Table 1: Three Available ETMs
As Table 1 shows, Measure B is the most welfare-enhancing from the viewpoint of PPE. The total of AC and HT are the least in Measure B, whereas EB are the same in all the measures.
Then, we examine the divergence between Country D’s private welfare (PW) and world overall welfare (WW). Suppose that a third of the HT is allocated to Country D (PHT). Table 2 shows the divergence between D’s (PW) and the world’s welfare (WW). All of the AC are to be borne by Country D. Country D obtains all of the EB.
Table 2: Divergence between PW and WW
As Table 2 shows, Country D actually prefers Measure C because PW is maximized in Measure C even though the world welfare is maximized when it invokes Measure B. This result is not surprising because Country D has to bear all of the AC, while the HT are shared by other countries. Country D prefers measures in which the HT are large but the AC are small. This is how the behavior of ETM-imposing countries tends to derogate from optimal even when we think of appropriate levels of health or environmental protection as given.
In these ways, ETM-invoking countries try to externalize the costs of ETMs to outsiders. However, as Coase suggests, this incentive itself does not necessarily mean that the invocation of the ETMs will end up with sub-optimal results.162
In the above hypothetical example, the exporting countries have to bear additional trade harms of $9.7 million ($26.3m x 2 ¸3 - $11.7m x 2 ¸ 3), which is greater than additional Country D’s benefits of $4.4 million ($28.8m - $24.4m) by taking Measure C instead of Measure B. Therefore, the exporting countries can be better off by bribing Country D any amount between $4.4million and $9.7million, suggesting Country D to invoke Measure B instead of Measure C.
However, this Coase-type bargaining between exporting and ETM-invoking countries does not necessarily occur because of the following reasons.
First, trade harms on exporters can be split among many countries. As a result, none of the exporting countries would have an incentive to bargain with ETM-invoking countries. For instance, in the above hypothetical example, suppose that the HT on foreign countries are divided over ten countries, each of which suffers the same amount of harm. That is, the additional amount of harms these countries suffer as a result of the invocation of Measure C instead of Measure B is $0.97 million which is much less than additional Country D’s benefits of $4.4 million. Therefore, individual exporting countries do not have incentives to seek to bargain with Country D. To be sure, bargaining is still possible if the exporters can cooperate for the purpose of bargaining. However, if the number of involved countries is large, then their ability to come together for bargaining may be small since the difficulties of cooperating tend to rise with the number of countries involved.
Second, bargaining would not occur due to the existence of asymmetric informational problems. Assume in the above hypothetical example that all of the HT on foreign countries concentrate on one country, say Country E, and Country E can make only a single take-it-or-leave-it offer to Country D. Country E is risk-neutral. Country E is not exactly sure how much Country D gains additionally by invoking Measure C instead of Measure B. Suppose that Country E think that there are two possibilities: the additional gain by taking Measure C might be $2 million with a probability of 80%, or it might be $6 million with a probability of 20%. In other words, Country E expect that, if it offers $2 million to Country D, it would be accepted with a probability of 80% and be rejected with a probability of 20%. If its offer is rejected, Country E still has to suffer the loss of $9.7 million. Therefore, Country E’s expected loss would be $3.54 million ($2 million x 0.8 + $9.7 million x 0.2) if it offers $2 million to Country D. On the other hand, if it offers $6 million, even though bargaining will occur for sure, Country E’s expected loss is also $6 million, which is greater than its expected loss when it offers $2 million. Therefore, its best offer is $2 million. However, the actual additional gain by invoking Measure C for Country D is $4.4 million, which is greater than Country E’s offer. Consequently, its offer will be rejected by Country D. As this example shows, bargaining would not occur since the parties to the bargaining would not have perfect information about the costs and benefits of available policy options.
Third, even if exporting countries have perfect information regarding the costs and benefits of available ETM options, they would not try to bargain with Country D; rather they would choose to bring an action before the WTO Dispute Settlement Body (DSB) because the initial allocation of “property rights” is unclear. If one of the rules under the WTO clearly admits a right of Country D to invoke Measure C, exporting countries would try to bargain with Country D without bringing the case to the DSB. However, in most of the cases, it is unclear whether the GATT, the SPS Agreement, or the TBT Agreement allows Country D to impose Measure C in order to protect its domestic environment. In fact, it might be also possible that the AB&P interprets these agreements as they guarantee exporting countries’ right of not being disturbed by ETMs like Measure C. As a result, the exporting countries may bring the case to the WTO prior to attempting to bargain with Country D in order to clarify whether they have a right to be protected from ETMs like Measure C. If they win, they have to pay nothing to force Country D to discard Measure C. Even if they lose, then they can try to bargain with Country D. In sum, it is rational for exporting countries to bring a case to the WTO in order to clarify the allocation of “property rights” prior to bargaining with ETM-invoking countries insofar as adjudication costs for the exporting countries are small.163
Because of these reasons, exporting countries and ETM-invoking countries might not bargain, or not be able to bargain successfully. Therefore, we need “trade-off devices” by which we alter the behavior of ETM-imposing countries appropriately.
C. Analysis of the LTRA and “Means and Ends” Tests
PPE implies that comparative cost-benefit analysis (CCBA) is the best way to maximize the net benefits obtained by ETMs. That is, in conducting CCBA, we compare the costs and benefits of available policy options and select the one which improves the net benefit (EB-AC-TH) the most. If the AB&P have perfect information regarding the costs and benefits of available ETM options, they can accomplish the optimal use of the world’s resource in the sense of PPE by striking down ETMs which are not the most efficient among available policy options. However, the AB&P have never invoked CCBA in analyzing ETMs. Instead, they have utilized the LTRA test and the “means and ends” test.
In the following, we compare the LTRA test and the “means and ends” test with CCBA.
AB&P has defined the LTRA as the test which examines whether ETM-invoking countries choose the least trade-restrictive measure among reasonably available ones. “Reasonably available” has two meanings: one is whether a less trade-restrictive measure is administratively available at reasonable cost; the other is whether a less trade-restrictive measure can effectively accomplish the purpose of the measure. The analysis of the “reasonably available” standard has to be done using inherently value-laden judgments; the strictness of the LTRA test may vary depending on how the “reasonably available” standard is applied. It should be noted that the LTRA test does not pay attention to the magnitude of EB. Regardless of the magnitude of EB, ETMs would be upheld as long as the ETM is the least trade-restrictive among reasonably available options. Furthermore, the LTRA test neither “compares” nor “sums up” any value at all. That is, in the process of LTRA scrutiny, AC, EB, and HT are examined independently without comparing them to each other or aggregating them.
As Trachtman points out, the LTRA test is “overbroad and underinclusive” in light of world welfare maximization in the sense of PPE.164 The optimal measure under CCBA could be struck down under the LTRA test if there exists a less trade-restrictive alternative than the optimal option.165 On the other hand, it is also possible that world welfare would deteriorate as a result of the invocation of the least trade restrictive measure if the EB obtained by the measure is extremely small and the HT is relatively huge. Furthermore, it is also pointed out that the least trade-restrictive test is imperfect in the sense hat it excludes policy options by thinking of “appropriate levels of protection” as given.166 That is, there would be measures which can improve the net benefits more than any options which can accomplish the “appropriate levels of protection”.
The “means and ends” test focuses on the relative relationship between HT and EB. AC is excluded from the scrutiny. When HT that occurs as a result of the invocation of an ETM is disproportionately large in comparison with the magnitude of EB obtained by the ETM, the ETM is struck down. Like the standard of the “reasonably availability” in the LTRA test, scrutinizing the “proportionality” between EB and HT inherently requires value-laden judgments. We have to wait for future cases to know how the “proportionality” standard is applied.167 In the process of the “proportionality” analysis, an examiner has to “compare” HT and EB. This is a big difference between the LTRA and “means and ends” tests.168
It should be noted that there is a big difference between the LTRA test and CCBA on the one hand and the “means and ends “ test on the other hand. In the process of the LTRA test and CCBA, multiple measures are examined. Meanwhile, in the scrutiny of the “means and ends” test, the examiner does not pay attention to the availability of other policy options; ETMs are upheld as long as HT is not disproportionately large in comparison with the magnitude of EB even if there is another measure which improves world welfare more than the measure in question. Therefore, the “means and ends” test is imperfect from the viewpoint of world welfare maximization. Furthermore, it can EB said that the “means and ends” test would be inferior to static cost-benefit analysis, which examines whether benefits from the invocation of a policy option exceeds the costs of the policy implementation in the sense of PPE, since administrative costs are excluded from the consideration of the “means and ends” test.
As we have seen, neither the LTRA test nor the “means and ends” test can guarantee the realization of the maximization of world welfare in the sense of PPE. However, this result does not necessarily mean that the use of the LTRA test and the “means and ends” test by the AB&P is inadequate. As Dunoff and Trachtman point out, there are some limitations in the use of the law-and-economics methodology.169 The LTRA test and the “means and ends” test could be a good solution to these limitations in a particular situation.
D. Evaluation of AB&P’s Use of the LTRA and “Means and Ends” Tests
In the following Section, we examine whether the use of the LTRA test and the “means and ends” test could be supported in light of the distributional, ethical, and philosophical problems PPE has.
1. Critiques of PPE
There could be three main criticisms of the use of PPE concept in the context of international law: lack of attention to distributional problems among countries, the incommensurability of values, and problems with interstate comparison of the utility.
First, PPE assumes that there are no transaction costs in redistributing benefits regardless of whether winners actually compensate losers. PPE focuses on whether net benefits increase as a result of the invocation of a measure without regard to redistribtional results of the measure. For example, suppose that there is a measure which benefits the developed countries more than it harms the least developed countries. Assume that the redistribution of the benefits and costs are impossible because of high transaction costs. Under usual Pareto efficiency analysis, this measure cannot be considered to be efficient because the developed countries cannot be better off without making the undeveloped countries worse off. However, this measure is efficient under PPE because the benefits enjoyed by the developed countries exceed the costs suffered by the developing countries.170 PPE itself cannot philosophically, morally, or politically justify why harming others by conducting an action should be allowed when the benefits of the action are greater than the harms.
Second, in conducting CCBA or looking for a measure which accomplishes potential Pareto optimal, we have to “compare” and “amalgamate” various benefits and costs. However, as many have pointed out, it is a very difficult job to quantify all of the social values such as the environment, moral order, and security on a unified scale (for example, a monetized scale) in order to calculate the net benefits gained by a policy option.171 For example, in the context of the protection of endangered species, people value these species primarily for their existence per se, as opposed to values which could be enjoyed by the exploitation or use of the species. That is, people can obtain economic utilities from the very fact that the endangered species are preserved.172 Furthermore, some people also obtain “altruistic” utilities from the fact that endangered species are protected from cruelty or death. However, how can we quantify these kinds of values?173 As Sunstein argues, in real life, we evaluate these values in non-economic ways. As a result, describing these values in economic language would obscure the real nature of non-economic valuations which we are actually using.174 Furthermore, Dunoff maintains that the use of economic language in evaluating these “non-use values” means not only describing them inaccurately but also creating the risk of changing our understanding of these social values.175 In the context of ETMs, the problem of incommensurability of values is the most significant when we evaluate the benefits derived from environmental protection. While some of the environmental regulations are primarily based on use value, for example the protection of human health, the others refer to non-use values, for example, the protection of endangered species. In conducting CCBA, we have to quantify the benefits derived from ETMs so as to render these benefits and costs commensurable with those of trade harms and administrative costs. That is not an easy task.
Third, PPE entails the problem of interpersonal comparison of utilities. In the process of CCBA, an analyst compares benefits obtained by one person with costs suffered by another. However, there is no widely accepted method to evaluate the benefits/costs to another. In order to compare utilities between people, it is required not only that benefits or costs have to be quantified in a comparable scale but also that these quantifications have to accurately reflect the utilities of individuals. In the context of ETMs, this problem appears in the form of interstate comparison of utility. We are not sure how we can compare the benefits of an ETM-invoking country with the harms inflicted by exporting countries.
In the following, we analyze whether the use of the LTRA and “means and ends” tests are good methods for avoiding these problems that PPE has.
2. The LTRA Test
It should be noted that the AB&P have utilized the LTRA test on ETMs which aim at protecting the domestic environment, regardless of which agreement the disputes were brought under. Under the SPS Agreement, Salmon and Apple, the cases in which the LTRA test was applied, were all about the protection of their nationals’ health from risks associated with food consumption. Under the GATT, in Asbestos, the Panel applied the LTRA test to a French decree designed to protect their nationals from health risks arising from the use of asbestos. In this regard, the AB findings in Gasoline, in which the purpose of the U.S. measure was to ameliorate air pollution caused by gasoline combustion in major population areas in the U.S., seems to be somehow odd in light of this general trend that the LTRA test is applied to ETMs for protecting the domestic environment since the “primary aimed at” test was applied in Gasoline. However, we should bear in mind that the chapeau analysis in Gasoline was very close to the LTRA test. The AB’s approach in Gasoline presents a clear contrast to the AB’s findings in Shrimp/Turtle, where the object of the U.S. measure was to protect the global environment and the AB employed a balancing approach in the chapeau scrutiny. Furthermore, we should be mindful of the general tendency that the chapeau tests seem to be more LTRA-like when measures in dispute are in pursuit of the realization of domestic interests. In Bovine Hide, the Panel found that the application of the Argentinean measure in issue, which was aimed at preventing tax evasion of income and value-added taxes, amounted to “unjustified discrimination” because an “alternative course of action” in which the Argentinean government would reimburse importers for the additional interest forgone would be available and administratively feasible.176 The prevention of tax evasion can be regarded as a measure designed to protect “domestic” interests even though it is irrelevant with environmental protection. In sum, regardless of whether under the independent paragraphs of GATT Article XX, under the chapeau of GATT XX, or under Article 5.6 of the SPS Agreement, the AB&P have invoked the LTRA test when the case is about measures designed to protect the domestic interests.
In order to evaluate whether the use of the LTRA test in the context of the domestic environmental protection is appropriate, we need to analyze the above-mentioned distributional, ethical, and philosophical weaknesses of PPE, which raise great concerns in the context of ETMs preserving the domestic environment.
First, as a general proposition, the distributional problem is significant when countries invoke ETMs in order to protect the domestic environment. The benefits of the measure concentrate on the measure-imposing countries by sacrificing the trade interests of foreign exporters. To be sure, the consumers in measure-imposing countries also suffer harms derived from the restriction on international trade. However, countries do not have an incentive to invoke ETMs unless benefits from environmental protection outweigh the total of harms on their consumers plus administrative costs, assuming the ETMs are “genuine”(that is, there is no protectionistic motivation in the ETMs). In other words, ETMs for protecting the domestic environment could be regarded as a kind of “beggar-thy-neighbor” policy; measure-imposing countries are better off by making other countries worse off, even though the net benefits to the measure-imposing countries might exceed the trade harms borne by other countries.177
However, we should also take into account national sovereignty rights to deal with domestic environmental problems. It is an ETM-imposing country, not other countries, which should be able to decide how the ETM-imposing country deals with its domestic environmental problems. The WTO seems to concur with this proposition. The AB&P have repeatedly underscored the point that WTO members are free to set their appropriate levels of health or environmental protection as they wish.178
In this regard, it could be said that the LTRA test would strike an exquisite balance between the sovereign right of a measure-imposing country to protect its domestic environment and a right of other countries to enjoy the benefits of free trade; the LTRA test does not intrude on ETM-imposing countries’ discretion in determining its appropriate levels of environmental protection, whereas requiring them to minimize trade harms on other countries. The LTRA accomplishes the balancing of Members’ rights by not balancing values. It is not coincidental that the AB&P emphasized the sovereign right of Members to decide their appropriate level of protection when they invoked the LTRA test under GATT Article XX (b) or Article 5.6 of the SPS Agreement. To be sure, distributional problems remain unsolved to some extent insofar as outsiders have to suffer trade harms, even though they are minimized. However, as Trachtman points out, compensation for redistribution could be indirect; the application of the LTRA test for all the ETMs for preserving the domestic environment implies that all Members would receive roughly equivalent payoffs. All Members are guaranteed the right to decide their appropriate levels of protection whereas they simultaneously have the right not to suffer trade harms which are more than minimized.179
Second, with regard to the problem of the incommensurability of values, it should be noted that the LTRA test does not “compare” any values at all: it evaluates the magnitude of trade harms independently from any other values; it does not pay attention to the magnitude of benefits derived from environmental protection at all180; furthermore, it considers administrative costs only to the extent of determining whether alternatives are economically feasible. The LTRA test has a strong moral advantage by neither comparing values nor quantifying the value of environmental protection. However, examiners of the LTRA test should bear in mind that there is a sloppy pitfall in the analysis of administrative feasibility. It is possible to argue that administrative feasibility should be determined in light of the magnitude of benefits derived from environmental protection.181 However, if we take into account the benefits of environmental protection in evaluating administrative feasibility, the LTRA test loses its valuable advantages: unnecessity of comparing environmental value with other values and of quantifying the magnitude of environmental benefits. In other words, the LTRA test becomes a mere subset of cost-benefit analysis.182 In this regard, we should be careful of how we interpret the AB findings in Beef with regard to the relationship between the LTRA test and the purpose of trade measures. In Beef, in the analysis of GATT Article XX (d), the AB noted that “[t]he more vital or important…interests or values” are pursued, the easier trade measures would be deemed to be “necessary” to achieve their ends.183 This AB’s comment is subject to many interpretations. It might mean that the regulatory purpose should be taken into account in analyzing the administrative feasibility of alternatives. On the other hand, another possible interpretation is that the AB merely pointed out the general tendency that the more important the regulatory purpose is, the fewer available alternatives there are. Analysts of the LTRA test should avoid the former interpretation since this interpretation would erode the moral weight of the LTRA test. In this regard, it should be criticized that the AB examined the degree of importance of the French measure in question in Asbestos, while the analysis of the French regulatory purpose did not seem to play a significant role in examining the availability of alternatives.184
Third, as to the problem of interstate comparison of utilities, we should be mindful of the AB’s approach when evaluating the magnitude of trade restrictions as a gestalt without distinguishing between trade harms suffered by foreign exporters/producers and those borne by ETM-imposing countries’ consumers.185 By so doing, the AB&P avoid “comparing” utilities among countries.
In conclusion, the LTRA test is a good method for avoiding the distributional, philosophical, and moral problems that CCBA has. In particular, regarding distributional problems which are significant in the context of ETMs for protecting the domestic environment, the LTRA test significantly ameliorates the distributional disadvantages of the domestic ETMs, while guaranteeing the sovereign rights of ETM-imposing countries to decide their appropriate levels of health or environmental protection.
3. The “Means and Ends” Test
Share with your friends:
The database is protected by copyright ©essaydocs.org 2020