State Liability for Regulatory Change: How International Investment Rules are Overriding Domestic Law

With governments around the world pushing efforts to negotiate and approve mega-investment treaties, it is important to be clear on just what these investment treaties do and do not mean. One issue that is increasingly apparent is that investment treaties are not merely tools to provide protections against abusive regimes and egregious conduct, but are mechanisms through which a small and typically powerful set of private actors can change the substantive content of the law outside the normal domestic legislative and judicial frameworks.

Some might counter that contention. Indeed, the European Commission recently issued a statement enthusiastically supporting investment treaties and investor-state dispute settlement, and labeling as flatly “untrue!” concerns that investor-state dispute settlement “subverts democracy,” “takes place behind closed doors,”  “undermines public choices” and is handled by “a small clique of lawyers”.[1] But, evidence from decisions regarding state liability for regulatory change shows something different.

This article, which draws from a more detailed study, compares U.S. domestic law and international treaty rules on state liability for regulatory changes. It shows that arbitral tribunals have interpreted investment treaty rules in a manner far more favorable to the interests of investors than the approaches adopted in U.S. courts.[2]

Investor-state arbitration and state liability for regulatory change

When are states liable for regulatory change that hurts the profitability or value of an investment? The answer to that question in domestic law reflects lawmakers’ decisions regarding how to appropriately balance public and private interests, and has very real implications for a government’s willingness and ability to introduce, monitor, and enforce measures that regulate private conduct in order to serve broader public goals. Arbitral tribunals interpreting and applying investment treaties, however, are issuing decisions that override those domestic choices.

These tensions between domestic law and international investment treaties are particularly evident when looking at the issue of state liability for changes in the general legal framework that impact an existing investor-state contract or quasi-contractual relationship, such as a permit, license or authorization issued by the government to a private entity. On this issue, arbitral tribunals have stated that one core obligation in investment treaties—the fair and equitable treatment (FET) obligation—protects the “legitimate expectations” of investors made at the time of the investment;[3] and if the legal framework governing the investment changes in a way that was not anticipated or foreseen by the investor at the time of making the investment, then the investor should be compensated for the cost of complying with those changes.[4] This means that if a new law is adopted, or an existing law is revoked or interpreted or applied in a new way,[5] those changes can trigger state liability. Various tribunals have refined and arguably softened that rule of “legitimate expectations,” stating that investment treaties do not generally act to freeze the law unless those changes are contrary to a specific commitment made by the state.[6] For those tribunals, the key to whether they will require governments to compensate investors for regulatory change is their view of what constitutes a “specific commitment” to refrain from making such changes.

In a number of cases decided to date, tribunals have interpreted the concept of a “specific commitment” broadly. In cases such as EDFI v. Argentina,[7] Enron v. Argentina,[8] LG&E v. Argentina[9] and Occidental v. Ecuador (2004),[10] the tribunals have found provisions in general domestic laws and regulations to constitute non-revocable commitments. The commitment, they have thus concluded, need not be so “specific.”[11] Tribunals have also bound governments to “promises” they have found or inferred from statements by government officials and representatives of state-owned enterprises, positions taken by agencies, and even illegal contracts or deals involving procedural or other irregularities.[12]

How this differs from domestic law—example of the gap between tribunal decisions and U.S. courts  

Notably, the broad rule that governments should compensate investors for changes in the general regulatory framework that impact their expectations and profitability as well as the narrower interpretation that governments are only liable to compensate for regulatory change that is inconsistent with a “specific commitment” given by a state to an investor, both privilege private rights over governmental regulatory freedom in a way that is inconsistent with domestic rules, such as those of the United States.

Comparing investment arbitration decisions regarding liability for regulatory change with U.S. case law addressing similar factual circumstances, for instance, illustrates that U.S. law takes a much narrower view of private rights. U.S. cases addressing the precise issue of state liability for regulatory change impacting investor-state contracts and quasi-contracts show that:[13]

–        The general rule is that the state will not be liable to private parties for economic harms suffered as a result of general regulatory change; and

–        The government may in certain cases have to compensate an investor for losses suffered as a result of general regulatory changes that impact a contractual or quasi-contractual relationship with the government, but, due to strict application of the doctrine of “unmistakability” and related rules, the government is largely shielded against liability in these cases.

More specifically, under U.S. law and its doctrine of “unmistakability,” liability will only be found when an official or entity with the (1) actual authority to make a promise regarding future regulatory treatment, (2) makes that promise in a clear and unmistakable way and (3) in a manner fully consistent with relevant procedural requirements for entering into investor-state contracts, and (4) does so with the intent to bind itself to that particular commitment regarding future regulatory treatment.

Case law has elaborated upon each of these requirements. On the requirement of actual authority, for instance, courts have explained that even if a government entity has authority to set tariffs for water use, that does not mean that it also has the authority to give up or restrict its sovereign power to set those tariffs.[14] The power to set rates is not the same as the power to promise to freeze or stabilize them, and for an agency to exercise the latter power it must have been clearly delegated that ability. Notably, the doctrine of estoppel is largely unavailable under U.S. law to protect investors in cases of mistaken reliance on promises made by government actors that exceeded the bounds of their authority.[15]

The requirement of intent has also been interpreted in a way that shields the government from liability. Courts have concluded, for example, that clear intent to induce investment by promising a certain type of regulatory treatment is different from and does not establish intent to induce investment by promising continued enjoyment of that regulatory treatment.[16]

Similarly, any alleged promise by the government to compensate an investor for the effect of a sovereign act must be “unmistakable.” This requirement acts as a “rule of strict construction that presumes that the government, in making an agreement regarding its regulation of a private party, has not promised to restrain future use of its sovereign power, unless the intent to do so appears unmistakably clear in the agreement.”[17] In one case illustrating the force of this rule, the Supreme Court found that a promise in a legislative act to “forever exempt” a water services system from taxes did not unmistakably establish a promise to never “exercise the reserved power of amending or repealing [that] act.”[18] The Supreme Court reasoned that the “utmost” that could be said was that when the legislature passed the law “forever” exempting the water system from taxes it had no intent “to withdraw the exemption from taxation; not that the power reserved would never be exerted … if in the judgment of the legislature the public interests required that to be done.”[19]

In addition to having to comply with substantive legal requirements, promises made by government entities to waive or compensate for regulatory change must also strictly comply with applicable procedural rules designed to prevent impropriety in the contracting process. Agreements concluded with the U.S. government in violation of those rules have traditionally been declared void ab initio. No actual collusion or fraud need be shown.[20]

There is notable divergence between international investment tribunals’ and U.S. courts’ respective assessments of the scope of enforceable “commitments” and government liability for interference with those undertakings. Both arbitral tribunals and U.S courts declare deference to sovereign acts of general applicability; both also recognize that governments do not have unbounded authority to exercise their sovereign power to the detriment of investor-state contracts. Nevertheless, they differ in terms of the respective tests they apply to determine whether the government promised not to exercise its authority or to provide compensation for future regulatory changes.[21]

Key points of distinction between the two systems include tribunals’ apparent willingness to find implied enforceable and non-revocable commitments against regulatory change, and to hold governments to particular undertakings that, under domestic law, may not be legally binding on either the government or the investor due to substantive or procedural failings. Similarly, tribunals have read ambiguity in the contract in favor of affirming, rather than rejecting, the existence of a commitment to waive future use of sovereign power. The Enron tribunal, for instance, stated that if the legal framework existing at the time “was intended to be transitory[,] it should have also been clearly advised to prospective investors.”[22] Likewise, the EDFI tribunal asserted that if Argentina had not intended to bear the risk of loss for future regulatory changes, it “could have said so” in its contract.[23] Both cases required the states to explicitly reserve future exercises of sovereign power, and thus stand in stark opposition to the U.S. unmistakability cases, which will only enforce promises to refrain from future exercises of sovereign power if there is mutuality of intent behind the promises and the commitments themselves are clearly expressed.

Another area of divergence relates to how a finding of unmistakability or a specific commitment can be impacted by the purpose or type of regulatory action that it purports to freeze. With respect to the purpose of the regulatory action, the early U.S. cases indicate that courts strictly applied the “unmistakability” test when applying a more relaxed rule could have threatened development of the new nation and its efforts to construct and operate crucial infrastructure. Likewise, courts today appear reluctant to find “unmistakable” promises of legal stability where the existence and enforcement of such promises would hinder the government’s ability to respond to crises, react to matters of public interest, and address harms caused or negative externalities imposed by private actors once the problems are discovered.

U.S. courts also appear to base the strictness with which they apply the “unmistakability” test on the type of action at issue, evidencing heightened concern regarding interfering with the government’s exercise of its taxation powers. By contrast, in international investor-state arbitration, neither the purpose nor type of the regulatory action at issue has seemed to impact the level of scrutiny tribunals have applied to determine whether the government had in fact guaranteed to waive its powers. Indeed, tribunals have found implied promises of stability that barred government action taken in response to financial crises and through the exercise of fiscal policy.

Finally, a fifth area of divergence is in the relevance U.S. courts and investment tribunals respectively assign to the temporal scope of the alleged commitment. In U.S. decisions, courts have emphasized that a commitment to accord a specific form of treatment does not imply a commitment to accord that treatment over the life of the contract. The degree of the waiver seems to affect scrutiny of the “unmistakable” nature of government guarantees that purport to restrict the authority of future administrations to respond to changing constituents, policies, and circumstances. Indeed, a number of cases finding no “unmistakable” promise of regulatory stability involved alleged promises that purported to last for decades, if not indefinitely.[24]

Decisions by investment tribunals to date reflect less unease with strictly enforcing long-term promises. In a number of cases, a framework established in law has been interpreted to be a framework that persists over time. Tribunals have also further elevated the importance of stability and of maintaining promises in accordance with their original terms by awarding lost profits over the originally foreseen life of intended deals and in accordance with the legal regimes applying to those arrangements at the time of their conclusion.[25]

How this impacts and overrides domestic law

In short, U.S. domestic rules regarding government flexibility to change the applicable regulatory framework differ from rules being developed and applied by investment tribunals. [26] The question this raises is whether tribunals will apply these rules on “specific commitments” to such domestic jurisdictions that take a different view of limits on sovereign powers.

The answer appears to be “yes”. Through this approach, tribunals have evidenced that they view investment treaties and, more specifically, the FET obligation, as implicitly creating a new category of investor rights that the investors would not have received under the relevant contracts/quasi-contracts or the domestic legal frameworks governing those instruments.

Tribunals have thus attached “new legal consequences” to pre-existing contractual or quasi-contractual relationships between investors and states, retroactively changing the rights and obligations of those actors.[27] Domestic delineations of private property rights are thus vulnerable to being overridden by arbitral tribunals with their own interpretation of what rights economic actors have been given under investment treaties.

Reining in claims seeking damages for regulatory change?

Because there is no system of binding precedent in international investment law, the fact that tribunals have taken certain approaches to “specific commitments” in the past does not mean that they will continue to do so in the future. Thus, future tribunals could soften the rule that has been applied in the past, and look to domestic law when defining the scope of property rights investors claim were harmed by conduct breaching the investment treaty. But there is no guarantee that tribunals will do so. Investment treaties give private arbitrators significant powers of interpretation, and other international treaties (i.e., the New York Convention and the ICSID Convention) largely insulate tribunals from formal or informal checks on their power. Even where the state parties to the treaty take a common and consistent position on their view of the meaning of a given treaty provision, that is no guarantee that the tribunals will follow those states’ mutually-agreed positions;[28] and where tribunals issue an interpretation with which states disagree, there are few, if any, mechanisms through which states can set tribunals back on the correct path.[29]

Moreover, through past case law, tribunals are sending signals to investors about how investment treaties can be used to challenge regulatory change. One of many examples of how investors are picking up these signals is Eli Lilly v. Canada, a dispute in which the investor is challenging rulings of Canadian courts interpreting Canadian intellectual property law, arguing that those judicial decisions improperly changed the host country’s legal regime in violation of the investor’s “legitimate expectations.”[30] Similarly, in Guaracachi v. Bolivia, the investor argues that the host country breached the FET obligation when it “effected a fundamental change to the regulatory regime that attracted” the investor’s investment.[31] Of course, not all claims will succeed. But if leading law firms are signing these filings, this illustrates that at least some experts believe these claims have enough legal merit to launch a costly case.

The conclusion this produces is that investment treaties—as they are being used by investors and applied by some tribunals—are not merely instruments to protect foreign investors against outrageous and discriminatory conduct by host states, but to expand the rights that investors have, and to do so in a way that shifts the risk of regulatory change from the investor to the government.

Authors: Lise Johnson is a senior legal research at the Vale Columbia Center on Sustainable International Investment. Oleksandr Volkov is an Associate with Egorov Puginsky Afanasiev & Partners Kiev.


[1] See European Commission, Incorrect Claims about Investor-State Dispute Settlement, Oct. 3, 2013, available at http://trade.EC.europa.EU/doclib/docs/2013/october/tradoc_151790.pdf.

[2] This essay draws from the following longer paper: Lise Johnson and Oleksandr Volkov, Investor-State Contracts, Host-State “Commitments” and the Myth of Stability in International Law, 24 Am. Rev. of Int’l Arb. 361 (2013)

[3] Técnicas Medioambientales Tecmed, S.A. v. United Mexican States, ICSID Case No. ARB(AF)/00/2, Award, May 29, 2003, para. 154.

[4] For example, this reading of the meaning of the fair and equitable treatment obligation as protecting against unforeseeable changes in the law was taken by Professor Rudolf Dolzer in his presentation at the VCC’s Spring Speaker Series on March 14, 2013, at Columbia University. See also, e.g., Suez, Sociedad General de Aguas de Barcelona S.A. and InterAgua Servicios Integrales del Agua S.A. v. Argentine Republic, ICSID Case No. ARB/03/17, Decision on Liability, July 30, 2010, para. 207; Total v. Argentina, ICSID Case No. ARB/04/1, Decision on Liability, Dec. 27, 2010, para. 122.

[5] Occidental Exploration and Production Company v. Ecuador, LCIA Case No. UN 3467, Final Award, July 1, 2004.

[6] See, e.g., AES v. Hungary, ICSID Case No. ARB/07/22, Award, Sept. 23, 2010, para. 9.3.34.

[7] EDFI v. Argentina, ICSID Case No. ARB/03/23 Award, June 11, 2012.

[8] Enron v. Argentina, ICSID Case No. ARB/01/3, Award, May 22, 2007.

[9] LG&E v. Argentina, ICSID Case No. ARB/02/1, Decision on Liability, Oct. 3, 2006.

[10] Occidental Exploration and Production Company v. Ecuador, LCIA Case No. UN 3467, Final Award, July 1, 2004.

[11] See also, e.g., M. Kinnear, “The Continuous Development of the Fair and Equitable Treatment Standard,” in A. Bjorklund, I. Laird, S. Ripinsky (eds.), INVESTMENT TREATY LAW, CURRENT ISSUES III (2009), at 228 (“The weight of authority suggests that an undertaking or promise need not be directed specifically to the investor and that reliance on publicly announced representations or well known market conditions is a sufficient foundation for investor expectations.”).

[12] See, e.g., Kardassopoulos v. Georgia, ICSID Case No. ARB/05/18, Award, March 3, 2010.

[13] These cases are reviewed in detail in Lise Johnson & Oleksandr Volkov, Investor-State Contracts, Host-State “Commitments” and the Myth of Stability in International Law, 24 Am. Rev. of Int’l Arb. 361 (2013).

[14] See, e.g., Home Telephone & Telegraph v. Los Angeles, 211 U.S. 265 (1908).

[15] See Johnson & Volkov at pp. 400-401; Fed. Crop Ins. Corp. v. Merrill, 332 U.S. 380, 384 (1947). Even where actual authority is established, estoppel claims against the government still place a heavy burden on the plaintiff to succeed. The general rule under U.S. federal law is that for equitable estoppel to apply against the government, a plaintiff must establish the basic elements of an estoppel claim, and show that the government engaged in affirmative and egregious misconduct. See, e.g., Sanz v. U.S. Sec. Ins. Co., 328 F.3d 1314 (11th Cir. 2003). Courts have also said that the equitable estoppel doctrine will only be applied against the government if doing so is necessary to avoid a serious injustice that would outweigh the damages to the public interest. See, e.g., Bolt v. United States, 944 F.2d 603, 609 (9th Cir. 1991). U.S. states similarly restrict estoppel claims against the government. See, e.g., Greece Town Mall, L.P. v. New York, 964 N.Y.S.2d 277 (April 25, 2013) (addressing the issue under New York state law); DRFP, LLC v. Republica Bolivariana de Venezuela, Case No. 2:04-CV-00793 (S.D. Ohio, May 14, 2013) (addressing the issue in a case against Venezuela addressing a contract governed by the law of Ohio).

[16] See, e.g., Suess v. United States, 535 F.3d 1348, 1362 (Fed. Cir. 2008).

[17] Alan R. Burch, Purchasing the Right to Govern: Winstar and the Need to Reconceptualize the Law of Regulatory Agreements, 88 Ky. L.J. 245, 248 (2000).

[18] Covington v. Kentucky, 173 U.S. 231, 238-239 (1899).

[19] Id. at 238-39.

[20] Alan I. Saltman, The Government’s Liability for Actions of its Agents that Are Not Specifically Authorized: The Continuing Influence of Merrill and Richmond, 32 Public Contract L.J. 775, 796 (2003).

[21] Cf. Glamis Gold Ltd. v. United States, NAFTA/UNCITRAL Ad hoc, Award, para. 800-02 (June 8, 2009) (finding no “specific inducements” the repudiation of which could potentially give rise to a breach of NAFTA Article 1105); para. 22 & n.24 (noting that although it viewed a repudiation of specific assurance as potentially giving rise to liability under the NAFTA, it would take no position on the “type or nature of repudiation measures that would be necessary to violate international obligations”).

[22] Enron Corp. v. Argentina, supra note 7, para. 137.

[23] See EDFI v. Argentina, supra note 6, para. 960.

[24] See, e.g., Bridge Proprietors v. Hoboken Co., 68 U.S. (1 Wall.) 116 (1883); Rogers Park Water v. Fergus, 180 U.S. 624 (1901); Century Exploration New Orleans, LLC v. United States, 110 Fed. Cl. 148, 172 (Fed. Cl. 2013) (“[N]othing in plaintiffs’ lease can be read to provide static treatment for their activities in perpetuity.”).

[25] See, e.g., Occidental Petroleum Corp. v. Ecuador, ICSID Case No. ARB/06/11, Award, Oct. 5, 2012, pp. 185-221, 308.

[26] While our research has focused on US law, preliminary analysis of the laws of other countries (e.g., India, Japan, and Canada) appears that other jurisdictions also take a view that is narrower than investment tribunals’ regarding the existence of purported promises by governments to restrict their abilities to take future regulatory action.

[27] See, e.g., Quantum Entertainment Ltd. v. US Dept. of Interior, 714 F.3d 1338 (Ct. APP. DC. 2013) (upholding finding that contract was “null and void” at the time it was entered into and that subsequent enforcement of that contract was impermissible retroactive change in the legal consequences of the deal).

[28] This can be easily seen by comparing states’ positions in briefs and non-disputing party filings with tribunals’ decisions. Briefs are regularly disclosed by the parties to the NAFTA and CAFTA.

[29] Some treaties have a mechanism making clear that if states agree to and issue interpretations reflecting their understanding of the agreement, those interpretations will be binding on tribunals. A well-known example of this is Article 1131 of the NAFTA.

[30] Eli Lilly and Co. v. Canada, Notice of Arbitration, Sept. 12, 2013, para. 82-84 (filed by Covington & Burling LLP and Gowling Lafleur Henderson LLP).

[31]  Guaracachi v. Bolivia, Claimants’ Post-Hearing Brief, May 31, 2013, para.116 (filed by Freshfields Bruckhaus Deringer).