First appearing in the 1970s, denial of benefits (DoB) clauses have proliferated, became more sophisticated, and evolved significantly and even beyond recognition as in the 2017 Colombian Model BIT. This post discusses such evolution and provides a taxonomy of the different versions of the clause.


Denial of Benefits Clauses in Outline

Traditionally, DoB clauses permit the Host State to exclude from treaty protection companies that formally have the nationality of the other Party, but are controlled or owned by nationals of a third State.

The notions of “control” and “ownership” are far from being uncontroversial and some treaties provide some clarification (Art. 21(3) BIT Japan Jordan). Sometimes DoB clauses combine nationality requirements with the absence of substantial business activities in the territory of the Host State (see, for example, Art. 17(1) Energy Charter Treaty).

Tribunals have not been entirely coherent in dealing with DoB clauses, often due to their different or vague wording. They have treated them as matter of jurisdiction (i.e. Ulysseas v. Ecuador; Guarachi v. Bolivia), or merits (i.e. Yukos v. Russian Federation; Ascom v. Kazakhstan; Bridgestone v. Panama). Furthermore, they have taken different views as to whether a denial of benefits can be exercised at any time (i.e. Ulysseas v. Ecuador; Guarachi v. Bolivia), or only before the institution of arbitral proceedings (i.e. Plama v. Bulgaria; Ascom v. Kazakhstan).

In recent years, investment treaties have extended the scope of DoB clauses to situations of corporate restructuring clearly intended to gain access to treaty protection (i.e. Art. 20, 2015 Indian Model BIT). The clause follows the original logic behind DoB clauses by allowing the Host State to neutralise aggressive treaty shopping. It is consistent with recent arbitral awards that have considered such practice as an abuse of rights (i.e. Philip Morris v. Australia).

Several other treaties require control or ownership by nationals of third States in combination with other situations such as

  • absence of diplomatic relations with the third State (Art. 21.1(a) BIT Israel Japan). The application of such clauses tends not to be problematic as the absence of diplomatic relations can be identified objectively.
  • absence of normal economic relations with the third State (Art. XII(a) BIT US Bahrain). The expression “normal economic relations” is rather vague. It has been expressly associated with unilateral economic sanctions and situations such as those existing in 1998 between the US and Cuba or Libya (Message to the US Congress on the ratification of the BIT US Mozambique (1998), Art. XII). Presumably, the clause also covers mandatory measures adopted by the UN Security Council as they are meant to upset “normal economic relations”. From this perspective, 103 of the UN Charter comes into play.
  • existence of measures with respect to third States, or even persons of third States (i.e. Art. 20, BIT Israel Japan), that prohibit transactions with the enterprise, or would be circumvented in case of application of treaty to such enterprise (Art. 14.14(2), USMCA). This category largely overlaps with the criterion of “normal economic relations”, as it relates to unilateral countermeasures and collective sanctions. It seems even broader as it may cover measures adopted directly against investors.
  • existence of measures related to the maintenance of peace and security (i.e. Art. 8.16(b)(i) CETA). The expression evokes the powers of the UN Security Council and again must be read in conjunction with Art. 103 of the UN Charter. The inclusion in some treaties of the protection of human rights within those measures remains rather obscure (see Art. 8.13, EU Japan).

These clauses inject some political content into DoB clauses. Yet, they do not normally provide for any procedural guarantees. It may be argued that these clauses are not entirely self-judging and that arbitral tribunals have at least some role in determining the existence of any of the situations referred to in the clause.


Italian Model BIT

Article 18 of the Italian Model BIT (on file with authors) incorporates quite a sophisticated DoB, which combines some of the categories included in the classification above. It requires ownership or control by nationals of a third state, cumulatively with (1) absence of normal economic relations with the third State or existence of measures related to the maintenance of international peace and security adopted against such State; or (2) “a proportionate reaction” by the Host State to “the serious deterioration of the political situation in the other country with respect to the rule of law, democracy and human rights” in the third State.

Paragraph 1 essentially combines the situations referred to in the taxonomy under (c) and (d) (see also Art. 13, BLEU Model BIT 2019; Art. 8.16 CETA). The language used in Art. 18(1) suggests that the measures related to international peace and security may be adopted unilaterally, or within the collective security system.

Paragraph 2 constitutes a novelty as the investor becomes the direct target of the unilateral response to the alleged violations committed by the third State. Its content and implications, however, raise concern. The situation that may trigger such response appears to be difficult to define in legal terms. Moreover, the treaty contains no procedural guarantees in favour of the investor, who is inevitably exposed to the unilateral political decision of the Host State. It is true that the investor may challenge the invocation of Art. 18 before an arbitral tribunal, but the level of deference such tribunal will exercise is difficult to anticipate.


Colombian Model BIT

The DoB clause contained in the 2017 Colombia Model BIT offers an interesting blend of traditional and new elements. The clause applies to enterprises of the other Party controlled or owned by nationals of third States if the Host State has no diplomatic relations with the third State, or maintains measures that prohibit transactions with the enterprise, or would be circumvented in case of application of treaty to such enterprise – paragraph (a); as well as to enterprises that do not have substantial business in the Host State – paragraph (c).

The scope of the clause has been further extended to two new categories. Under paragraph (b), the Host State may invoke the clause when the enterprise is controlled or owned by nationals of a third State and shareholders submit a treaty claim without a written authorization of the enterprise. The clause is meant to prevent parallel proceedings and any compensation granted to shareholders shall constitute the final compensation to the enterprise.

Paragraphs (d) and (e) are even more intriguing and innovative. Paragraph (d) allows the Host State to deprive the investor of treaty protection where the investor is an enterprise that has:

  • committed serious human rights violations;
  • sponsored persons or organisations sentenced for serious human rights violations or violations of humanitarian law, or sponsored internationally-listed terrorist organisations;
  • caused serious environmental damage;
  • committed serious tax and fiscal fraud;
  • committed acts of corruption;
  • caused grave violations of labour laws;
  • engaged in money laundering activities.

Importantly, paragraph (d) is triggered by a judicial or administrative decision certifying the commission of wrongdoings. From this perspective, a precedent can be found in the 2006 ECOWAS Supplementary Act. Art. 18(1) deprives the investor found in breach of its obligations concerning corruption of the right to initiate any dispute settlement process under the treaty.

Furthermore, under paragraph (e), the DoB clause applies when a local court has found the investor or the executives of an enterprise responsible for the violation of criminal laws.

With paragraph (d) and (e), the clause takes a completely different connotation, disconnected from the traditional requirements of control or ownership. It assumes the function of a sanction, which directly targets the enterprise for its wrongful practices.

Unlike the Italian Model BIT, the Colombian Model BIT provides a detailed procedure for the application of the clause. The Host State must promptly inform the investor and the Home State of its intent to invoke the DoB clause. All pending proceedings are suspended for 90 days and terminated in the absence of any objection by the Home State. Otherwise, and with exception of Paragraph 1 (a), the Bilateral Investment Council will try to reach a friendly solution within 6 months, during which proceedings are further suspended. If the Council intervention is not successful, the effectiveness of the DoB is to be settled by the dispute settlement mechanism invoked by the investor (be that a local court or tribunal). The procedure appears to provide a solid protection against abusive application of the clause.


Concluding remarks

DoB clauses were designed as technical clauses giving the Host State the possibility of limiting treaty protection to genuine investors of the other Party. They have evolved significantly in different directions, with the introduction of politically charged situations connected with the conduct of third States, such as economic sanctions or maintenance of peace and security.

With the Colombian Model BIT, the clause performs an additional new function as the Host State may deprive the enterprise of the treaty protection in response to a comprehensive catalogue of wrongful acts committed by the enterprise itself. Ownership and control become less dispositive. While the procedure set in the treaty seems to adequately protect the investor against abuses, the clause ultimately offers the Host State an important tool to recalibrate its relationship with investors.


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.