A Proposal for Reforming the Calculation of Damages in Investment Treaty Arbitration
The Significance of Causation in Investment Treaty Arbitration
Causation is a core component of responsibility, whether legal or moral. Blame will usually only be placed on a person when their conduct has a causal influence on the consequence in question. This basic idea has found its way into investment treaty arbitration. When an arbitral tribunal is considering whether the state must pay compensation to the investor, the key determination is whether the state’s conduct had a causal influence on the investor’s loss. Unless this causal relationship is proven, the investor’s claim for compensation will be futile.
For this reason, what it means for a state’s conduct to “cause” an investor’s loss is of paramount importance. The meaning of “cause” will be determined by the test of causation that is used. In investment treaty arbitration, the most commonly used test is called “but for causation” or “sine qua non causation.” Applying this test, the question is: if x (“antecedent”) had not occurred, would y (“consequence”) have occurred?
The purpose of this blog post is to explore what “x” should mean in investment treaty arbitration. As detailed below, there are two possible definitions. Regarding the decision of which definition should prevail, the (financial) stakes are particularly high: while one definition will deliver more compensation to investors, the other will have the opposite effect.
What Does “X” Equal?
The first possible definition is that x equals “the state’s conduct.” This is the definition that most arbitral tribunals use (see, for example, Masdar v Spain (paras 549–552) and Vivendi v Argentina (para 8.2.7)). Practically speaking, if this definition is adopted, an arbitral tribunal tests for causation by asking, “If the state’s conduct had not occurred, would the investor’s loss have occurred?” The state’s conduct is that conduct that has been found to be in breach of the applicable investment treaty. For example, if the state’s enactment of a new regulation breaches the standard on fair and equitable treatment in the applicable investment treaty, then that act of enacting such regulation will be the “state’s conduct.” Accordingly, for such a case, the test of causation will ask, “If the state had not enacted the regulation in question, would the investor have suffered the loss that it did?”
A second possible definition is that x equals “the wrongful aspect of the state’s conduct.” In contrast to the first possible definition, this definition has found favour with only a few arbitral tribunals. One apparent reason for this lack of enthusiasm is a certain skepticism about whether conduct can have a non-wrongful aspect and a wrongful aspect. A tort law-inspired example can help alleviate this skepticism. The situation is (unfortunately) all too common: a car driver hits a pedestrian, which results in the pedestrian suffering a broken leg. The speed limit for the car driver was 30 km/h, although the car driver was driving 40 km/h at the time of the collision. What was the wrongful aspect of the car driver’s conduct? The fact that they were driving 10 km/h over the speed limit. If the goal is to test whether this wrongful aspect of the car driver’s conduct is causal, the question becomes: if the car driver was driving at 30 km/h, would the pedestrian have suffered a broken leg? By imagining a world where the wrongful aspect of the car driver’s conduct is not present, then it is possible to determine if it was necessary for the occurrence of the consequence.
Is a state’s conduct also divisible between wrongful aspects and non-wrongful aspects? The answer depends on which investment treatment standard the state is in breach of. Importantly, if the state is accused of breaching the FET standard, this division will be possible. Investment treaty arbitrations where the state has breached this standard by lowering a subsidy that it pays to the investor offer good examples. Suppose that an investor formerly received a subsidy payment of 100 ducats per month. Upon reduction of this subsidy payment, the investor received 60 ducats per month, which an arbitral tribunal subsequently determined was an excessive reduction and hence amounted to a breach of the FET standard. Logically, if the reduction was excessive, then there must have been a point between 100 ducats and 60 ducats where it was not excessive. For illustrative purposes, let us suppose that that figure is 85 ducats. In this case, the causal question would be: if the state had reduced the subsidy payment to 85 ducats, would the investor have suffered the loss that it did?
“X Equals State’s Conduct” or “X Equals Wrongful Aspect of State’s Conduct”?
In summary, there are two possible definitions of the antecedent in the test for causation, one of which is the “state’s conduct,” with the other being the “wrongful aspect of the state’s conduct.” Which one is to be preferred?
Readers will note that if the definition of the “wrongful aspect of the state’s conduct” is preferred, then the practical outcome will be that states pay less compensation to investors in the future. This practical outcome will manifest itself in two ways.
First, there will be fewer findings of causation. To exemplify this point, consider the question: if the state had reduced the subsidy payment to 85 ducats, would the investor have suffered the loss that it did? Suppose that the investor in question is financially unhealthy, meaning that it would have become insolvent in a world where the subsidy payment was only reduced to 85 ducats. That investor cannot establish causation. The same investor might, however, establish causation if the antecedent is the state’s conduct. In this case, the task is to imagine a world where the state does not reduce the subsidy payment at all, and then ask: in that world, does the investor still go insolvent? Probably not, in which case the state has caused this loss.
Even in cases where the antecedent is the wrongful aspect of the state’s conduct and there is causation, states will usually pay less compensation to investors compared to the situation if the antecedent is the state’s conduct, which counts as the second manifestation of the phenomenon of states paying less compensation. To illustrate this, suppose in this case that the investor is financially healthy, and it does not go insolvent, but the reduction of the subsidy payment down to 60 ducats does cause a financial loss of 1,000,000 ducats. According to most arbitral tribunals’ interpretation of the rule on full reparation, the state must pay compensation equalling 1,000,000 ducats to the investor. Some arbitral tribunals (see, for example, Kruck v Spain (para 354)) have : if the state may reduce a subsidy payment without breaching the FET standard, then it should not pay compensation on the part of the investor’s loss caused by any such “permitted reduction.” Following this logic, the investor is owed 625,000 ducats, assuming that the state was permitted to reduce the subsidy payment to 85 ducats without breaching the FET standard. The investor receives full reparation, but only for that part of its loss caused by the aspect of the state’s conduct that renders such conduct in breach of its obligation.
This logical coherence counts as a major reason for preferring the “wrongful aspect of the state’s conduct” over the “state’s conduct” as the antecedent when testing for causation in investment treaty arbitration. It is illogical for arbitral tribunals to insist that states can take measures that adversely affect the value of investments without breaching the FET standard, yet when states go too far and end up breaching this standard, then perform a switch by ruling that they have to pay for all losses incurred by investors, including those losses that they were lawfully permitted to inflict. If states’ compensation was limited to those losses that it unlawfully (or wrongfully) inflicted, then another advantage comes into view, specifically a potential thawing of the regulatory chill that investment treaties sometimes induce. To illustrate this potential, consider a situation where a state is considering the enactment of a regulation that it knows will devalue certain foreign investments in its territory. Aware that those losses might provoke investment treaty claims, it carefully designs this new regulation to make it compliant with the FET standard. Law, however, is not an exact science, meaning that there is always a risk that this new regulation will be found to be unlawful. Despite the care that has been taken to make this new regulation compliant, it is determined that the magnitude of the risk is too much, meaning that this new regulation is put on ice. But would this decision be different if the same state was told that any compensation owing to investors will be limited to the losses caused by the extent to which the new regulation goes too far? Clearly, the decision calculus relating to the enactment of this new regulation becomes different because the magnitude of the risk facing the state is much reduced.
The Way Forward for Adopting the “Wrongful Aspect of the State’s Conduct” as the Antecedent
But how can this proposal for testing causation with reference to the wrongful aspect of the state’s conduct be compatible with the principle of full reparation? This principle derives from the Chorzów Factory Case. There, Poland’s wrongful conduct was an expropriation, which is fundamentally different from a breach of the FET standard. As regards expropriation, the division of the state’s conduct into a non-wrongful aspect and a wrongful aspect is impossible—there is nothing “non-wrongful” about unlawfully taking another’s asset, meaning that all of the state’s conduct is wrongful, with the result that the state has to pay compensation on all of the investor’s losses. As various cases involving the application of the FET standard have shown, there is a line at which a state’s conduct becomes unlawful, which then brings it into breach. The goal is to add this nuance to the principle of full reparation. That does not involve a dramatic change but only a small addition to the effect that investors must receive full reparation for all of their losses caused by the wrongful aspect of the state’s conduct.
The question now remains, “How can that change be effected?” The issue of calculating damages should come on the agenda at the 49th session of UNCITRAL Working Group III. As the proposal to test causation with reference to the wrongful aspect of the state’s conduct was put forward as a potential reform at the 46th session, it will be one option for states to choose. If they want to make this calculation logically coherent and thaw out the regulatory chill that investment treaties sometimes cause, then they should get behind this proposal.
Authors
Martin Jarrett is Senior Research Fellow, Max Planck Institute for Comparative Public Law and International Law.