After a review process that lasted three years, expectations ran high for the revised model U.S. bilateral investment treaty (“BIT”), which was released last month. Stakeholders from many parts of society — the U.S. Congress, environmental organizations, labor groups, business groups, trade associations, academia, the public, and investment experts — weighed in during the review process, with wildly divergent opinions and in some cases, hopes for radical changes. In the end, however, the unveiling proved somewhat anticlimactic, as the new model BIT did not diverge greatly from its 2004 predecessor.

First, what did not change?

The new model BIT made no changes to the substantive investment law protections. Critics had recommended certain reforms, such as limiting the scope of expropriation exclusively to direct appropriations of property by the host state, restricting the scope of claims for violation of the minimum standard of treatment, and limiting the fair and equitable treatment protection to the standard articulated in Glamis Gold. Moreover, they had asked the administration to bar application of other international agreements through the BIT’s most favored nation (MFN) clause, and to limit the application of national treatment provisions to cases in which regulatory measures were enacted primarily for discriminatory purposes. In the end, however, the Administration rejected all such proposals. Accordingly, the new model BIT’s provisions on national treatment, MFN, minimum standard, expropriation, and transfers of funds remain unchanged from the 2004 model BIT.

The arbitration clauses were also kept largely untouched. Critics of investment treaties had requested that the U.S. government either replace investor-State arbitration with State-State dispute resolution, or delay or limit the recourse to investor-State arbitration (for example, by requiring the exhaustion of local remedies and narrowing the definitions of “investor” and “investment”). However, the administration opted to retain unaltered the investment arbitration clauses of the 2004 model BIT. The sole exception was the clause in Art. 28(10) requiring the Parties “to strive to reach an agreement” for appellate review in the event that an appellate mechanism were established; that obligation was watered down in the new model to a requirement that the Parties “consider” whether arbitral awards under the BIT should be subject to any new appellate mechanism. (Consistently with this change, the new model BIT also eliminates Annex D of the 2004 version, which had required the Parties to commence negotiations within three years on an appellate mechanism for investment arbitration; such period had been significantly shortened to three months in CAFTA, was then restored to three years in the U.S.-Rwanda BIT, but has now been completely removed in the new model.) The new U.S. approach on investor-State arbitration contrasts with that of Australia, which has opted to forgo altogether any investor-State dispute resolution mechanisms in its future investment and trade treaties, and to rely instead on its domestic courts. The U.S. position also diverges from that of the European Parliament, which recently adopted a resolution urging that future dispute settlement mechanisms include a requirement that foreign investors must exhaust domestic remedies before initiating investor-State claims.

So what did change significantly in the new model BIT?

First, the administration added new transparency requirements. The new model BIT requires Parties to consult periodically on how to improve transparency practices with regard to publication of laws, decisions respecting investment, and arbitration proceedings. Additionally, a completely new section was added to Article 11 establishing transparency requirements relating to proposed and adopted regulations. Specifically, the model BIT requires Parties to publish proposed regulations, explain their purpose and rationale, allow public comments on them, and address such comments when adopting the final regulations. The new model also adds a requirement that each Party allow persons of the other Party to participate (e.g., by allowing them to comment) in standards-setting exercises; certain exceptions apply, however, including with respect to sanitary and phytosanitary measures. Both NGOs and investors should be pleased with the new transparency provisions, which could have the effect of forestalling a certain percentage of investor-State arbitrations, since the new provisions give investors an opportunity to discuss the effects of regulatory amendments — and host states a chance to reevaluate proposed changes — before final promulgation. In essence, these modifications allows for exchanges of views before the conditions for a dispute actually arise.

Second, the administration expanded the scope of labor and environmental obligations, by imposing on the Parties an affirmative obligation to “ensure” that the Parties do not waive or derogate from domestic labor and environmental laws; the 2004 model had contained a more aspirational provision, calling only for the Parties to “strive to ensure.” The new model also adds a requirement that the Parties ensure that they “effectively enforce” their labor and environment laws. At the same time, however, the model BIT adds a new clause explicitly recognizing the Parties’ right to exercise regulatory discretion (so long as it is reasonable) and to make resource allocation decisions (so long as it is done in good faith). Finally, the new model adds a definition of “environmental law” and expands upon the definition of “labor law.”

Labor and environmental groups had asked the administration to make labor and environmental obligations enforceable through State-State dispute resolution, as in the recent U.S. free trade agreements (FTAs) with Panama, Colombia, and Peru. However, the administration demurred, opting for the detailed and extensive consultation procedure found in some older FTAs. Predictably, the amendments left all interested sectors at least partly dissatisfied: labor and environmental groups consider the modifications too mild, while business groups have expressed concern that these particular provisions may hinder negotiation of new BITs with certain countries.

Third, the administration introduced certain key changes designed to address investments in countries with state-led economies. Such changes include a new clause in the performance requirements article barring host States from requiring the use of domestic technology; a provision (akin to that in the U.S.-Korea BIT) requiring a Party to allow investors of the other Party to participate in the development of technical and similar standards on a non-discriminatory basis; and a clarification that the obligations under the model BIT fully cover state-owned enterprises, as well as other entities or persons that act under delegated governmental authority.

Fourth, the new model includes a number of revisions to the financial service provisions, largely modeled on the 2008 U.S.-Rwanda BIT. For example, the text clarifies that measures relating to payment and clearing systems are within the scope of measures relating to financial services that States may adopt for prudential reasons and that are exempted from the scope of the BIT. The new model also adds a provision entitling a respondent State to ask a tribunal to promptly decide on prudential measures and the applicability thereto of the financial services exception. Further, the model BIT introduces a clarification that the Parties may adopt financial sector measures necessary to secure compliance with laws and regulations otherwise consistent with the BIT (including those designed to prevent deceptive or fraudulent practices, or to address the effects of a default on financial service contracts), subject only to the requirement that they not be applied in a discriminatory fashion. Ultimately, the administration did not include sovereign debt annexes such as those included in CAFTA, and the Panama, Peru, and Colombia FTAs, leaving the matter to future individual treaty discussions.

Finally, the administration included a small but potentially valuable change to the definition of “territory of a Party,” by expressly including within the scope of that term the territorial sea, as well as high seas areas in which the Party may exercise sovereign rights or jurisdiction under customary international law. This amendment should sweep within the scope of the BIT such investments as off-shore oil and gas projects, and fish farms.

The foregoing are but the more salient revisions; the new model includes other revisions, which are less far-reaching or merely formal. Ultimately, the 2012 model BIT changed only moderately from its 2004 predecessor. As investor protections were not diluted, business groups are generally pleased and hope that the new model will jumpstart BIT negotiations. The U.S. is currently a party to 40 BITs — far fewer than Germany and China, for example, which each have approximately 100 — and it lacks a BIT with major trading partners like China, India, Russia, and Brazil. Some have argued that the enhanced transparency, labor, and environment provisions in the new model may make it more difficult to negotiate long-awaited BITs with nations such as China and India. Others counter by noting that a model BIT is simply a starting point for negotiation, and that the amendments in the new model may not actually find their way into the final texts of future BITs. Thus, it shall remain to be seen whether the new model will in fact hinder current or future U.S. BIT negotiations.


The author wishes to acknowledge the contribution of his Arnold & Porter colleagues Dawn Yamane Hewett and Geraldina Mendoza in the preparation of this post.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Kluwer Arbitration
This page as PDF

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.