of test goes a significant step beyond simple means-ends rationality testing. It inquires whether there is a less trade restrictive means to accomplish the same end. The definition of the end is often outcome-determinative. In some cases necessity testing is qualified by requiring that the means be the least trade restrictive alternative that is reasonably available. In addition, necessity testing is sometimes combined with limitations on the categories of ends permitted.

(iv) Proportionality. Proportionality stricto sensu⁶ examines whether the means are `proportionate' to the ends: whether the costs are excessive in relation to the benefits. It might be viewed as cost-benefit analysis with a margin of appreciation, as it does not require that the costs be less than the benefits. Proportionality may be either static or comparative, in the same way as cost-benefit analysis. A comparative approach to proportionality testing would include in its calculus the costs and benefits of alternative rules.

(v) Balancing tests. These tests purport to decide whether a measure that impedes trade is acceptable, balancing all of the factors. Balancing may be viewed as a kind of amorphous or imprecise cost-benefit analysis.⁷ More charitably, and perhaps more correctly, it may be viewed as a kind of cost-benefit analysis that recognizes the difficulty of formalizing the analysis, and seeks to achieve similar results informally.⁸

(vi) Cost-benefit analysis. Static cost-benefit analysis in the context at hand⁹ juxtaposes the regulatory benefits and the trade costs of regulation, as well as other costs involved, and would strike down regulation where the costs exceed the benefits. Cost-benefit analysis in this context may be viewed as stricter scrutiny than the domestic cost-benefit analysis that has recently become popular, as it adds a cost dimension not normally included, i.e. detriments to trade. Adding trade detriments to the calculation would presumably have the marginal effect of causing some regulation to fail a cost-benefit analysis test. It is worth comparing static cost-benefit analysis, simply juxtaposing the costs and benefits of a single rule, with a more dynamic comparative cost-benefit analysis, comparing the net benefits of multiple rules, and recommending the rule with the greatest net benefits.

D Toward Comparative Institutional Analysis

This article recognizes, following Coase,¹⁰ Demsetz,¹¹ Komesar,¹² Wolf¹³ and countless others, that neither the market nor the state are perfect or perfectible institutions, but that our existential task is to choose the least imperfect combinations of institutions. `In a world of institutional alternatives that are both complex and imperfect, institutional choice by implication, simple intuition, or even long lists of imperfections is deeply inadequate.'<sup>14</sup> However, this article extends the analysis of these authors by asking which state and which market. In seeking an answer to this question, it evaluates different markets (from the private market to the `market' among competitive governments, and beyond) and different governmental entities at different vertical levels as appropriate repositories of authority.

The article examines strategies for management established pursuant to constitutional or treaty language and used in dispute resolution fora in the EU's common market, in the multilateral trade system under the GATT and WTO and in the United States' internal common market.¹⁵ While these strategies are based on legislative or constitutional texts, the texts are consistently indeterminate - perhaps more than most laws - and thus the task of constructing strategies has often fallen on dispute resolution bodies.

Within this comparative analysis, it is important to keep in mind the two leading alternatives to the trade-off exercise as a means to moderate between trade values and other social values: (i) laissez-régler (used here to denote a permissive attitude taken by the international system, allowing local governments freedom to regulate in the domestic sphere) and (ii) international regulation (a decision to moderate between these values in a more specific, and in a positive,¹⁶ international legislative manner). The first alternative may allow the erosion of international commitments in ways that may be unacceptable in at least some international economic law settings, but may be acceptable in other settings where few externalities exist or where states may make ad hoc bargains at low transaction costs. In fact, mechanisms for managing the conflict between trade values and other social values have the effect of constraining state intervention, either in favour of laissez-faire, or, where combined with international legislative devices, in favour of international regulation. `The modern regulatory state inevitably produces burdens on trade, if only because of the unavoidable lack of regulatory uniformity.'¹⁷ Petersmann argues in favour of international disciplines on national regulation - against laissez-régler - in order to protect laissez-faire.¹⁸

A `laissez-régler' approach to local regulation means decentralizing decisions about regulation. Kitch explains why decentralization is not necessarily the enemy of free trade, arguing that centralized supervision or control is only one way that local units can cooperate to achieve their goals. The fact that there is decentralized authority over the laws and government practices affecting commerce does not mean that there will not be free trade. Free trade among decentralized authorities will result from voluntary cooperation, motivated by the fact that it will produce greater wealth for all to share. In the short run, this approach to free trade may cause significant bargaining instability, as each jurisdiction tries to establish a bargaining position through bluff, threat and implemented threat. But in the long run, it may provide more free trade than centralized authority because it places stronger incentives on each jurisdiction to promulgate efficient rules for both its internal and external commerce.¹⁹

Kitch implicitly compares two different centralizing structures: one mandatory and the other voluntary. As North has pointed out, Kitch's perspective seems to be based on an assumption that it is cheaper in transaction cost terms for states to get together on an ad hoc basis to cooperate than it is for this cooperation to be imposed by the federal government: He makes an assumption as to which is the more efficient instrument of cooperation.²⁰ North responds that we do not `know that decentralized authority would promote more efficient rules than would centralized authority'.²¹ The trade-off devices examined herein may be viewed as heuristics for determining, in particular settings, whether decentralized or centralized authority is more satisfactory.

As instruments of negative integration, these trade-off devices may serve another dynamic purpose, providing incentives for positive international regulation where they strike down domestic regulation. Furthermore, they clarify and cull the appropriate topics of, and scope for, international regulation by indicating what domestic regulation is acceptable. Once domestic regulation is identified as acceptable pursuant to the rules applied by courts, it is for the legislative process to determine whether the international values are great enough to justify superseding domestic law by international regulation. In this respect, these trade-off devices may serve to allocate work between adjudicative and legislative decision-making processes.²²

From a horizontal, as opposed to vertical, perspective, these trade-off devices may be viewed as intended not to limit local autonomy, but to restrain `state interference in the affairs of other states'.<sup>23</sup> Thus, local autonomy is on both sides of the equation, although in some instances it is represented by international institutions. `Interference arises from two basic causes, state protection of local commerce against external competition, and extra costs that result when more than one sovereign regulates or taxes the same person or transaction. The latter costs are of two kinds - multiple burdens, and conflict costs caused by inconsistent regulation.'²⁴

In current or static terms, trade-off devices serve as heuristics for determining when domestic regulation should be suppressed. They moderate between the domestic (laissez-régler) and the international on a case-by-case basis. In intertemporal terms, perhaps they serve in some cases as bridges through time from laissez-régler to international regulation.

Pursuing a roughly comparative methodology,²⁵ this analysis finds significant similarities in the texts and approaches applied in the three jurisdictions examined. Beginning with comparative cost-benefit analysis as a presumptively best alternative, the article seeks to comprehend moves to other approaches based on problems with cost-benefit analysis and seeks to explain variations among these other approaches. These relationships cannot be drawn precisely, as there are many variables and only a small number of cases to compare, but it is hoped that lines of further inquiry will emerge.

2See Dunoff, `Reconciling International Trade with Preservation of the Global Commons: Can We Prosper and Protect?', 49 Wash. & Lee L. Rev. (1992) 1407, at 1449-50. See also Stewart, `International Trade and Environment: Lessons from the Federal Experience', 49 Wash. & Lee L. Rev. (1992) 1329 ( `It thus seems appropriate, for the present, to proceed on the premise that environmental protection policy as well as trade policy are both appropriately aimed at promoting, in different ways, human welfare, broadly understood.')

3 The careful reader will object that anti-discrimination or national treatment rules do not relate regulatory benefits to trade burdens, but simply prohibit application of more burdensome standards to non-nationals. However, as will be shown below, in cases other than those of de jure or intentional discrimination, these rules inevitably compare regulatory benefits and trade burdens.

4 This is a central problem in any federal, regional or international free-trade system. Other scholars have addressed these issues in important articles, including Barcelo, `Product Standards to Protect the Local Environment - the GATT and the Uruguay Round Sanitary and Phytosanitary Agreement', 27 Cornell Int'l L.J. (1994) 755, at 759-760; Farber and Hudec, `Free Trade and the Regulatory State: A GATT's-Eye View of the Dormant Commerce Clause', 47 Vand. L. Rev. (1994) 1401 (comparing dormant commerce clause analysis with GATT Art. XX analysis); Klabbers, `Jurisprudence in International Trade Law: Article XX of the GATT', 26 J. World Trade (1992) 63; Stewart, supra note 1; Kommers and Waelbroeck, `Legal Integration and the Free Movement of Goods: The American and European Experience', in M. Cappelletti, M. Seccombe and J. Weiler (eds.), Integration through Law: Europe and the American Federal Experience (1986).

5 See, e.g., Trachtman, `International Regulatory Competition, Externalization and Jurisdiction', 34 Harv. Int'l L.J. (1993) 47.

6 N. Emiliou, The Principle of Proportionality in European Law: A Comparative Study (1996), 6. A wider definition of proportionality developed in the EU context includes three tests: (i) proportionality stricto sensu, (ii) a least trade restrictive alternative test, and (iii) a simple means-ends rationality test. This article will consider only the narrower type of proportionality.

7 See Smith, `State Discriminations against Interstate Commerce', 74 Cal. L. Rev. (1979) 1203, at 1205 (`. . . the Justices take all relevant circumstances into account and render judgment according to their overall sense of the advantages and disadvantages of upholding the regulation'). At their most precise, balancing tests are the same as cost-benefit analysis. See Maltz, `How Much Regulation is Too Much - An Examination of Commerce Clause Jurisprudence', 50 Geo. Wash. L. Rev. (1981) 47, at 59-60.

8 `If we had a way of quantifying all the appropriate inputs, and a way of comparing them, and a theory that told us how to do so, we would not call it balancing. Rather, it would be called something like "deriving the most cost-effective solution," or just "solving the problem."' Gottlieb, `The Paradox of Balancing Significant Interests', 45 Hastings L.J. (1994) 825, at 839. See also Aleinikoff, `Constitutional Law in the Age of Balancing', 96 Yale L.J. (1987) 943, at 1002-1004.

9 For more general and technical treatment of cost-benefit analysis, see, e.g., P.S. Menell and R.B. Stewart, Environmental Law and Policy (1994), at 81-160. D. Pearce and C. Nash, The Social Appraisal of Projects: A Text in Cost-Benefit Analysis (1981); R. Tresch, Public Finance: A Normative Theory (1981); E. Stokey and R. Zeckhauser, A Primer for Policy Analysis (1978); E.J. Mishan, Cost-Benefit Analysis (1976); H. Raiffa, Decision Analysis (1968).

10 R. Coase, The Firm, the Market and the Law (1988), at 95-185 incorporating and commenting upon earlier work, including Coase's seminal articles: `The Nature of the Firm', 4 Economica (1937) 386, and `The Problem of Social Cost', 3 J.L. & Econ. (1960) 1. See also Idem, `The Nature of the Firm: Influence', 4 J.L. Econ. & Org. (1988) 33, at 33.

11 Demsetz, `Information and Efficiency: Another Viewpoint', 12 J. L. & Econ. (1969) 1.

12 N. Komesar, Imperfect Alternatives (1994). For an application of Komesar's approach to the types of problems addressed here, see Poiares Maduro, `Reforming the Market or the State? Article 30 and the European Constitution: Economic Freedom and Political Rights', 3 European Law Journal (1997) 55.

13 C. Wolf, Markets or Governments: Choosing Between Imperfect Alternatives (1988).

14 Komesar, supra note 11, at 6.

15 See Collins, `Economic Union as a Constitutional Value', 63 NYU L. Rev. (1988) 43.

16 It is common to distinguish between negative integration, by virtue of invalidating local rules that burden commerce, and positive integration, by virtue of central legislation that preempts or supercedes the local law.

17 Farber and Hudec, supra note 3, at 1402.

18 E.-U. Petersmann, Constitutional Functions and Constitutional Problems of International Economic Law (1991), at 210-221; see also Tumlir, `Need for an Open Multilateral Trading System', 6 World Econ. (1983) 393, at 406.

19 Kitch, `Regulation and the American Common Market', in A.D. Tarlock (ed.), Regulation, Federalism and Interstate Commerce (1981), at 13-14. But see Olson, `The Principle of Fiscal Equivalence: The Division of Responsibilities among Different Levels of Government', 59 Am. Econ. Rev. (1969) 479, at 480-81 (explaining why simple bargaining is insufficient to achieve Pareto optimality under circumstances of positive transaction costs). See also Inman and Rubinfeld, `A Federalist Fiscal Constitution for an Imperfect World', in H.N. Scheiber (ed.), Federalism: Studies in History, Law and Policy (1988).

20 North, `An Economist's Perspective on the American Common Market', in Tarlock, supra note 18, at 78.

21 Ibid. See also Trachtman, `The Theory of the Firm and the Theory of the International Economic Organization: Toward Comparative Institutional Analysis', 17 Northw. J. Int'l L. & Bus. (1997) 470.

22 `If the Contracting Parties were to decide to permit [environmental] trade measures ..., it would be preferable for them to do so not by interpreting Article XX, but by amending or supplementing the provisions of the General Agreement ...' United States - Restrictions on Imports of Tuna, 39 BISD (1993) 155, 204, para. 6.3 reprinted in 30 ILM (1991) 1594 [hereinafter `First Tuna Panel Report'].

23 Collins, supra note 14, at 109. In this regard, the problem can be viewed as having a reciprocal nature. For an analysis of the reciprocal nature of property rights, see Coase, supra note 9.

24 Farber and Hudec, supra note 3, at 1402.

25 See Lijphart, `Comparative Politics and the Comparative Method', 65 Am. Pol. Sci. Rev. (1971) 682.

© 1990-2004 European Journal of International Law
All comments and suggestions should be sent to webmaster
This site is part of the Academy of European Law online, a joint partnership of the Jean Monnet Center at NYU School of Law and the Academy of European Law at the European University Institute.
This file was last modified: Tuesday, October 14, 2003 12:47PM