Expropriation Jurisprudence under NAFTA: The test for the application of the police powers
Note: there is more information about the Expropriation Jurisprudence under NAFTA tests here. See also the entry about international regulatory expropriation.
By Dr. Howard Mann and Dr. Julie A. Soloway (March 31, 2002)
The establishment of the test for expropriation begins the debate on the expropriation and regulation relationship. The next step is to establish whether the test for expropriation itself, or a related test, addresses the regulation half of the equation more directly. Absent such a second test, it is clear that, almost by definition, a test based solely on the degree of interference of a measure should always capture effective public welfare measures, and especially environmental measures. This in turn would entitle a foreign investor to receive compensation for even bona fides measures.
The second test here relates to the applicability of the police powers rule. As noted earlier, there is no single definition of the police powers rule. Its scope is generally understood to include measures taken by a government under the normal or common functions of governments to protect the environment, human health, consumer protection, regulate hazardous products, and so on. Also as noted earlier, the police powers rule was generally understood in customary international law to be what would now be called a carve out in trade law terms. A measure adopted under a police power should therefore not be subject to compensation: George Aldrich, “What Constitutes a Compensable Taking of Property? The Decisions of the Iran-United States Claims Tribunal”, 88 AJIL 585-610 at 609 (1994). It is not the purpose of this paper to develop a precise argument on each aspect of the expropriation/regulation relationship, but we believe this approach, with its attendant consequences on burden of proof issues, is appropriate.
Of the three cases that have reached a conclusion on the expropriation question, only S.D. Myers has developed arguments that suggest a role for the police powers rule. The development of the legal reasoning on this particular issue may have been preempted by the Tribunals interpretation of the facts. Whether these rulings were correct or not is not the issue here. Whether they had an impact as factual rulings on the further legal reasoning of the Tribunal is a relevant issue. Unfortunately, it is not easy to determine this issue.
In S.D. Myers, for example, the Tribunal ruled that the environmental protection measure in question was really a disguised protectionist measure. Hence no issue of weighing a bona fide measure would arise, although the legal issue was still treated here. In Pope & Talbot, the legitimacy of the measure was questioned by the claimant but supported by the Tribunal. Still, the legal ruling on the lack of any significant impact may have forestalled any further legal analysis. In Metalclad, it was ruled that the environmental protection issues had been managed and approved by the federal level of government in Mexico, hence no real environmental issue was left for determination by the local government that blocked the construction of the hazardous waste site. Thus it found the denial of the permit in question to have been an unlawful measure under Mexican law and hence in its view an expropriation. This aspect of the judgment was one of several vitiated by the British Columbia Supreme Court review of the arbitral decision. United Mexican States v. Metalclad Corporation, (2001) 89 B.C.L.R. 3(rd) 359.
However, this aspect of the ruling does not extend to the Ecological Decree, upon which a finding of expropriation was also made independently of the denial of the operating permit. Despite these problematic factors, the existing cases still do show some treatment of the legal issue being addressed in this section.
Pope & Talbot states clearly that non-discriminatory measures within the scope of the police powers are covered by Article 1110. A reason for this is stated as being the concern that a blanket exception would allow creeping or indirect expropriations to be left unchallenged due to the form of the measure. But the reasoning appears to stop here in the sense of the absence of any test that might address a distinction between bona fide public welfare measures under the police powers rule and other government measures that are confiscatory in a more classic sense. As already described in the previous section, the only criteria used by the Tribunal went to the degree of interference, based on a test of significant impact. However, this may have resulted from the finding that the significance test was not met.
Moreover, it is important to keep in mind that the absence of an established test to distinguish bona fides public welfare measures from protectionist measures does not mean that tribunals will override the police powers rule. It is in fact extremely difficult to reduce a “smell” test into a series of discrete legal questions. This problem is not isolated to international investment, but is a problem of international trade law as well. However, panellists are given discretion in order to be able to have some flexibility in making this determination.
In Metalclad, the Tribunal appears to have gone farther. On one reading at least, it appears to have rejected the applicability of a purpose and effect analysis outright. Doing so, for all practical purposes, has the effect of ending a police powers role, as it is essentially a purpose-based analysis.
Para. 111: “The Tribunal need not decide or consider the motivation or intent of the adoption of the Ecological Decree”
One reading of this position could suggest it may have been taken because a finding on the ecological decree, in relation to which this statement is specifically made, was not essential to the case due to the other findings. Some have argued the whole section of the decision on expropriation was obiter, though this was completely rejected by the BC Superior Court decision that reviewed the Tribunal’s award.(United Mexican States v. Metalclad Corporation, (2001) 89 B.C.L.R. 3(rd) 359; paras. 86-105). More realistically, however, this paragraph must be read in conjunction with the original test for expropriation cited above. The rejection of any need to assess purpose is consistent with its choice of a singular test based on degree of interference. It is also consistent with the absence of any reference to a consideration of purposes in the Pope & Talbot case.
Metalclad presented a simple opportunity to examine the Ecological Decree in the light of the expropriation/police powers dynamic. An analysis of this type might well have produced a conclusion that recognized the right of governments to enact measures pursuant to the police powers rule without paying compensation. The analysis could then have considered whether the measure fell within the scope of police powers. Such an analysis could have then distinguished measures whereby governments acquire title or prevent the continued occupation and any use of land in order to create a state asset – an ecological reserve here- from measures that prevent the use of property in certain ways as to injure others or the environment. It could have done so without much need for potentially more difficult lines to be drawn. It might have noted that in most countries, at least, when land is set aside for public use, as was the case here, that is seen as an expropriation and hence outside the normal scope of the police powers rule. Such an analysis could have reached the same result, without rejecting the basic element of the police powers analysis: looking at why a measure is adopted. Whether deliberately or not, the Tribunal chose simply to reject it. It is this legal choice that remains troubling.
All this being said, the Metalclad case remains controversial. There is widespread disagreement among analysts on the findings of fact, and on the impact those findings have on different parts of the arbitral decision, disagreement which exists between the present two authors. These disagreements lead to differing views on the extent to which the Tribunal could or should have undertaken a serious police powers analysis.
This leaves S.D. Myers as the only case where the consideration of purpose was called for. Unfortunately, all the criteria then used in the main decision went to a degree of interference analysis. Only in the separate opinion of Schwartz does one find a deeper analysis of this issue.
Schwartz notes that regulation is something that owners ought to expect.(S.D. Myers, Schwartz separate opinion, para. 213). Further, he argues that Article 1110 must be read in the context of NAFTA’s overall environmental construction, including the preamble, Article 1114 and the adoption at the same time of the North American Agreement on Environmental Cooperation. All these factors, he argues favour the strong right of governments to protected the environment. Consequently, he concludes, Article 1110 “does not constitute a generous invitation to impose liability for regulatory activity in the ordinary course of government business.”(S.D. Myers, Schwartz separate opinion, para. 214). Towards this end, as noted in the previous sub-section, Schwartz does provide some criteria to identify regulatory measures as distinct from non-regulatory confiscatory measures.
There is no statement that non-discriminatory, bona fide public welfare measures are not within the realm of expropriation. Recall here that his original statement on this issue was that regulatory conduct is not remotely the subject of legitimate complaints in most cases. Nothing is readily apparent in this judgement to determine when one may be in the presence of such a case. This is made a more relevant point by his subsequent refusal to take a final position on the expropriation/regulation relationship in this case in order to preclude possible political impacts from his potential decision. However, as stated above, it would be dangerous to prima facie exclude any form of regulation from the scope of review.
Whether existing tests recognize the police powers rule implicitly is another question to explore, though this has not been done to date in the cases. At least one phrasing of the test for expropriation includes the substantial deprivation of an investor’s reasonably-to-be-expected benefit from the property. Certain regulations designed to protect the public interest may arguably be “reasonably expected” and thus not compensable. For example, a manufacturer of chemicals in a highly regulated industry may not be able to argue that he reasonably expected to sell certain chemicals, especially in the light of the steps leading up to an environmental measure (i.e., research, consultations, etc.) In this way, the reasonable expectations portion of the expropriation test may provide a second line of defence after the requirement that a measure substantially deprive an investor. While we both agree this would be arguable, it has not been addressed as yet in any of the cases.