Jean Monnet Center at NYU School of Law



Previous |Next |Title


3. Comparison of Actual Tradeoff Devices


We continue our evaluation of CCBA by examining the alternatives extant in the same terms by which we evaluated CCBA, albeit more briefly. The tradeoff devices examined here, other than CCBA, are in use, to varying extents, in varying combinations and with varying effect, in the three jurisdictions considered here. The following table summarizes, in gross, the use of these tradeoff devices in the EU, GATT/WTO and U.S. A more detailed analysis would reveal great diversity within each category.

Table A: Tradeoff Devices in EU, GATT/WTO and U.S. Law





European Union

GATT/WTO

U.S.

Textual Sources

Arts. 30 and 36

Arts. III, XI and XX

Commerce Clause

NTR

required

required

required

SMERT

required

only required under XX after finding of violation of III or XI; included in "like products" analysis under III

required

LTRAT

required

only required under certain provisions of XX after finding of violation of III or XI

required after a finding of de jure discrimination or "discriminatory effect"--rarely satisfied

NP

required

not required

required, but liberal

BT/CBA

perhaps required

not required

perhaps required, but liberal

a. Textual and Political Contexts of Analysis

None of EU, GATT/WTO or U.S. law explicitly prescribes any of the above tests, with the exception of NTR.[142] All of the tests analyzed below have been judicially[143] cultivated on relatively stark textual bases, at least at first. They have met with political acquiescence and in some cases political approval, but have suffered attacks alleging illegitimacy on varying grounds, including the lack of a textual basis.[144]

Thus, even if legislatures or framers of constitutions and treaties did not intend to mandate these tradeoff analyses, judges invented them. They did so not necessarily to increase their bureaucratic power, but in order to fill a gap that required filling in order to decide cases: the gap in clarity of allocation of competences between the center and the periphery. Only in a limited number of areas are all local impediments to free trade invalidated; only in a limited number of areas are local actions invulnerable to central judicial review. In these clear areas, allocation of authority along the vertical axis requires little judicial analysis, but only categorization. However, in other areas, notably the field where the majority of problems involving local initiatives that impede free trade are located, the local initiatives are neither always prohibited nor always permitted. The devices described below are judicially-created devices to discriminate between those to be prohibited and those to be permitted.

Of course, saying that these devices lack strong textual foundation and are judicially-created is not to say that the language of the texts on which they are based is unimportant. However, it is fair to say that these texts serve only as a starting point of analysis. In the case of the U.S. Constitution and the Treaty of Rome,[145] it was recognized by the relevant judicial bodies that in order to create a common market, local laws would need to be disciplined. In the U.S., federal legislation was from an early point available to discipline local laws, but the pre-emption doctrine significantly enhanced protection from localism:

I do not think the United States would come to an end if we lost our power to declare an Act of Congress void. I do think the Union would be imperiled if we could not make that declaration as to the laws of the several States. For one in my place sees how often a local policy prevails with those who are not trained to national views and how often action is taken that embodies what the Commerce Clause was to end.[146]

In the European Union, central legislation by majority vote was not and is not always possible, putting more pressure on judicial supervision of local law.[147] The ECJ "did not receive the power to declare the law of a Member State void . . . . but went as far as it could to reach the same practical outcome . . . ."[148] Nor did the WTO receive the power to declare the law of a member state void, or even perhaps to require a member state to change its law, but it does have the power to declare a member state law in violation of WTO law.

i. Arts. 30 and 36 of the Treaty of Rome

Article 30 of the Treaty of Rome is similar to article XI of GATT discussed below: it prohibits quantitative restrictions and all measures having equivalent effect.[149] This language was interpreted quite broadly in Dassonville,[150] but this breadth of prohibition required a countervailing principle to protect certain national measures: it was not desirable that all national measures that affect EU trade be struck down. Article 36, providing exceptions to the free trade principles of article 30, may be analogized to article XX of GATT discussed below.[151] While the exceptions in article 36 have been interpreted as limited to those explicitly mentioned,[152] and have been construed narrowly,[153] the ECJ has developed the "rule of reason" under article 30 to provide broader exceptions.[154]

Obstacles to movement within the Community resulting from disparities between the national laws relating to the marketing of the products in question must be accepted in so far as those provisions may be recognized as being necessary in order to satisfy mandatory requirements relating in particular to the effectiveness of fiscal supervision, the protection of public health, the fairness of commercial transactions and the defense of the consumer.[155]

Cassis de Dijon confirmed that the proscription of article 30 is applicable even to indistinctly applicable measures.[156] However, the rule of reason articulated in Cassis de Dijon allows measures that are indistinctly applicable to be justified under article 30 on the basis of the general interest.[157] These national measures must be "reasonable," meaning that they must be necessary in order to promote the general interest. Thus the same principle of proportionality that circumscribes the use of exceptions under article 36 also circumscribes the rule of reason. The rule of reason may be viewed as both (a) an expansion of the category of permitted ends in national legislation from those listed in article 36, and (b) an extension of the principle of proportionality. Its mandatory requirements--permitted ends--are not a closed set.[158]

It is perhaps worth noting here that the 1993 Keck and Mithouard judgment of the ECJ may be interpreted as a curtailment or categorical reduction of the scope of article 30. By reserving non-discriminatory marketing regulation to the exclusive jurisdiction of the member states, the ECJ establishes a category of cases that will not be subject to further judicial review. This allocation of jurisdiction avoids the costs of more specific determinations, and is at least partly congruent with a territorial approach to legislative jurisdiction. In other words, the category of restraints on marketing methods is an area "where inter-state trade considerations would seem to be of peripheral importance."[159] It is in this sense that the Keck judgment renders "explicit what was implicit in Cassis de Dijon: the 'home state control' principle provides not just an institutional structure for the internal market but also delimits the scope of the internal market, or at least what activities should be regulated by the Community institutions."[160]

It is instructive that the GATT has not developed a rule of reason under article XI to perform a similar function. GATT has not needed a rule of reason because article III has been interpreted to insulate non-discriminatory internal product regulation from further scrutiny under article XI, and because article XI of GATT has not been interpreted as broadly as article 30 of the Treaty of Rome.[161] Thus, as the scope of GATT's aspirations were smaller, at least prior to the conclusion of the Uruguay Round, it did not scrutinize such non-discriminatory regulation at all, adhering to laissez-régler, and therefore, it might be argued, needed no heuristic to differentiate between permissible and impermissible non-discriminatory measures. Important categorical exceptions to the GATT rule of laissez-régler have developed, such as the product-process distinction, and revised treatment under the Standards Agreement and S&P Agreement, exerting more pressure on the exception provisions of article XX and of these agreements.

ii. Necessity under GATT Article XX and under the Uruguay Round Agreement on Technical Barriers to Trade

Article XX of GATT provides general exceptions from any relevant obligations under the GATT, and thus, inter alia, provides exceptions to obligations of national treatment under article III of GATT, and from obligations to eliminate quantitative restrictions under article XI of GATT. National measures that are subject to article III and that are found to satisfy the requirement of national treatment-type non-discrimination seem to be excluded from the stricter scrutiny of article XI.[162] The recent GATT panel decisions relating to U.S. restrictions on imports of tuna pursuant to the U.S. Marine Mammals Protection Act[163] are examples of the converse of this principle: article III was found inapplicable, and therefore article XI was applicable. It is of critical importance to the comparison with the U.S. commerce clause and EU law that, at least under GATT itself, non-discriminatory domestic regulatory measures that relate to the product as such,[164] even if applied at the border to imported products, are subject to no further scrutiny. Such measures may be disproportionate in the sense that their cost in trade terms may be far greater than their benefit in regulatory terms, or there may be clearly less trade restrictive alternatives available, but they are not subject to such scrutiny. As discussed below, however, proportionality analysis may be brought to bear to determine whether the domestic measure is indeed non-discriminatory.

Given the relatively unforgiving scrutiny of article XI, due to the lack of a "rule of reason," where measures are subject thereto, as in the two tuna cases, the central discourse addresses the availability of an exception under article XX. Of course, where discrimination violating article III, or any other GATT violation for that matter, is found, the discourse would also generally turn to article XX. The recent WTO panel decision and SAB decision regarding the U.S. Clean Air Act reformulated gasoline regulations are examples of discrimination being found under article III, shifting the discourse to article XX.[165]

The general exceptions under article XX that will be discussed here are those for measures necessary to protect human, animal or plant life or health (XX(b)); necessary to secure compliance with laws or regulations which are not inconsistent with the provisions of the GATT (XX(d)); and relating to the conservation of exhaustible natural resources if such measures are made effective in conjunction with restrictions on domestic production or consumption (XX(g)). In addition, the chapeau of article XX requires that measures qualifying for relief thereunder are not applied in a manner that would constitute a means of arbitrary or unjustifiable discrimination, or a disguised restriction on trade.

Furthermore, the Uruguay Round Agreement on Technical Barriers to Trade (the "Standards Agreement") and the Uruguay Round Agreement on the Application of Sanitary and Phytosanitary Measures (the "S&P Agreement") contain similar and additional disciplines. For example, the Standards Agreement provides that technical regulations may not be adopted or applied with a view to or with the effect of creating unnecessary obstacles to international trade. The next sentence seems to interpret this restriction as LTRAT--LE: "For this purpose, technical regulations shall not be more trade restrictive than necessary to fulfil a legitimate objective, taking account of the risks non-fulfillment would create."[166] Thus, this agreement seems to eliminate the possibility that non-discriminatory technical regulations[167] would be protected from further scrutiny. It may also be possible to interpret the necessity language as calling for BT or CBA.[168]

The S&P Agreement, like the Standards Agreement, is intended to elaborate on particular provisions of GATT, especially on article XX(b) of GATT.[169] The basic provisions on necessity are similar to those in article XX itself and in the Standards Agreement.

iii. The U.S. Commerce Clause and Dormant Commerce Clause.

"[A] chief occasion of the commerce clauses was 'the mutual jealousies and aggressions of the States, taking form in customs barriers and other economic retaliation.'"[170] The Commerce Clause of the U.S. Constitution is on its face simply a positive allocation of legislative power to Congress, to "regulate Commerce with foreign Nations, and among the several States. . . ."[171]

This plenary allocation to Congress of power over commerce begs two relevant questions. First, what is meant by interstate commerce, given that any commerce that is interstate is also at least in part intrastate, and that it is possible to find an interstate link for virtually all intrastate commerce?[172] This is an example of functional overlap, and of the indeterminacy of categories used to allocate power. Second, to what extent is this power exclusive, depriving the states of power to regulate whatever is deemed interstate commerce?[173] (Note that this latter problem was raised, in a somewhat different sense, by Dassonville in the EU.) It is well-understood that the commerce clause provides pre-emptive power to Congress, when Congress acts through positive legislation.[174] "But it has been settled for more than a century that the [commerce] Clause prohibits States from taking certain actions respecting interstate commerce even absent congressional action."[175] Thus, the commerce clause serves in this context as a basis for pre-emption of state regulation. This aspect of the commerce clause is known as the "dormant" or "negative" commerce clause.[176]

Of course, as indicated above, the commerce clause has not been self-executing. Rather, it has been necessary for the Supreme Court to determine what kinds of state actions would be struck down as unconstitutional interferences with the federal commerce power. In Brown-Forman Distillers v. New York Liquor Authority,[177] the Court described its two-tier approach:

When a state statute directly regulates or discriminates against interstate commerce, or when its effect is to favor in-state economic interests over out-of-state interests, we have generally struck down the statute without further inquiry. When, however, a statute has only indirect effects on interstate commerce and regulates evenhandedly, we have examined whether the State's interest is legitimate and whether the burden on interstate commerce clearly exceeds the local benefits. We have also recognized that there is no clear line separating the category of state regulation that is virtually per se invalid under the Commerce Clause, and the category subject to the Pike v. Bruce Church balancing approach. In either situation the critical consideration is the overall effect of the statute on both local and interstate activity.[178]

On the other hand, Prof. Tribe cautions that the "Supreme Court's approach to commerce clause issues, despite such structuring devices as the emphasis on less restrictive or discriminatory alternatives, often appears to turn more on ad hoc reactions to particular cases than on any consistent application of coherent principles."[179]

b. National Treatment Rules (NTR)

i. NTR Defined

All three jurisdictions studied here forbid discrimination that violates NTR. A number of scholars and some judges argue that NTR is the only tradeoff rule needed or desirable.[180] The NTR requirement seems to be a sine qua non of integration, and perhaps represents the fundamental economic hallmark of political integration: treat foreigners (or their goods) at least as well as you treat locals (or their goods).[181] It might be argued that while NTR rules seem to be motivated more strongly by political considerations,[182] and while NP and SMERT are ways of identifying de facto discrimination, BT and CBA have more explicitly economic motivations. But NTR is inherently unstable, and with time and pressure seems to metamorphose into more demanding tests, including SMERT, NP, LTRAT and BT or CBA. Furthermore, NTR is insufficient to resolve the problems of multiple regulation, regulation that fails to take account of legitimate diversity and regulation that is eccentric and thus inappropriately different from other states' regulation. We cannot accept the proposition that NTR is sufficient without assessing the costs of leaving these problems unaddressed.[183]

(1) European Union

Article 7 of the Treaty of Rome prohibits discrimination on grounds of nationality, including discrimination against goods of foreign origin. However, discrimination on the basis of national origin is sufficient to establish a violation of article 30 of the Treaty of Rome, which, being more specifically related to goods, would cover a cause of action relating to denial of NTR more specifically.

However, as noted above, discrimination is sufficient, but not necessary, for article 30 scrutiny (except, after Keck, with respect to regulation of marketing arrangements). In addition, arbitrary discrimination would render unavailable any exception under article 36, or under the "rule of reason," as would a finding that de jure non-discriminatory measures constituted a "disguised restriction on trade." Unlike the case under GATT, there is no distinction between internal measures and border measures.[184] This is to be expected, as the EU customs union is intended to eliminate border measures entirely.

Under EU law, it is recognized that "[t]he different treatment of non-comparable situations does not lead automatically to the conclusion that there is discrimination. Material discrimination would consist in treatment of either similar situations differently or different situations identically."[185] Thus, the principle of non-discrimination is not easily applied in cases in which either the discrimination is not de jure or the de jure discrimination has a justification. As explained in more detail below in the context of the "like products" concept under GATT/WTO law, this review of the rationality of the regulatory categories in order to find discrimination involves the ECJ in an intrusive review of national regulatory categories.[186]

(2) GATT/WTO

As noted above, GATT contains in article III a rule of national treatment. National regulatory measures that are subject to but do not violate article III are thought to be protected from the otherwise applicable prohibition under article XI (prohibiting quantitative restrictions, including regulatory provisions that exclude the sale of the relevant goods).[187] Thus, importantly, under GATT itself, non-discriminatory domestic measures need not comply with the requirements for an exception under article XX. This is an important distinction from article 30 of the Treaty of Rome, which applies even to measures that are indistinctly applicable. This protection against article XI scrutiny, and the need for article XX justification, is removed under the Standards Agreement, pursuant to which even non-discriminatory measures must, inter alia, not create "unnecessary obstacles to international trade."[188]

Under the chapeau provisions of article XX, however, no exception is available for discriminatory measures that are applied in a manner that constitutes arbitrary or unjustifiable discrimination, or a disguised restriction on international trade. For example, in its recent reformulated gasoline decision, the SAB found that in addition to constituting a violation of article III, U.S. discrimination was also "unjustifiable" and a "disguised restriction," thereby rendering unavailable the exception under article XX(g).[189] Under the SAB's analysis, contradicting that of the original panel, the U.S. Clean Air Act regulations at issue would otherwise have fallen within the terms of article XX(g), relating to the "conservation of exhaustible natural resources."[190]

Where the discrimination is not de jure,[191] but is hidden, disguised, indirect or incidental, a more difficult analysis is required. However, the standard applied by recent GATT panels is to require treatment consistent with "effective equality of opportunities for imported products in respect of the application of laws, regulations and requirements affecting their internal sale, offering for sale, purchase, transportation, distribution or use of products."[192]

In order to find either de jure or de facto discrimination, it is necessary to determine that the foreign products and the relevant domestic products are "like products" within GATT.[193]

Furthermore, the recent GATT Tuna decisions[194] have held that article III applies to discrimination with respect to products as such, but does not apply to regulatory categories that are not determined by reference to the product as such: regulation of production processes. The GATT panels have refused to apply article III to regulation of production processes because these regulations do not apply to the product as such, denying otherwise non-discriminatory process regulations the protection from article XI scrutiny afforded by article III. Interestingly, there has been a debate regarding the extent of coverage by the Standards Agreement of process standards as "technical regulations." The language of the definition of "technical regulations" includes process standards relating to the product, which appears somewhat ambiguous. The intent of at least some negotiators was to exclude from coverage process standards that are not somehow carried with the product.[195] The exclusion of such process standards from coverage under the Standards Agreement, as well as from protection under article III as described above, would have the result of relegating them to possible salvation under article XX, if at all.[196] Interestingly, the S&P Agreement specifically covers sanitary and phytosanitary measures that relate to processes and production methods, so long as they are applied to protect human, animal or plant life within the territory of the regulating state.[197]

(3) U.S.

In U.S. law, several of the tests described here are combined. Under commerce clause analysis, legislation that (i) discriminates de jure against out-of-state interests, or (ii) discriminates de facto in a way that is explicitly intentional (discriminatory purpose),[198] or (iii) has such a great disparity of effect that it indicates intentional discrimination[199] (described by the Court as having a "discriminatory effect") will be subjected to strict scrutiny, which it will rarely survive.[200] This strict scrutiny consists of a requirement that the state implementing the discriminatory measure justify it in LTRAT terms.[201] Even if these types of measures satisfy SMERT, they cannot be sustained "unless the discrimination is demonstrably justified by a valid factor unrelated to economic protectionism."[202] This proviso seems similar to the position under EU law: "comparable situations must not be treated differently and different situations must not be treated in the same way unless such treatment is objectively justified."[203] In fact, however, the Court has applied a "virtually per se rule of invalidity" to provisions that patently discriminate against interstate trade.[204]

Where it finds neither discriminatory purpose nor "discriminatory effect," the Court will characterize the burdens on interstate commerce as incidental or indirect, and apply a more relaxed and deferential proportionality test.[205] "A facially nondiscriminatory regulation supported by a legitimate state interest [SMERT--LE] which incidentally burdens interstate commerce is constitutional unless the burden on interstate trade is clearly excessive in relation to the local benefits [NP]."[206] Note that this differs from the "discriminatory effect" cases only in the way the Court characterizes the effect--in these cases, no discriminatory intent is suspected.

The Supreme Court is less concerned about externalization of regulatory costs alone, without economic protectionism, and will often uphold state statutes that may impose disproportionate costs on out-of-state interests but that are not intended to protect local markets.[207] However, where a regulatory goal, such as road safety, is sought to be achieved by diverting commerce from the regulating state--through isolationist techniques--the Court will invalidate the state regulation under the commerce clause.[208] "[E]conomic localism cannot be characterized as a symptom of breakdown in a local democratic process."[209] This perspective, implicitly discounting the possibility of local consumer interests interceding, is consistent with a standard public choice analysis of international trade politics.[210]

ii. Maximization of Net Gains of Trade and Regulation

As noted above, NTR rules decline to maximize the net gains from trade and regulation, in the sense that they fail to discipline local regulation that is non-discriminatory but (i) results in multiple regulation, (ii) refuses to accept foreign regulation as "equivalent" or (iii) simply is more costly in trade terms than it is valuable in regulatory terms.

iii. Administrability: Standards versus Rules

NTR seems like a rule, but is often operated as a standard. While NTR may seem on its face to be extremely easy to administer, this will often not be the case, as the "like products" issue is a proxy for a judicial examination of the rationality of regulatory categories. Often, NTR examinations shade into SMERT, NP, LTRAT, BT and CBA. The inability to confine NTR to a formally realizable rule may eviscerate the argument in favor of NTR on the ground that it avoids the administration costs of more intrusive tests.

Anti-discrimination rules intersect with SMERT insofar as the failure of a local rule to pass the SMERT test is often a minimal test for discrimination. Furthermore, as we have seen in the case of the dormant commerce clause, anti-discrimination rules may use failure of NP as a proxy for, or perhaps evidence of, latent discrimination, where there is no patent discrimination. Thus, both SMERT and NP may be used to identify discriminatory intent.[211] However, they are both over-broad and underinclusive as indicators of discrimination: that is, measures may fail these tests without being intentionally discriminatory, and may pass these tests when the intent that motivated their passage was to discriminate. In addition, arbitrary discrimination is "by definition, not the least restrictive way to achieve a legitimate public policy objective"[212] and therefore would violate LTRAT.

While rules of non-discrimination, such as national treatment, are different from NP, BT and CBA, they may under certain factual circumstances shade into a similar analysis. They may do so in considering the degree of distortion of competition, and in determining the "likeness" of products. Stated another way, at least under GATT/WTO law, different treatment does not violate requirements of national treatment where (i) it does not affect the conditions of competition for imported products adversely, or (ii) the relevant foreign and domestic products are not "like" and therefore provide no basis for determining discrimination. Similarly, in connection with discrimination in EU law[213] and under the U.S. commerce clause, we sometimes speak of "similarly situated" foreign persons or goods.[214] Here again, the determination of what is similarly situated may be outcome-determinative, and depends on the tribunal's degree of deference to the legislative categories under examination.

At least one GATT panel, the Wine and Beer Panel, has recognized the potentially extensive effects of the like products issue.

The Panel recognized that the treatment of imported and domestic products as like products under Article III may have significant implications for the scope of obligations under the General Agreement and for the regulatory autonomy of contracting parties with respect to their internal tax laws and regulations: once products are designated as like products, a regulatory product differentiation, e.g. for standardization or environmental purposes, becomes inconsistent with Article III even if the regulation is not "applied . . . so as to afford protection to domestic production".[215]

When a GATT or WTO panel finds that like products are treated differently, it invalidates regulation that may have been premised on the different characteristics of these products that are found by the panel to be insufficient to make them "un-like". The types of difference that will be permitted to constitute the basis for un-likeness constitute the range of permissible legislation. In the reformulated gasoline case,[216] the U.S. argued that the foreign character of Venezuelan and Brazilian manufacturers made them different for regulatory purposes, and the WTO panel and SAB rejected this categorization as illegitimate. The U.S. argument was that the regulatory category to which the foreign gasoline belonged was that for which the manufacturer could not accurately establish an individual historical baseline of chemical content.[217] The panel implicitly invalidated this regulatory category, finding that "chemically-identical imported and domestic gasoline are like products under Article III:4."[218] Here, the reason the U.S. felt it could not establish an individual baseline accurately was because of the very foreign character of the producers: the difficulty of auditing compliance and enforcing regulation as to foreigners.

The Wine and Beer Panel recognized that the power to supervise regulatory categorization is the power to supervise regulation, and suggested that "it is imperative that the like product determination in the context of Article III be made in such a way that it not unnecessarily infringe upon the regulatory authority and domestic policy options of contracting parties."[219] Where WTO dispute resolution panels review and reject domestic regulatory categories, they sit in judgment of the rationality of those categories, and of the regulation itself.[220]

This is how anti-discrimination may shade into proportionality (even in cases where there is no showing of discriminatory purpose).[221] In addition, this is how seemingly simple rules relating to discrimination are transmuted to standards forming the basis for Lochnerian judicial intervention. Thus, the purportedly neutral and formally realizable principle of non- discrimination is revealed to be value-laden and complex of administration.[222] And yet, a narrowly construed NTR alone would not provide sufficiently free trade in a zero transaction costs world (it is theoretically possible that in a positive transaction cost world, the inaccuracy of using only NTR might be acceptable in order to avoid the cost of more precise tests). In conclusion, it should be noted that definitions of NTR also shade into stipulations against economic protectionism and externalization.[223] While the impetus to do so is understandable, it should be noted that protectionism and externalization are not self-defining terms,[224] and that their use requires the expansion of NTR analysis to include a broad array of difficult factors that relate to, or include, SMERT, LTRAT, NP, BT and, ultimately, CBA.

iv. Distributive Concerns

NTR, by definition, allows externalization, so long as the regulatory standard chosen is not discriminatory. Where it accepts local regulation, NTR may allow detriments to be conferred on outsiders without compensation. Where the NTR test is applied centrally, or is applied locally on a reciprocal basis (reciprocal national treatment), broad or constitutional-type compensation may be seen to be made. In fact, none of the tradeoff devices considered in this article satisfy distributive concerns. However, one argument in favor of NTR is that it refers distributive issues to the political process-- the federal or EU legislative process or the international treaty process--accepting the advantages of political institutions over judicial institutions as instruments for revelation and exchange of preferences.

v. Moral Concerns

For those who attack CBA on moral grounds, one response is to ask which tradeoff device has greater moral legitimacy in a context in which not to decide is to decide. None of the tradeoff devices in use today seems to occupy a moral ground superior to CCBA. While NTR declines to make the kinds of explicit choices associated with CCBA, it chooses, and does so with less regard for all of the concerns at stake. Those who place non-trade values presumptively ahead of trade values might prefer NTR, as it displays a bias toward the non-trade values. However, the rest of us may seek a less biased instrument. Similarly, it might be argued that NTR has great moral legitimacy insofar as it simply protects non-discriminatory local regulation from attack. The moral legitimacy of NTR, in this sense, is derived from the moral legitimacy of the local legislative process. However, in order to assign moral superiority to NTR, we would have to assign moral superiority to the domestic legislative process, as opposed to the central legislative or adjudicative process. Furthermore, while there may be a social value in avoiding or mystifying tragic choices, in the interest of social cohesion, the "trade and . . . problem" context does not seem to be a valuable place to do so.

vi. Theoretical Concerns: Avoidance of Interpersonal Comparison of Utility

Again, NTR, if operated as advertised, avoids judicial interpersonal comparisons of utility, and refers the problem to the legislative realm, which may provide a more direct instrument for assertion of utility. Of course, for de jure discrimination that has no possible legitimate justification, no such comparisons are necessary, but such discrimination truly represents the "easy" case, and is declining in use. Where NTR seeks to address de facto discrimination, it often incorporates evaluation of the rationality of regulatory categories, SMERT, NP, LTRAT or even BT. SMERT does not seem to involve significant interpersonal comparisons of utility. However, NP and BT do, and LTRAT is often applied as LTRAT--RA, in which case it also involves significant interpersonal comparison of utility, albeit avoiding the most difficult evaluation of the domestic regulatory goal per se.

c. Least Trade-Restrictive Alternative Tests (LTRAT)

i. LTRAT Defined

(1) European Union

One component of the wider concept of proportionality in EU jurisprudence under arts. 30 and 36 is LTRAT. In the De Peijper case, the ECJ found a violation of article 30 where a Dutch law required that parallel importers of pharmaceutical products obtain documentation from the manufacturer or from the manufacturer's distributor. This requirement had the obvious collateral effect of chilling parallel importation. Even though there was a purported health justification for the requirement, the ECJ refused to allow the Netherlands to rely on the exception in article 36, because less trade restrictive measures were available to the Netherlands. Importantly, the Netherlands could have cooperated with the state of production of the imported pharmaceuticals. Thus, in this and other EU contexts, a LTRAT requirement gave rise to a duty of cooperation. We have seen similar duties arise under the LTRAT test applied pursuant to article XX of GATT.[225] In some cases the least trade restrictive measure is international regulatory cooperation.

(2) GATT/WTO: Necessity

Recall that the "necessity" qualifications contained in article XX(b) and (d) of GATT have been interpreted to require the national measure to be the least trade restrictive alternative reasonably available: LTRAT--RA--LE.[226] A fundamental question in connection with the necessity analysis is the scope of the "measure" under review: is it the entire regulatory scheme or only the trade-restricting component? In the Section 337 GATT dispute resolution report, the panel found that each inconsistency with GATT must be considered, rather than the necessity of the regulatory scheme as a whole.[227] Of course, if the necessity determination were required to be made only as to the regulatory scheme as a whole, it would be a much more deferential test. For example, in the recent panel decision relating to the U.S. Clean Air Act reformulated gasoline regulations, the panel examined only the discriminatory aspect of the regulations as the "measure" to be justified under article XX.[228] Thus, the question of the scope of the measure to be evaluated introduces a certain degree of outcome-determinative discretion, perhaps similar in its corrosive effect to the like products question raised in NTR analysis.

A second fundamental question arises from the inclusion of RA in the LTRAT test: what is reasonable?[229] If the reasonableness test amounts to a requirement that the least trade restrictive alternative not be so costly as to countervail the benefits of the regulatory measure, then it bears some resemblance to CBA; excluding from its truncated maximizing analysis only the measurement of benefits of the regulatory measure. If, alternatively, it amounts to a comparison that requires that the regulatory costs not be disproportionately great in comparison to the trade benefits, then it is a kind of NP. In the U.S. Reformulated Gasoline decision, the WTO dispute settlement panel and SAB declined to accept the U.S. position that auditing of refineries and other enforcement activities could not be effected in Venezuela and Brazil. Thus, "extraterritorial" enforcement measures (along the lines of the enforcement of the U.S. antidumping laws),[230] or as in the tuna cases, international cooperation, were viewed by these panels as reasonably available alternatives.[231]

The Standards Agreement adds a curious phrase to the LTRAT: it provides that "technical regulations shall not be more trade-restrictive than necessary to fulfill a legitimate objective, taking account of the risks non-fulfillment would create."[232] On its face, the italicized language appears non-sequacious: what part of LTRAT analysis would consider the risks of non-fulfillment of the regulatory goals? However, if the necessity test is thought of as more of a BT or CBA,[233] considering the potential costs of regulatory failure as part of its calculus, then this language may make sense. CBA would ordinarily discount a risk by its probability in order to calculate its "cost." In addition, if the necessity test under this provision is thought of as NP, the magnitude and probability of risk becomes relevant. On the other hand, the S&P Agreement contains what appears to be a non-sequacious LTRAT--RA formulation, requiring that sanitary and phytosanitary measures be "not more trade restrictive than required to achieve their appropriate level of protection, taking into account technical and economic feasibility."[234] The related footnote indicates that this standard disciplines two of the three components of regulatory cost and benefit. First, it asks whether there is a regulatory alternative that incurs significantly reduced trade costs. Second, it asks whether that regulatory alternative is reasonably available. It declines to discipline the extent to which the measure maintains its ability to meet the appropriate level of protection; that is, it does not require any reductions in protection, no matter how costly in trade terms.[235]

It should be noted here that article XX in general, and articles XX(b), (d) and (g) in particular, refer to specific regulatory goals: LTRAT--RA--LE. Thus, a dispute resolution panel applying these provisions must determine whether the regulatory goal sought to be achieved fits within these limited categories. On the other hand, the Standards Agreement refers only to "legitimate objectives."[236]

(3) U.S.

As noted above, where the trade restriction amounts to discrimination, either in purpose or in practical effect,[237] the U.S. Supreme Court has imposed LTRAT-RA. Discriminatory effect seems to be treated the same way as intentional discrimination. Finding that a municipality's regulation discriminated against interstate commerce, the Court held that it could not do so, "even in the exercise of its unquestioned power to protect the health and safety of its people, if reasonable alternatives, adequate to conserve legitimate local interests, are available."[238] Another formulation of LTRAT, with the same results, requires that discrimination be "demonstrably justified by a valid factor unrelated to economic protectionism."[239] It is interesting that the Court uses LTRAT as strict scrutiny, once discrimination is demonstrated. This limited application of LTRAT, combined with the Court's limited use of BT, indicates a central role for NTR in the U.S., the jurisdiction in our comparison that has the greatest legislative capacity.

ii. Maximization of Net Gains of Trade and Regulation

As a tool to maximize net gains of trade and regulation, LTRAT is overbroad and underinclusive. First, it seems to elevate trade values to a pre-eminent status,[240] by insisting that any non-trade measure be designed to meet its goal using the means that is least trade restrictive, no matter what the domestic regulatory cost. LTRAT--RA seems to address this problem by placing a cap on the domestic regulatory cost. Second, there may be circumstances where even the least trade restrictive alternative is not worthwhile: where the non-trade measure is worth far less than the trade costs it imposes. LTRAT engages in truncated maximization, or CCBA, by keeping the regulatory benefit relatively constant, and working on the trade detriment side. It thus evaluates a much more limited range of options, ignoring other groups of options that may be superior.

In the leading U.S. case of Pike v. Bruce Church, after prescribing a balancing or proportionality test that would consider whether the burden on interstate commerce "is clearly excessive in relation to the putative local benefits," Justice Stewart stated that "the extent of the burden that will be tolerated will of course depend on . . . whether [the local interest] could be promoted as well with a lesser impact on interstate activities."[241] Justice Stewart thus added to the balancing test a search for less trade restrictive alternatives.[242] This combination, and LTRAT generally, amounts to a start in the transition from static CBA toward CCBA, limiting the universe of comparison to measures that reduce the trade restriction costs without modifying the regulatory benefit. It only amounts to a start, as Stewart did not add a search for an alternative that would provide more regulatory benefits at a fixed trade cost, or at a trade cost that would be less than the increased regulatory benefits. Nor did he add a search for an alternative that would provide less regulatory benefits, but diminish trade detriments by an amount greater than the reduced regulatory benefits.

Perhaps it is worth noting here the categorical nature of LTRAT, and its relatively modest use in the U.S., where federal legislation has historically been easier to enact than perhaps EU legislation or amendments to GATT. Where local regulation can more readily be invalidated legislatively by legislative pre-emption, there is less need for a judicially-created and policed doctrine of LTRAT. National public policy goals can be achieved without judicial intervention. This categorical nature gives rise to the overbreadth and underinclusiveness referenced above.

Of course, the definition of the regulatory goal is critical to these tests. The more precisely defined the regulatory goal, and the more it refers both to maximizing regulatory benefits and minimizing regulatory costs, the more difficult it will be to find a less trade restrictive alternative that will achieve the same complex goal. However, implicit in LTRAT as applied is exclusion of consideration of domestic regulatory costs. This exclusion provides an incomplete picture of costs, giving weight to trade costs, but ignoring regulatory costs. This distortion may be reduced in the case of LTRAT--RA, by the implicit inclusion of regulatory costs in the consideration of what alternatives are indeed reasonably available.

Finally, LTRAT shades into recognition. That is, where the home country regulation is adequate to achieve the host country's regulatory goal, recognition is the least trade restrictive alternative. This relationship has been recognized since the Cassis de Dijon case.[243]

iii. Administrability: Standards versus Rules

LTRAT raises significant administrative problems, but these problems are simply a subset of those raised by CCBA. This is because LTRAT is itself a subset of CCBA, as noted above. In order to engage in LTRAT evaluation, it is necessary to calculate trade restriction effects for each of the regulatory measures evaluated.

iv. Distributive Concerns

LTRAT does not provide for any direct compensation to outsiders in cases where local legislation is upheld, nor does it provide direct compensation to locals where legislation is invalidated. On the other hand, it (i) minimizes the trade cost on outsiders (while accepting regulatory goals) and thus minimizes the distributive concern with respect to outsiders, and (ii) takes advantage of the same type of broad or indirect reciprocity that each tradeoff device studied in this paper entails.

v. Moral Concerns

LTRAT seems subject to some of the same moral critiques as CCBA, but avoids evaluation of the relative value of regulatory goals, and thus avoids the strongest moral critique.

vi. Theoretical Concerns: Avoidance of Interpersonal Comparison of Utility

LTRAT, by providing a truncated analysis that does not evaluate local regulatory goals, seems to limit interpersonal comparison of utilities. While its assessment of trade restriction involves some interpersonal comparison of utilities, it does so only within the utilities of the complaining state, and does not consider the utilities of the allegedly offending state. Finally, its comparison of trade restriction involves calculations that are frequently required in trade relations, and thus perhaps are more susceptible to economic analysis.

d. Narrow Proportionality (NP)

i. NP Defined

Narrow proportionality may be viewed as a special type of BT or CBA. It is different from a full BT or CBA, insofar as it generally provides a margin of deference to the local regulation, asking whether it is disproportionate, or excessive in its costs considering its benefits, as opposed to whether its detriments simply outweigh its benefits. Second, it is different as it does not purport to weigh specific factors, or to engage in CBA. Rather, like BT, it involves more of a gestalt. On the other hand, the definition and method of application of each test is critical. For example, Sunstein has defined proportionality as requiring that aggregate social benefits are proportionate to the aggregate social costs.[244] This definition merges the analytical tasks of proportionality and CBA, while providing a somewhat different mathematical formulation: proportionate costs and benefits instead of benefits in excess of costs.

(1) European Union

Proportionality is a general principle of EU law,[245] and was incorporated in the Treaty of Rome by the Maastricht Treaty.[246] The locus classicus of the principle of proportionality in EU law is the Internationale Handelsgesellschaft case, which articulated proportionality as follows: "the individual should not have his freedom of action limited beyond the degree necessary for the public interest."[247] The affinity between this formulation of proportionality and the formulation of subsidiarity in Catholic social doctrine is apparent.[248] Proportionality was critical to the decision in the well-known Danish bottles case, where the ECJ found the trade detriments arising from one component of the Danish regulatory scheme disproportionately large in relation to the regulatory benefits.[249]

The doctrine of proportionality as elaborated at Maastricht builds upon this jurisprudence. Article 3b, the provision of the Treaty of Rome added at Maastricht to incorporate the principle of subsidiarity, also states the principle of proportionality: "Any action by the Community shall not go beyond what is necessary to achieve the objectives of this Treaty."[250] This has been articulated further by the European Council: "Any burdens, whether financial or administrative, falling upon the Community, national governments, local authorities, economic operators and citizens, should be minimized and should be proportionate to the objective to be achieved."[251] Of course, these two statements simply incorporate different weasel words--"necessary" and "proportionate"--without defining precisely how compliance with these requirements is to be determined.

In EU law, the wider concept of proportionality includes (i) LTRAT, (ii) SMERT, and (iii) NP.[252]

(2) GATT/WTO

GATT/WTO jurisprudence has not explicitly embraced NP, except to the extent, as discussed below, that LTRAT may be viewed as shading into NP.

(3) U.S.

In cases of differential treatment that the U.S. courts do not characterize as having "discriminatory effect"--where there is no sense of intentional discrimination, but only incidental or indirect burdens on interstate commerce--the local legislation is upheld unless the burden on interstate commerce is clearly excessive in relation to the local regulatory benefits.[253] NP is thus reserved for cases of non-discrimination, and generally has the result of validating the local measure.

ii. Maximization of Net Gains of Trade and Regulation

The Supreme Court has combined NP with LTRAT,[254] but LTRAT has a complex relationship with NP. Where the difference between the measure under review and another available measure that is less trade restrictive is great in terms of the amount of trade restriction, it might be argued that NP as formulated above would have the same invalidating result as LTRAT. This is because the burden on interstate commerce could be characterized as "clearly excessive in relation to the putative local benefits" by reference to the less restrictive alternative. This is a kind of implicitly comparative NP, with the same relationship to non-comparative NP as that between CBA and CCBA. Comparative NP includes LTRAT, with a margin. Due to this margin, it is possible that a measure would fail LTRAT, but not be clearly excessive under NP, if it were only marginally more trade restrictive than another alternative. It is also possible that a measure that is not the least trade restrictive alternative would satisfy a non-comparative NP test.

NP may be viewed as a truncated CBA (but not necessarily CCBA) test. NP does all the same work as CBA, but provides greater deference to local regulation. In this sense, it is by design less capable of maximization of gains from trade and regulation. Or perhaps one might argue that given the margin for error involved with CBA, NP's presumption in favor of local regulation simply applies this margin for error, taking advantage of the gains from conservatism. NP may become more like CCBA the more that it evaluates various alternatives as part of its determination of whether the particular measure under scrutiny is proportional.

iii. Administrability: Standards versus Rules

As it requires the same data as CBA, NP would seem no more administrable than CBA. However, its margin of deference may allow a looser analysis, yielding greater predictability and ease of administration.

iv. Distributive Concerns

NP considers the interests of local interests alongside the interests of outsiders, while weighting the analysis in favor of the local interests. Like the other tradeoff devices, NP does not provide compensation, but may be viewed as incorporating a kind of broad reciprocity, similar to other tradeoff devices, wherein the losing party may expect recompense by virtue of the possibility of being the winning party in another similar case.

v. Moral Concerns

NP cannot claim significantly greater moral legitimacy than CBA or CCBA, although it accords greater deference to local regulation, which might be viewed by some as commanding greater moral legitimacy than negative regulation. On this view, the margin of deference afforded by NP may be viewed as a cushion of protection for this higher morality. However, this view assumes, with little explicit basis, the greater morality of local decision-making.

vi. Theoretical Legitimacy: Avoidance of Interpersonal Comparison of Utility

NP suffers from the same problems of interpersonal comparison of utility as CBA and CCBA, although again, the margin of deference might be viewed by some as reducing this problem.

e. Balancing Tests (BT)

i. Defined

As noted above, BT and CBA have much in common with one another, and with NP: each compares regulatory benefits with trade detriments. The major difference between NP on the one hand, and BT and CBA on the other hand, is that NP seems to specify a particular margin of deference for local regulation: the trade detriments that the local regulation entails cannot be disproportionately large.[255] This is why, in the EU context and in the U.S. context, terms like proportionality and balancing tests, and even cost-benefit analysis, are often used interchangeably.

(1) European Union

The European Union's jurisprudence relating to free movement of goods does not adopt BT or CBA per se. While proportionality in the EU sense seems to require some weighing, it is not full BT or CBA.[256] In the Stoke-on-Trent (Sunday Trading III) case, the ECJ stated that "[a]ppraising the proportionality of national rules which pursue a legitimate aim under Community law involves weighing the national interest in attaining that aim against the Community interest in ensuring the free movement of goods."[257] However, a U.K. court has expressed the following reservations:

It is readily assumed that the exercise required by the Cassis De Dijon exception in a case such as the present would involve a kind of cost-benefit analysis. Weights would be attributed to the interests respectively of free movement and the socio-cultural object of the particular measure, and the court would then decide whether the latter outweighed the former. Something of this kind is often involved in the legislative process, where political premises lead to a decision that one desirable aim must be subordinated to another. But to perform this task in a judicial context would in all but the most obvious case be a difficult matter.[258]

(2) GATT/WTO

GATT/WTO jurisprudence has not explicitly embraced BT or CBA, and has implicitly done so only in a very limited sense, as discussed above, by virtue of the like products analysis. Given the similarity between the language of article 30 of the Treaty of Rome and the language of article XI of GATT, and parallels between article 36 of the Treaty of Rome and article XX of GATT, it may be worth considering why no NP, BT or CBA have developed under GATT. In the Herring and Salmon case under the superseded US-Canada Free Trade Agreement, the panel suggested a balancing test under article XX(g) of GATT.[259] This related to the question, under article XX(g), of whether the measure was "primarily aimed at" conservation. This question, while unique to article XX(g), may be viewed as somewhat analogous to the necessity test in article XX(b) and (d). The Panel decided that the key to this question was whether the Canadian government would have adopted the measure for conservation reasons alone, and that in turn, this could be determined on the basis of a CBA relating trade detriment to domestic regulatory benefit.[260]

(3) U.S.

"While it has become standard practice at least since Pike v. Bruce Church . . . to consider, in addition to [issues of discrimination and inconsistent regulation], whether the burden on commerce imposed by a state statute 'is clearly excessive in relation to the putative local benefits,' . . . such an inquiry is ill suited to the judicial function and should be undertaken rarely if at all."[261] "[I]n no area has the Court engaged in the sort of open-ended balancing the scholars have recommended."[262] "Contrary to some suggestions, courts sometimes must attempt to 'weigh' non-comparables."[263] This debate continues.

As noted above, in the U.S. commerce clause context, BT amounts to lax scrutiny, and often validation, of the local measure. "[T]his lesser scrutiny is only available where other legislative objectives are credibly advanced and there is no patent discrimination against interstate trade."[264] The dichotomy established by the Court, between patent discrimination (including both de jure and certain de facto discrimination) and legislation with only incidental effects on interstate commerce, emphasizes the intent, whether patent or latent. In this sense, "incidental" means unintended. Where the adverse effects on interstate commerce are incidental, the Court will engage in BT,[265] but will almost always uphold the local legislation, avoiding difficult questions and avoiding challenges to the Court's legitimacy. In the Carbone case, Justice Souter, joined by Justices Blackmun and Rehnquist, argues that the purported BT that has become a part of the Court's jurisprudence is "not so much an open-ended weighing of an ordinance's pros and cons, as an assessment of whether an ordinance discriminates in practice or otherwise unjustifiably operates to isolate a State's economy from the national common market."[266]

The ECJ "seems to place a much heavier thumb on the [EU] side of the scale than the Supreme Court places on the federal side."[267] Indeed, the ECJ has seemed more willing to invalidate member state legislation based on failure of NP or LTRAT tests than the Supreme Court has seemed based on its BT.

ii. Maximization of Net Gains of Trade and Regulation

BT can be viewed as the CBA of resignation. It avoids the apparent arrogance of seeking, perhaps with false precision, to evaluate and commensurate various factors, but simply looks at them and decides. In this regard, like NP, it intentionally avoids the claims to precision of CBA or CCBA. Because BT, like CBA, is not explicitly comparative (or dynamic) in theory, and usually is not comparative in practice, it provides less scope for maximization of net gains of trade and regulation than CCBA.

One important distinction between LTRAT and NP, on the one hand, and BT and CBA, on the other hand, is the willingness of LTRAT and NP to engage in comparative evaluation: to consider at least one type of hypothetical alternative. That is, LTRAT and the comparative form of NP are willing to examine whether there is a less costly alternative in trade terms. Under this rubric, for example, they would consider whether international action--integration--to achieve the relevant local goal is a less trade restrictive alternative.[268] On the other hand, BT and non- comparative CBA balance actual factors and costs, without imagining the possibility of more efficient alternatives through integration. Of course, such speculation raises important questions of legitimacy, and when GATT or WTO panels have suggested that international cooperation was a less trade restrictive alternative under article XX, for example in the Tuna decisions,[269] environmentalists and others have criticized this apparent usurpation of the power of national governments to decide the means to their chosen ends.

iii. Administrability: Standards versus Rules

BT is perhaps the least predictable of our tradeoff devices, as it imposes the fewest constraints on decisionmaking, although one constraint it does impose is that all factors must be considered. BT is also relatively costly to administer, although because of its relative modesty of precision, it may be more administrable than CBA or CCBA.

iv. Distributive Concerns

BT raises the same distributive concerns as CBA and CCBA.

v. Moral Concerns

Interestingly, BT may have greater claims to moral legitimacy than CBA. BT does not seek to commodify values, and unlike some forms of CBA, has room for consideration of non-monetizable and non-commensurable values. Rather, BT may have more in common with deliberative decision-making, providing room for all considerations to be included in the decision.

vi. Theoretical Legitimacy: Avoidance of Interpersonal Comparison of Utility

BT raises concerns regarding interpersonal comparison of utility similar to those raised by CBA, although the concerns may be reduced because BT explicitly avoids claims of precise comparison.

f. The Move From CCBA

We began with the proposition that CCBA unqualifiedly maximizes the net sum of gains from trade and gains from regulation. However, at least as a tradeoff device, CCBA experiences real problems of administrability, and raises distributive, moral and theoretical concerns. This article has attempted to compare other tradeoff devices with CCBA in these terms, with a view toward suggesting that CCBA is not used because of these problems, which may be reduced, at a cost in terms of maximization of net gains, by use of other tradeoff devices. If these tradeoff devices are chosen with social welfare in mind, we may presume that the cost in terms of maximization is less than the gains in terms of addressing distributive, moral and theoretical concerns. However, it may be that our capacity to optimize social welfare is limited by institutional constraints. It may be that we as a global society have not yet developed institutional solutions that facilitate greater use of CCBA--that we have not extended the Pareto frontier by institutional innovation. With greater evaluation and institutional innovation, it may be that CCBA will play a greater role in international decision-making.

For many commentators and judges, simple NTR is the appropriate fall-back position. However, this article has shown that NTR, in the absence of explicit discrimination or evidence of intentional unjustified discrimination, may suffer from some of the same problems as CCBA. Furthermore, there seem to be significant instances where discipline is worthwhile in the absence of explicit discrimination or evidence of intentional discrimination. SMERT adds little to the depth of scrutiny provided by NTR. NP is quite similar to CBA, with a greater margin of deference, and consequently provides some of the same benefits and is susceptible to some of the same problems as CBA. LTRAT seems overbroad and underinclusive, but nevertheless is frequently used. LTRAT--RA operates on two parameters: trade cost and regulatory cost. However, it declines to include regulatory benefit in its analysis. It accepts a degree of inaccuracy in exchange for the benefits of avoiding the greatest moral and theoretical concerns that may come of evaluating and comparing the benefits of domestic regulation. BT, as a fuzzy CBA or CCBA, seems to have some real benefits, including the avoidance of attempts to commensurate in mathematical terms. Where BT includes considerations of whether a less restrictive alternative exists, it incorporates some of the benefits of LTRAT--RA.

All of the tradeoff devices considered here have distributive problems: they all make binary decisions that may leave costs on outsiders. This can be rationalized on a broad basis: in a community, sometimes I lose and sometimes you lose. On the other hand, this problem of distribution--of compensation to losers--may be reduced by referring the decision to the political process. In this connection, the political process--the legislature in particular--may be viewed as a specialized institution for the transfer of value, especially under conditions of incommensurability. Courts may transfer more difficult problems to the legislature simply by declining to settle them, or by settling them in an unsatisfactory way (particularly where not to decide is to decide). In this way, use of a device such as NTR that declines to discipline a range of local legislation that seems inefficient may be justified as a referral of the linked tasks of interpersonal comparison of utility and distribution to the legislature. It is in this sense only that those who argue for NTR as the main tradeoff device can be correct. Of course, this argument for NTR is persuasive only where the legislature is available to act; this is not the case in the early history of the European Community, and in the present days of the WTO.

However, LTRAT--RA seems to provide some of the same benefits as NTR, with greater ability to maximize net gains from trade and regulation, while avoiding the greatest problems of administrability, moral concerns and theoretical concerns. Morever, LTRAT--RA accepts more responsibility for the adjudicator than does NTR, standing ready to fill gaps in legislative capability. This fact may support the use of LTRAT--RA in the GATT/WTO context.

One might then argue that NTR is appropriate for circumstances where a well-developed legislative capacity exists. And indeed, in the U.S., the basic rule seems to be NTR, with other tests serving as mere proxies for or adjuncts to NTR. Conversely, a basic rule of LTRAT--RA seems more appropriate where central legislative capacity is more limited.

Table B summarizes the comparison described above, while Table C analyzes the factors addressed by each of the tradeoff devices by way of maximization.

Table B: Summary of Analysis


Maximization

Administr-

ability

Moral Concerns

Theoretical Concerns

NTR

No direct maximization

Good in theory, but flows into other tests

Often refers difficult decisions to legislature

Not raised until flows into other tests

LTRAT--RA

Maximizes at only trade cost and cost of regulation parameters

Problem

Reduced due to abstention from assessment of regulatory benefits

Reduced due to abstention from assessment of regulatory benefits

NP

BT with a margin of deference

Larger problem

Significant as seeks to commensurate

Significant as seeks to engage in interpersonal comparison

BT/CBA

Non-dynamic; therefore limited maximization

Larger problem

Significant as seeks to commensurate

Significant as seeks to engage in interpersonal comparison

CCBA

Full maximization

Largest problem

Significant as seeks to commensurate

Significant as seeks to engage in interpersonal comparison

Table C: Analysis of Maximization


Discipline

patent discrimination

Trade restriction

--static (S) or comparative (C)

Regulatory cost

--static (S) or comparative (C)

Regulatory benefit

--static (S) or comparative (C)

Relate trade restriction to regulatory cost

Margin of appreciation

All factors--

static (S) or comparative (C)

NTR

X







LTRAT

X

C




X


LTRAT --RA

X

C

C


X

X


NP


S

S

S

X

X

S

BT/

CBA

S

S

S

X


S

CNP

X

C

C

C

X

X

C

CCBA

X

C

C

C

X


C


[142] It is worth noting, however, that the Uruguay Round agreements, including GATT 1994, the General Agreement on Trade and Services, the Standards Agreement, and the S&P Agreement, specifically include language that is designed to incorporate the prior GATT jurisprudence of "necessity." The Dunkel Draft of the Standards Agreement would have gone further in suggesting proportionality review, as it contained the following footnote after section 2.2: "This provision is intended to ensure proportionality between regulations and the risks non-fulfillment of legitimate objectives would create." 3 THE GATT URUGUAY ROUND: A NEGOTIATING HISTORY (1986-1992) 527 (Terence P. Stewart ed., 1993). Of course, the meaning of this proposed footnote, as well as of its exclusion, is open to interpretation. In referring to proportionality, it could be referring to NP, WP or to LTRAT. By deleting this footnote in the final text, the parties may have taken the view that it was unnecessary to accomplish what they wanted, or they may have decided that they meant none of these things. Barcelo suggests that the inclusion of the footnote would have authorized CBA. Barcelo, supra note 9, at n.88, citing Ernst-Ulrich Petersmann, International Competition Rules for the GATT-MTO World Trade and Legal System, 27 J. WORLD TRADE 35, 45 (1993). Barcelo further takes the view that the parties intended to avoid a move to BT or CBA. Id., citing a telephone conversation with a representative of the Office of the United States Trade Representative confirming this interpretation. While this may well have been the U.S. view of the intent, it is open to a WTO panel to find a different interpretation.

[143] This term refers to the European Court of Justice ("ECJ") and the U.S. Supreme Court, as well as to GATT dispute resolution panels.

[144] See Farber & Hudec, supra note 9, at 1408-1409, and sources cited therein.

[145] Treaty Establishing the European Community, Feb. 7, 1992 [1992] 1 C.M.L.R. 573, 602 [hereinafter "Treaty of Rome"], incorporating changes made by Treaty on European Union, Feb. 7, 1992, O.J. C 224/1 (1992), [1992] 1 C.M.L.R. 719, 31 I.L.M. 247 [hereinafter "TEU"]. The TEU amended the Treaty Establishing the European Economic Community, Mar. 25, 1957, 298 U.N.T.S. 11, 1973 Gr. Brit. T.S. No. 1 (Cmd. 5179-II), as amended by the Single European Act, O.J. L 169/1 (1987), [1987] 2 C.M.L.R. 741, in Treaties Establishing The European Community (EC Off'l Pub. Off. 1987).

[146] OLIVER WENDELL HOLMES, COLLECTED LEGAL PAPERS 295-96 (1920).

[147] See Joseph Weiler, The Transformation of Europe, 100 YALE L.J. 2403 (1991). See also Case 804/79, Commission v. United Kingdom (Fisheries), [1981] E.C.R. 1045, [1982] C.M.L.R. 543 (a member state could not take measures on its own in an area committed to EU jurisdiction, even though EU legislation had not been effected).

[148] Lenaerts, supra note 109, at 256.

[149] See Council Directive 70/50 on the Abolition of Measures Which Have an Effect Equivalent to Quantitative Restrictions on Imports, 1970 O.J.SPEC. ED. (I) 17. See also Eric L. White, In Search of the Limits to Article 30 of the EEC Treaty, 26 COMM. MKT. L. REV. 235 (1989).

[150] Case 8/74, Procureur du Roi v. Dassonville, 1974 E.C.R. 837, 852, [1974] 2 C.M.L.R. 436 ("[a]ll trading rules enacted by Member States which are capable of hindering, directly or indirectly, actually or potentially, intra-Community trade are to be considered as measures having an effect equivalent to quantitative restrictions"). See also Case 788/79, Gilli & Andres [1980] E.C.R. 2071 [1981] 1 C.M.L.R. 146. But see Keck and Mithouard, Cases C-267 and C-268/91, [1993] 1 E.C.R. 6097, [1995] 1 C.M.L.R. 101 (art. 30 does not apply to simple selling restrictions, as opposed to product or presentation restrictions). On the Keck judgment, see Damian Chalmers, Repackaging the Internal Market --The Ramifications of the Keck Judgment, 19 EUR. L. REV. 385 (1994); Norbert Reich, The "November Revolution" of the European Court of Justice: Keck, Meng and Audi Revisited, 31 COMM. MKT. L. REV. 459 (1994).

[151] See PETER OLIVER, FREE MOVEMENT OF GOODS IN THE EEC 168 (2d ed. 1988). The varying interpretations of this language in the EU and in the GATT presents an interesting case study in the different practical legal effect of similar texts, perhaps with an explanation based at least in part on institutional differences.

[152] See, Case 95/81, Security for Import Advances, [1982] E.C.R. 2187, 2204, : "It must be recalled that in accordance with the settled case law of the Court, Article 36 must be strictly interpreted and the exceptions which it lists may not be extended to cases other than those which have been exhaustively laid down and, furthermore, that Article 36 refers to matters of non-economic nature."

[153] Art. 36, like art. XX of GATT, does not reserve powers to the member states, so much as establish what would be a plausible justification for an exception to the principle of unimpeded trade, and the bases for its use. See Case 5/77, Simmenthal, 1977 E.C.R. 1555, 1557.

[154] See, e.g., Case 302/86, Commission v. Denmark, [1988] E.C.R. 4607, [1989]1 C.M.L.R. 619 (environmental protection is one of the mandatory requirements protected under art. 30); Jan Steiner, Drawing the Line: Uses and Abuses of Article 30 EEC, 29 COMM. MKT. L. REV. 749, 752-53 (1992).

[155] Case 120/78, Rewe-Zentral AG v Bundesmonopolverwaltung fur Branntwein, [1979] E.C.R. 649, [1979] 3 C.M.L.R. 494. See also note 37, supra.

[156] I.e., non-discriminatory. See Commission Directive of December 22, 1969, O.J. L 13/31 (January 19, 1970). See also White, supra note 148, at 241.

[157] Distinctly applicable measures would violate art. 30 of the Treaty of Rome unless they fall within one of the exceptions provided explicitly under art. 36. Case 113/80 Commission v. Ireland, [1981] E.C.R. 1625, [1982] 1 C.M.L.R. 706; Cases C-1/90 & C-176/90 Aragonesa de Publicidad Exterior S.A. v. Departamento de Sanidad y Seguridad Social de Law Generalitat de Catalu_a (APESA) [1991] 1 E.C.R. 4151, [1994] C.M.L.R. 887.

[158] See, e.g., JOSEPHINE STEINER, TEXTBOOK ON EC LAW 86-87 (1992).

[159] Chalmers, supra note 149 at 387.

[160] Id. at 402 (citations omitted).

[161] But see Standards Agreement, supra note 6.

[162] See Barcelo, supra note 9, at 759-760 (listing and explaining panel decisions supporting this proposition). This interpretation of art. III as exclusive is based in part on language in the Note Ad Art. III to the effect that border measures intended to enforce domestic regulation are to be "subject to art. III." The implication drawn by some is that such border measures are exclusively subject to art. III. In Canada--Administration of the Foreign Investment Review Act (FIRA), 30 B.I.S.D. 140, 162, para. 5.14 (1984), the panel stated as follows:

The Panel shares the view of Canada that the General Agreement distinguishes between measures affecting the "importation" of products, which are regulated in Article XI:1, and those affecting "imported products", which are dealt with in Article III. If Article XI:1 were interpreted broadly to cover also internal requirements, Article III would be partly superfluous. Moreover, the exceptions to Article XI:1, in particular those contained in Article XI:2, would also apply to internal requirements restricting imports, which would be contrary to the basic aim of Article III.

[163] United States--Restrictions on Imports of Tuna, 39 B.I.S.D. 155 (1993), reprinted in 30 I.L.M. 1594 (1991) [hereinafter "First Tuna Panel Report"]; United States--Restrictions on Imports of Tuna, DS29/R (1994), reprinted in 33 I.L.M. 839 (1994) [hereinafter "Second Tuna Panel Report"]. Neither panel decision was adopted, and so neither has formal legal effect.

[164] In the First Tuna Panel Report, id., the Panel held that art. III only applies to regulations applicable to the product as such, not to regulation that relates to the process by which the product is produced, such as the restrictions on fishing technique applicable under the U.S. Marine Mammals Protection Act. See text accompanying notes 193 to 195, infra.

[165] United States--Standards for Reformulated and Conventional Gasoline, WT/DS2/R, January 29, 1996 (panel report) [hereinafter "Reformulated Gasoline Panel Report"], reprinted in 35 I.L.M. 274 (1996), modified by the SAB report, WT/DS2/AB/R, April 29, 1996, reprinted in 35 I.L.M. 603 (1996) [hereinafter "Reformulated Gasoline SAB Report"].

[166] Standards Agreement, supra note 6, art. 2.2.

[167] The correct term of art under the Standards Agreement for standards that are backed by law is "technical regulations."

[168] See DANIEL C. ESTY, GREENING THE GATT: TRADE, ENVIRONMENT, AND THE FUTURE 222 (1994); Barcelo, supra note 9, at 769-771, not making this argument himself, but citing Richard Stewart, The NAFTA: Trade, Competition, Environmental Protection, 27 INT'L LAW. 751, 760 n.29 (1993); Runge, supra note 48, at 18 n.12.

[169] S&P Agreement, supra note 6, preamble.

[170] Baldwin v. G.A.F. Seelig, Inc., 249 U.S. 511, 522 (1935) (Cardozo, J., quoting Judge Hand in the opinion below).

[171] U.S. CONST. art. I, sec. 8, cl. 3.

[172] For a recent case evaluating the limits of the commerce clause, see United States v. Lopez, 115 S.Ct. 1624 (1995).

[173] See Tribe, supra note 40, at 403.

[174] In addition, under the supremacy clause, a valid act of Congress (pursuant to Congress' powers under the commerce clause or otherwise) supersedes conflicting state law. U.S. CONST. art. VI, cl. 2.

[175] CTS Corp. v. Dynamics Corp. of America, 481 U.S. 69, 87 (1986). Initial commerce clause cases include Gibbons v. Ogden, 9 Wheat. 1, 227 (1824) and Cooley v. Board of Wardens, 53 U.S. (12 How.) 299 (1851). In Gibbons v. Ogden, Justice Johnson's concurring opinion suggested the exclusive Congressional power over interstate commerce. In Cooley v. Board of Wardens, the Court accepted the exclusive Congressional power.

[176] See Cooley v. Board of Wardens, 53 U.S. (12 How.) 299, 319 (1852): "[w]hatever subjects of this power are in their nature national, or admit only of one uniform system, or plan of regulation, may justly be said to be of such a nature as to require exclusive legislation by Congress." Given the broad scope of the positive commerce clause, see Wickard v. Filburn, 317 U.S. 111 (1942), it has been difficult to determine when states are excluded from action and when they may act. See Tyler Pipe Industries v. Washington State Dep't of Revenue, 483 U.S. 232, 260 (1987) (Scalia, J., concurring in part and dissenting in part) ("applications of the doctrine have, not to put too fine a point on the matter, made no sense").

[177] 476 U.S. 573, 578 (1986).

[178] Id. at 578-79 (citations omitted).

[179] Tribe, supra note 40 at 439. See also Sykes, supra note 104, at 104: "As shall be seen, however, these 'principles' are, in fact, exceptionally vague and do not generate an entirely coherent body of decisions." Farber and Hudec concur with this perspective, explaining the variation in terms of an impressionistic initial determination of whether the intent of the local measure is discriminatory or not. Farber & Hudec, supra note 9, at 1415, 1426. See also Regan, supra note 74 (arguing that intent should be the touchstone of decisions under the dormant commerce clause). In fact, local measures that are viewed as non-discriminatory are subjected to the type of analysis described by Tribe, and generally upheld. Local measures that are viewed as discriminatory are subjected to LTRAT, and rarely survive.

[180] See, e.g., CTS Corp. v. Dynamics Corp., 481 U.S. 69, 95-96 (1987) (Scalia, J., concurring); Regan, supra note 74, at 1165; Giuliano Marenco, Pour une interprétation traditionnelle de la notion de mesure d'effet equivalent á une restriction quantitative, CAH. DR. EUR. 291-364 (1984); Giuliano Marenco & Karen Banks, Intellectual Property and the Community Rules on Free Movement: Discrimination Unearthed, 15 EUR. L. REV. 224, 238-41 (1990).

[181] This article focuses on goods, as opposed to services, persons or other factors. Much of the analysis regarding goods is applicable to other factors.

[182]

The central rationale for the rule against discrimination is to prohibit state or municipal laws whose object is local economic protectionism, laws that would excite those jealousies and retaliatory measures the Constitution was designed to prevent.

C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1682 (1994). See also South Carolina Highway Department v. Barnwell Bros., 303 U.S. 177, 185 n.2 (1938).

[183] See note 97, supra.

[184] See White, supra note 148, at 241.

[185] Case 13/63, Italian Republic v. Commission, [1963] E.C.R. 165, 178, [1963] C.M.L.R. 289. See also Case 106/83, Sermide SpA v. Cassa Conguaglio Zucchero et al, [1984] E.C.R. 4209.

[186] For an example of validation of local law through "creative" use of regulatory categories, see the well-known Walloon Waste case, Case C-2/90, [1992] E.C.R. I4431, [1993] 1 C.M.L.R. 365 (Court finds that distinction between local waste and foreign waste is not discrimination, in part due to the policy of "proximity:" treating waste near its origin).

[187] See note 161, supra.

[188] Standards Code, supra note 6, art. 2.2.

[189] Report of the Appellate Body, United States--Standards for Reformulated and Conventional Gasoline, WTO Document WT/DS2/AB/R, April 29, 1996, at 28-29.

[190] It is worth noting here that art. XX(g) does not contain a "necessity" qualifier, and that the SAB rejected the prior panel report's finding that an exception under art. XX(g) was unavailable. The SAB criticized the panel report for finding that the very discrimination that violated art. III also made it impossible for the U.S. measure to be found "primarily aimed at" the conservation of exhaustible natural resources. Under panel decisions, art. XX(g) has been interpreted to require that measures be primarily aimed at such conservation. Canada--Measures Affecting Exports of Unprocessed Herring and Salmon, 35 B.I.S.D. 98, 114, para. 4.6, adopted March 22, 1988; Second Tuna Panel Report, supra note 39; United States--Taxes on Automobiles, DS 31/R (1994). The latter two panel decisions were not adopted by the GATT Council.

[191] As in, for example, the case of United States--Section 337 of the Tariff Act of 1930, 36 B.I.S.D. 345, 392, paras. 5.25-5.27 (1990) [hereinafter "Section 337 Panel Report"].

[192] Section 337 Panel, id., at 386, para. 5.11; see also United States Measures Affecting Alcoholic and Malt Beverages, 39 B.I.S.D. 279, para. 5.30 (1992) [hereinafter "Wine and Beer Panel Report"].

[193] See Gerald C. Berg, An Economic Interpretation of "Like-Product," 30 J. WORLD TRADE 195 (1996).

[194] First Tuna Panel and Second Tuna Panel, supra note 162.

[195] See Negotiating History of the Coverage of the Agreement on Technical Barriers to Trade with Regard to Labelling Requirements, Voluntary Standards, and Processes and Production Methods Unrelated to Product Characteristics, WTO Document WT/CTE/W/10 G/TBT/W/11, August 29, 1995, ¶¶ 146-148.

[196] It may be worthwhile here to point out that the product-process distinction, as thus applied under article III, is a regulatory jurisdiction allocating distinction. See Joel P. Trachtman, United States Restrictions on Imports of Tuna, 86 AM. J. INT'L L. 142 (1992). That is, as applied under article III, product regulation is subject to the relatively lax scrutiny of NTR, whereas process regulation must meet the strict scrutiny of LTRAT, under article XX. This method of distinction, and allocation of regulatory jurisdiction, is no more natural than any other formalistic categorization. It is thus subject to the same kind of evaluation applied here to other methods of allocation of regulatory jurisdiction.

[197] S&P Agreement, supra note 6, Annex A, para. 1.

[198] The intent is that of the legislators. See Smith, supra note 21, at 1241-43, citing Hunt v. Washington Apple Advertising Comm'n, 432 U.S. 333, 352 (1977); Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456, 463 n.7, 471 n.15 (1981).

[199] See Regan, supra note 74.

[200] In Justice O'Connor's words, "Where, however, a regulation 'affirmatively' or 'clearly' discriminates against interstate commerce on its face or in practical effect, it violates the Constitution unless the discrimination is demonstrably justified by a valid factor unrelated to protectionism." C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1688 (O'Connor, J., concurring) (1994). In Maine v. Taylor, 477 U.S. 131 (1986), the Supreme Court upheld a discriminatory statute, where Maine satisfied its burden of proof that the law prohibiting the import of live baitfish was necessary, in the sense that the legitimate local purpose of protecting local fisheries from disease could not be served by less discriminatory means, such as inspection. Strict scrutiny, in this field, applies a LTRAT, and almost invariably invalidates the state measure.

[201] See, e.g., Hunt v. Washington State Apple Advertising Commission 432 U.S. 333 (1977). See also Farber & Hudec, supra note 9, at 1413.

[202] New Energy Co. of Indiana v. Limbach, 486 U.S. 269, 274 (1988); Philadelphia v. New Jersey, 437 U.S. 617, 627 (1978) ("In each of these cases, a presumably legitimate goal was sought to be achieved by the illegitimate means of isolating the State from the national economy").

[203] Case 106/83, Sermide SpA v. Cassa Conguaglio Zucchero et al., 184 E.C.R. 4209, 4231.

[204] Associated Ind. of Missouri v. Lohman, 114 S.Ct. 1815 (1994), quoting Philadelphia v. New Jersey, 437 U.S. 617, 624 (1978). See also C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677 (1994).

[205] Id. See Pike v. Bruce Church, Inc., 397 U.S. 137 (1970).

[206] C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1688 (O'Connor, J., concurring) (1994), citing Brown-Forman Distillers Corp. v. New York State Liquor Auth., 476 U.S. 573 (1986); Pike v. Bruce Church, Inc., 397 U.S. 137 (1970). Note, however, that in dissent, Justice Souter, joined by Blackmun and Rehnquist, argue that at least in the context of processing services, all the laws invalidated under the commerce clause were "patently discriminatory," except in the case of Pike v. Bruce Church. C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1694 (Souter, J., dissenting) (1994). See also Philadelphia v. New Jersey, 437 U.S. 617, 624 (1978).

[207] See, e.g., Mintz v. Baldwin, 289 U.S. 346 (1933). See also Wyoming v. Oklahoma, 112 S. Ct. 789 (1992); Maine v. Taylor, 477 U.S. 131 (1986).

[208] Kassel v. Consolidated Freightways Corp., 450 U.S. 662 (1981); Philadelphia v. New Jersey, 437 U.S. 617 (1978).

[209] Tribe, supra note 40, at 411.

[210] See Tushnet, supra note 59, at 132-33.

[211] This is part of the explanation for Prof. Regan's ability to argue that when the Court balances, it is in actuality applying what he refers to as the "anti-protectionism principle." Regan, supra note 74, at 1108-1109.

[212] Sykes, supra note 104, at 118. The converse is not true: measures that are not discriminatory may well not be the least trade restrictive alternative.

[213] See text accompanying note 184, supra.

[214] See Donald H. Regan, Siamese Essays: (I) CTS Corp. v. Dynamics Corp. of America and Dormant Commerce Clause Doctrine; (II)Extraterritorial State Legislation, 85 MICH. L. REV. 1865, 1870 (1987).

[215] Wine and Beer Panel, supra note 191, at 294, para. 5.72.

[216] Reformulated Gasoline Panel Report, supra note 164.

[217] Id. at 9, para. 3.19.

[218] Id. at 34, para. 6.9.

[219] Id.

[220] Similar issues may arise in other settings. See Minnesota v. Clover leaf Creamery Co., 449 U.S. 456 (1981), where the Court upheld a Minnesota law banning non-returnable plastic milk containers, but not banning non-returnable containers made of other materials. This distinction could have been viewed as de facto discrimination, if the Court had determined to take a less benign view of the legislative categories. On the other hand, in several recent GATT and WTO decisions, under provisions other than art. III(4) of GATT, panels have interpreted the "like product" concept narrowly. See, e.g., Japan--Taxes on Alcoholic Beverages, 11 July 1996; Japan--Tariff on Import of Spruce-Pine-Fir (SPF) Dimension Lumber, 36 B.I.S.D. 167 (1990). It is relevant that these interpretations consider "like products" under different provisions, as it is recognized that the meaning of this term varies depending on the context in which it is used.

[221] In addition, Farber & Hudec have pointed out that art. III analysis of this type, examining the regulatory justification of the product categories, predetermines the outcome of a normally analytically subsequent art. XX analysis. That is, if the panel criticizes the regulatory categories to find discrimination by virtue of different treatment of like products, it is virtually impossible to find that the discrimination is supported by regulatory justification under art. XX. Farber & Hudec, supra note 9, at 1424. But see the Reformulated Gasoline SAB opinion, supra note 164, in which the SAB seemed to separate these two analyses, allowing that the U.S. regulatory categories, while discriminatory under art. III, might be exempted under art. XX(g) if not discriminatory or arbitrary within the meaning of the chapeau provisions.

[222] This perspective belies assertions of the illegitimacy of judicial balancing, arguing that balancing is inevitable. Thus, when Justice Scalia criticizes balancing and asserts that the U.S. Supreme Court's commerce clause role should be limited to policing discrimination, one response is simply that anti-discrimination requires balancing also. See CTS Corp. v. Dynamics Corp., 481 U.S. 69, 95-96 (1987) (Scalia, J., concurring). See also Regan, supra note 74, at 1160-67, 1174-82 (arguing that the Court should only police discrimination, and not engage in balancing). It is noteworthy that certain conservative law and economics theorists and judges recommend cost-benefit analysis, which has much in common with balancing tests, when applied in the domestic sphere to determine whether to regulate or to adopt a laissez-faire stance, and how to regulate, but criticize the same type of methodology when it is applied to determine whether a local regulatory measure has greater trade benefits than regulatory costs. Does the cost-benefit methodology raise fewer concerns in the local sphere than in the international or federal context? Or are these theorists and judges evincing a prejudice toward derogation of authority to smaller units: the market agent in the local regulatory context and the state in the international or federal context?

[223] See, e.g.,, Smith, supra note 21, at 1212-13. Smith argues for a definition of discrimination that is based on either economic protectionism or externalization. These categories are significantly wider than traditional discrimination. See also Farber & Hudec, supra note 9, at 1428-29. Farber and Hudec suggest that the formulation of NTR specified in the Uruguay Round Services Agreement may solve some of the problems with NTR as otherwise configured, by considering whether the national measure in question "modifies conditions of competition" in favor of domestic suppliers. This text also stipulates that it does not cover circumstances where "inherent competitive disadvantages" result from the foreign character of the relevant services or suppliers. General Agreement on Trade in Services, art. XVII(1) n.11. What modifies conditions of competition from "natural" conditions, and what is or is not inherent to foreignness, will be subject to dispute.

[224] See Smith, supra note 21, at 1219 ("It is not possible to specify how uneven the distribution of burdens and advantages must be to count as a discrimination").

[225] See, e.g., the First Tuna Panel Report, supra note 162, the Second Tuna Panel Report, supra note 162, and the Reformulated Gasoline Panel Report, supra note 164.

[226] Section 337 Panel Report, supra note 190, at 392, paras. 5.25-5.27 ("It was clear to the Panel that a contracting party cannot justify a measure inconsistent with another GATT provision as 'necessary' in terms of Article XX(d) if an alternative measure which it could reasonably be expected to employ and which is not inconsistent with other GATT provisions is available to it"); Thailand--Restrictions on Importation of and Internal Taxes on Cigarettes, 37 B.I.S.D. 200, 223 (1991) (citing the Section 337 Panel Report) [hereinafter "Thai Cigarettes Panel Report"]; the Wine and Beer Panel Report, supra note 191; First Tuna Panel Report, supra note 162, and Second Tuna Panel Report, supra note 162.

[227] Section 337 Panel Report, supra note 190, at 393, para. 5.27.

[228] Reformulated Gasoline Panel Report, supra note 164, at 38, para. 6.22: "The Panel noted that it was not the necessity of the policy goal that was to be examined, but whether or not it was necessary that imported gasoline be effectively prevented from benefitting from as favourable sales conditions as were afforded by an individual baseline tied to the producer of a product." This position seems consistent with prior GATT panel decisions. See the reports cited in Reformulated Gasoline SAB Report, supra note 164, n. 28.

[229] The S&P Agreement, supra note 6, specifically (although not unambiguously) adds a reasonableness qualification. See note 233, infra. These provisions leave some ambiguity in light of art. 2.2 of the S&P Agreement, which provides a necessity test in respect of the application of sanitary and phytosanitary measures, but lacks a reasonableness qualifier.

[230] Reformulated Gasoline Panel Report, supra note 164, at 40-41, paras. 6.26-6.28; Reformulated Gasoline SAB Report, supra note 164, at 27. Note the similarity to the decision in Dean Milk Co. v. City of Madison, 340 U.S. 349 (1951), where the Court found it reasonable to require that Madison send its inspectors further than five miles outside the city.

[231] But see Dunoff, supra note 4, at 1419: "the Panel's assertion that the United States had not engaged in multilateral efforts to address the tuna/dolphin problem is simply incorrect."

[232] Standards Agreement, supra note 6, art. 2.2 (emphasis added). However, see note 35, supra, for a possible explanation of this language based on the negotiating history.

[233] Indeed, Farber and Hudec argue that both the Standards Agreement and the S&P Agreement "call for a balancing analysis similar to what one finds in the opinions of U.S. courts in [dormant commerce clause] cases." Farber & Hudec, supra note 9, at 1431. See also Sykes, supra note 104, at 78-79.

[234] S&P Agreement, supra note 6, art. 5.6. Footnote 3 thereto states as follows: "For purposes of paragraph 6 of Article 5, a measure is not more trade-restrictive than required unless there is another measure, reasonably available taking into account technical and economic feasibility, that achieves the appropriate level of sanitary or phytosanitary protection and is significantly less restrictive to trade." This is LTRAT--RA. See also art. 2.2: "Members shall ensure that any sanitary or phytosanitary measure is applied only to the extent necessary to protect human, animal or plant life or health . . . ."

[235] However, art. 5.4 of the S&P Agreement, supra note 6, exhorts WTO Members, "when determining the appropriate level of sanitary or phytosanitary protection, [to] take into account the objective of minimizing negative trade effects."

[236] Standards Agreement, supra note 6, art. 2.2.

[237] Bacchus Imports, Ltd. v. Dias, 468 U.S. 263, 270 (1984); Hunt v. Washington State Apple Advertising Commission, 432 U.S. 333, 336 (1979).

[238] "Discrimination against interstate commerce in favor of local business or investment is per se invalid, save in a narrow class of cases in which the municipality can demonstrate, under rigorous scrutiny, that it has no other means to advance a legitimate local interest", C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1683 (1994), citing Maine v. Taylor, 477 U.S. 131 (1986); Philadelphia v. New Jersey, 437 U.S. 617 (1978); Dean Milk Co. v. City of Madison, 340 U.S. 349, 354 (1951).

[239] Fort Gratiot Sanitary Landfill, 504 U.S. 353, 359 (1992), quoting New Energy Co. of Indiana v. Limbach, 486 U.S. 269, 274 (1988).

[240] Dunoff, supra note 4, at 1449-50.

[241] 397 U.S. 137 (1970) (emphasis added).

[242] See also Henkin, supra note 102, at 1040: "The principle of balancing implies a sub-principle that the state use the means of meeting its needs which is least burdensome to interstate commerce." Id., citing Dean Milk Co. v. City of Madison, 340 U.S. 349 (1951).

[243] Case 120/78, Cassis de Dijon, [1979] E.C.R. 649, [1979] 3 C.M.L.R. 494.

[244] CASS R. SUNSTEIN, AFTER THE RIGHTS REVOLUTION: RECONCEIVING THE REGULATORY STATE 181 (1990)

[245] Emiliou, supra note 7, at 322-23, citing Case 11/70, Internationale Handelsgesellschaft GmbH v. Einfuhr-und Vorratsstelle fur Getreide und Futtermittel, [1970] E.C.R. 1125, 1127, [1972] 10 C.M.L.R. 255 (the principle of proportionality is "an integral part of international law and of supra-national law, such that a Community statute contrary to these concepts should be considered void").

[246] TEU, supra note 144.

[247] Case 11/70, Internationale Handelsgesellschaft GmbH v. Einfuhr-und Vorratsstelle fur Getreide und Futtermittel, [1970] E.C.R. 1125, 1127, [1972] 10 C.M.L.R. 255, 256.

[248] See Joseph Komonchak, Subsidiarity in the Church: The State of the Question, 48 THE JURIST 298, 299 (1988) (quoting Pius XI, Quadregismo Anno 79 (1931)).

[249] Case 302/86, Commission v. Denmark, [1988] E.C.R. 4607, 4632, [1989] 1 C.M.L.R. 619, 632.

[250] Treaty of Rome, supra note 144, art. 3b (as amended in 1992).

[251] European Council in Edinburgh, Conclusions of the Presidency, Dec. 12-13, 1992, Annex 1 to Part A, at 8 (summarizing results of the Edinburgh Summit).

[252] Case C-288/89, Stichting Collectieve Antennevoorziening Gouda v Commissariaat voor de Media, [1991] E.C.R. I-4007 at 4041 (para 15). See also George A. Bermann, Taking Subsidiarity Seriously: Federalism in the European Community and the United States, 94 COLUM. L. REV. 332, 386-87 (1994). The NP component described by Bermann shades closer to CBA than that described above: "the costs of the measure must not manifestly outweigh its benefits." Id., citing Case C-331/88, Regina v. Minister of Agriculture, Fisheries and Food, ex parte Federation Europeenne de la Sante Animale (FEDESA), [1991] 1 C.M.L.R. 507, 532-33.

[253] E.g., Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456 (1980). See also note 205, supra.

[254] Id. at 473.

[255] For a U.S. domestic case distinguishing cost-effectiveness in a static CBA sense from a statutory requirement that costs of environmental protection not be "wholly disproportionate" to benefits, see Chemical Mfrs. Ass'n v. EPA, 870 F.2d 177, 205 (5th Cir. 1989), cert. denied, 495 U.S. 910 (1990). See also, Eric T. Mikkelson, Comment: Earning Green for Turning Green: Executive Order 12,291 and Market-Driven Environmental Regulation, 42 KAN. L. REV. 243, 252 (1993).

[256] See, e.g., Case 302/86, Commission v. Denmark, [1988] E.C.R. 4607, [1989] 1 C.M.L.R. 619.

[257] Case 169-91, [1992] E.C. R. I-6635, recital 15.

[258] WH Smith Do-It-All Ltd. v. Peterborough City Council, [1990] 2 C.M.L.R. 577 (Queen's Bench 1990).

[259] United States-Canada Binational Panel, In the Matter of Canada's Landing Requirement for Pacific Coast Salmon and Herring, Panel No. CDA 89-1807-01, para. 6.13 (Oct. 16, 1989), available in LEXIS, INTLAW Library, USCFTA File, [hereinafter "Salmon and Herring Panel Report"]. The panel was applying GATT arts. XI and XX, which were incorporated into the U.S.-Canada Free Trade Agreement under arts. 407 and 1201, respectively, thereof.

[260] Id. at paras. 7.08-7.09.

[261] CTS Corp. v. Dynamics Corp. of America, 481 U.S. 69, 95 (Scalia, J., concurring) (1986) (citations omitted). Note that this quote refers to a NP test.

[262] Regan, supra note 74, at 1093.

[263] Tribe, supra note 40, at 421 (citations omitted; emphasis in original). See Northwest Cent. Pipeline Corp.v. State Corp. Comm'n of Kan., 489 U.S. 493 (1989); Kassel v. Consolidated Freightways Corp., 450 U.S. 662 (1981); Pike v. Bruce Church, Inc., 397 U.S. 137 (1970). In Raymond Motor Transportation, Inc. v. Rice, 434 U.S. 429, 441 (1978), the Court observed that "experience teaches that no single conceptual approach identifies all of the factors that may bear on a particular case."

[264] Chemical Waste Management, Inc. v. Hunt, 504 U.S. 334, 343, n.5 (1992), quoting Philadelphia v. New Jersey, 437 U.S. 617, 624 (1978).

[265] Hunt v. Washington State Apple Advertising Comm'n, 432 U.S. 333 (1977).

[266] C&A Carbone, Inc. v. Town of Clarkstown, 114 S.Ct. 1677, 1698 (Souter, J., dissenting) (1994). Souter argues that this BT "is similar to, but softer around the edges than, the test we employ in cases of overt discrimination." 114 S.Ct at 1698-99. See also Regan, supra note 74.

[267] Kommers & Waelbroeck, supra note 9, at 219.

[268] Another well-known example in the EU context is labelling requirements in place of substantive regulation. See, e.g., Case 179/85, Commission v. Germany (Re the Use of Champagne-Type Bottles), [1986] E.C.R. 3879, [1988] 1 C.M.L.R. 135 (1988); Case 16/83, Criminal Proceedings Against Karl Prantl, [1984] E.C.R. 1299, [1985] 2 C.M.L.R. 238.

[269] First Tuna Panel Report and Second Tuna Panel Report, supra note 162.


Previous |Next |Title

 

 


Questions or comments about this site?
Email Enfellows@exchange.law.nyu.edu

Top of the page