Treaties and Custom in Trade and Investment

Last week I had the privilege to attend an investment arbitration conference and FDI moot court competition at Pepperdine. Kudos to Murdoch University of Australia for winning the competition and my alma mater NYU for winning the highest overall ranking.

There was much to ponder in the conference from the likes of Andrea Bjorkland, Todd Weiler, Anna Joubin-Bret and Mark Kantor. For me at least, Andrew Guzman’s presentation was the most challenging, in which he summarized his theory that BITs do not create or reflect customary international law with respect to the standard of compensation for unlawful expropriations.

One of his most interesting arguments was to compare state practice in the trade context and the investment context. He notes that the WTO has 153 members and yet no one would argue that MFN or national treatment are standards that should be applied as a matter of customary international law vis-à-vis non-WTO members. If this is true (which it surely is), then why should we assume that BITs create or reflect customary international law?

Guzman has a point. If you run the numbers, the WTO reflects general and consistent practice among nations to a greater extent than BITs. Of the 18,721 possible country pairings (assuming there are 194 independent countries), the 153 members of the WTO reflect 11,628 country pairings, or 62 percent of all possible pairings. By contrast, according to the latest estimates from UNCTAD, there are 2,750 BITs, reflecting only 15 percent of all possible country pairings. Thus, 85 percent of all possible investment pairings are outside a BIT, whereas only 38 percent of all possible trade pairings are outside the WTO.

On this basis, can one say that the WTO standard of non-discrimination either reflects or creates a customary international law norm of non-discrimination in the treatment of goods from all countries? The answer has to be no. But why?

I would suggest that the reason is that WTO members clearly intended for non-discrimination to be a benefit that accrues only to members. The WTO is, to borrow a distinction from Jose Alvarez, a “contract” treaty rather than a “legislative” treaty such as the VCLT. We know this in part by the process of tariff negotiations in trade rounds and in accession talks. We also know this because there is a general and consistent practice of discriminating against non-WTO members in collecting duties. The Harmonized Tariff Schedules of virtually every country have a general column for the tariff rate of WTO members, and a separate column for higher tariffs for non-WTO members. Differential treatment among WTO members and non-WTO members is an uncontroversial feature of international trade. (Of course, this is a simplification. Each year the United States grants certain non-WTO countries, such as Russia, MFN treatment under Jackson-Vanik. Non-discrimination is a right WTO members enjoy; it is a privilege non-WTO members may receive as a matter of annual discretionary review.)

Can the same be said of international investment law? It certainly would be odd to devise a standard of compensation for all nations based on the state practice of 15 percent of nations. But that may not be the correct way to count countries. If virtually every country has signed a handful of BITs, then one might say this reflects a general consensus of nations, even though it is a small percentage of the total number of possible dyads.

As a rational choice scholar, Guzman’s essential question is why a developing country would sign a BIT. His answer is that BITs give a capital-importing host country a competitive advantage over other similarly-situated countries. Why establish a textile factory in country X without a BIT when a corporation can go to country Y with a BIT? On this theory, foreign investors follow BITs that afford greater protection than they would otherwise obtain.

But there may be other rational choice explanations for why a state would sign a BIT even though the standard for compensation for expropriations may not change. The purpose of BITs may not be to alter the standard of compensation enjoyed under customary international law (or under contract or constitutional law), but rather to provide an effective mechanism for enforcement of the recognized standard of compensation. On this theory, BITs are about effective procedure, not new substance. Another theory is that BITs offer a different type of substantive protection than customary international law. BITs incorporate new substantive rights, such as fair and equitable treatment. One need not deny the customary international law status of fair market compensation for takings to find a rational explanation for the proliferation of BITs.

Of course, the larger question of whether BITs reflect or create custom remains controversial. BITs standing alone arguably do not create custom, but they provide additional evidence of custom, combined with other indicia such as FCN treaties, the jurisprudence of international tribunals (such as modern ICSID awards, and older decisions from the Iran-United States Claims Tribunal and ad hoc tribunals), and state practice reflected in domestic takings laws and international and domestic government contracts. Moreover, unlike the WTO, it is hard to identify a general and consistent practice of states that distinguishes the standard for members as compared to non-members in the investment context. There is nothing akin to Harmonized Tariff Schedules that pronounces a differential standard for states that do not enjoy investment treaty rights.