First, we thank Susan Franck again for her very insightful comments on our recent article, Investment Protection in Extraordinary Times: The Interpretation and Application of Non Precluded Measures Provisions in Bilateral Investment Treaties. We are pleased that she agrees with us on many points and welcome the thought provoking comments she has offered.
Our response focuses on the three issues Professor Franck identifies in the second half of her post, namely the questions of risk allocation, whether NPM clauses constitute an exception to liability, and, finally, whether the Argentine cases have broader applicability.
First, Professor Franck suggests that our paper treats bilateral investment treaties (BITs) as risk allocation devices. We fully agree with Professor Franck and recognize in our paper that BITs perform (or at least are assumed to perform) a broad range of functions including many of those she mentions. If one looks to the origin of the US BIT program, however, it becomes apparent that BITs also represent a bargain between free traders, who favored strong investor protections, and protectionists, who sought to preserve state freedom of action, particularly in the national security domain. The Exon-Florio Amendment is an oft-cited example of this bargaining process. Our focus on the risk allocation function of BITs is in part a response to this early bargain that underlies many states’ BIT programs. While scholarship to date has largely focused on the investor protection side of that bargain, our analysis of non precluded measures (NPM) clauses looks at how states have sought to limit their risks under BITs while simultaneously protecting cross-border investment.
We do not, however, argue, as Professor Franck seems to suggest, that BITs are “akin to a form of insurance” to protect “bad business judgments.” Rather, we recognize that BITs containing NPM clauses strike a balance. On one side of that balance, the substantive protections of the BIT limit the risks faced by investors by providing them legal recourse when their investments are interfered with by host states. On the other side of the balance, the NPM clauses in those BITs limit the risks faced by host states in the kinds of emergency situations covered by the NPM clause. The resulting treaty thus reflects the bargain between the two sets of interests in both the capital exporting and host states that allocates risks between the host state and investors.
Second, Professor Franck raises the interesting question of whether NPM clauses should be thought of as an exception to or an exclusion from liability. In fact, as the comparative analysis in our article recognizes, different states frame their NPM provisions in different ways. The net effect of either framing, however, is to remove actions taken by the host state that fall within the scope of the NPM clause from the substantive protections of the BIT and thereby preclude liability.
Professor Franck is correct to note that in most cases the affirmative defense of necessity in customary international law should be unnecessary where a BIT contains an NPM clause. More specifically, where a treaty contains an NPM clause of comprehensive scope, the narrow necessity defense under customary law will generally not become relevant. NPM clauses are generally drafted to provide states greater flexibility to respond to emergency situations than would have been available under the customary law defense of necessity. Hence, if a measure fails to be in conformity with the broader treaty-based NPM clause, it will also fail the narrower customary tests, and where it is covered by the NPM clause, the question of whether it also meets the customary necessity requirements becomes moot. There are two situations, though, in which the customary defense may become outcome-determinative. The first is provided by cases in which the BIT at issue does not contain an NPM clause to begin with; in such a case, the respondent state is left with the defenses provided by the customary law of state responsibility. The second concerns cases in which the applicable BIT does include an NPM clause, but of the limited type. Here, the NPM clause may fail to cover a certain state measure because the latter infringes on a treaty provision that falls outside of the NPM clause’s scope, but that measure may still be covered by the necessity defense whose coverage is not as such limited to any specific provisions.
Third, Professor Franck asks whether the Argentine cases are representative and make good law. We suggest that in increasingly globalized world, the threats of economic collapse, public order emergencies, or pandemic diseases are becoming all too common and that the Argentine experience is, unfortunately, likely to be repeated elsewhere if in a somewhat different form. From the jurisprudence of the ICSID tribunals that have issued awards in the Argentine cases thus far, it appears that these cases are not, in fact, leading to good law. As the Annulment Committee in the CMS case observed, the CMS Tribunal provided an “erroneous interpretation” of the NPM clause in the U.S.-Argentina BIT that “could have had a decisive impact on the operative part of the award.” Hard cases may well make bad law and we must hope that the Annulment Committee report and the cases against Argentina still to be decided will push toward better law that reflects the treaty commitments of Argentina and the United States.
On this point, Professor Franck further suggests that the significant sums awarded against Argentina thus far are statistical outliers. Her own empirical work on this question indicates that many ICSID awards are in fact relatively small, particularly compared to the awards against Argentina that routinely run into the hundreds of millions of dollars. Even if these awards are statistical outliers, states within the ICSID system are certainly taking notice. In May 2007, for example, Bolivia notified the World Bank that it was withdrawing from the ICSID Convention and Bolivian President Evo Morales urged his Latin American counterparts to do the same. Other states such as Venezuela and Ecuador have noted the desire to limit ICSID jurisdiction and minimize potential BIT liability. Bad law, even in a few outlier cases, thus poses real challenges to the legitimacy and effectiveness of the ICSID system and investor-state arbitration more generally. By fully recognizing the risk allocation states sought to create when they included NPM clauses in their BITs, that threat to the legitimacy and effectiveness of ICSID arbitration may be avoided. We hope our paper can contribute to that process.