Disentangling Legal Stability from Legitimate Expectations: Towards Greater Deference to Regulatory Changes in Renewable Energy Transition Policies in Investment Arbitration

Abstract

The obligation of stability generally requires host States to maintain a relatively stable regulatory framework to mitigate political risks facing foreign investments. It has played a significant role in international investment tribunals’ review of host States’ renewable energy transition policies. This paper critically reviews tribunals’ interpretation of the obligation with a particular focus on the Spanish cases involving renewable energy incentive schemes. It canvasses the two ‘dimensions’ adopted by investment tribunals in the interpretation of stability, namely the protection of legitimate expectations and States’ right to regulate for public purposes. Examining the contents of the two dimensions separately, this paper argues that legal stability should be disentangled from the notion of legitimate expectations and be assessed through the reasonableness of regulatory changes per se. It further argues that an intrusive interpretation of legal stability lacks legal and institutional bases; instead, more deferential standards should be adopted in the review of renewable energy transition policies.

1. Introduction

It has long been controversial to what extent regulatory stability should be guaranteed for the purpose of foreign investment protection. The obligation of stability generally requires host states to maintain a relatively stable legislative and regulatory framework and refrain from making drastic changes that adversely affect investors’ interests.¹ In Investor–State dispute settlement (ISDS) practice, claims relating to stability may arise from two circumstances, namely breach of contract and general regulatory changes.² In the former circumstance, a host State explicitly undertakes in its contract with investors that the governing laws and regulations will remain unchanged for a period of time (which is known as ‘stabilization clauses’); consequently, if the host State breaches the undertaking, the investor may claim treaty violation through umbrella clauses or the notion of legitimate expectations.³ Stabilization clauses are widely used in international energy investments and have played a critical role in mitigating political risks of investments in energy sectors such as oil and gas.⁴

This article focuses on the second circumstance, that is, the general treaty-based obligation of legal stability, under which regulatory changes by a host State may still be found to violate treaty obligations even without the existence of stabilization clauses. Investment tribunals generally consider the treaty-based obligation of stability to be subsumed in the fair and equitable treatment (FET) standard, despite the fact that investment treaties typically do not refer to the term ‘stability’ in FET clauses. One notable exception is the Energy Charter Treaty (ECT), which mentions the terms ‘stable’ and ‘fair and equitable treatment’ hand-in-hand. Consequently, it is generally less contested that the obligation of stability exists in ECT. The disputes concerning renewable energy policies examined in this article are raised under the ECT; therefore, this article will not engage in the debate over whether the FET inherently includes the requirement of legal stability.

There is an emerging understanding among commentators and tribunals that stability is not an absolute obligation – it does not mean freezing of laws. Rather, it must accommodate reasonable regulatory changes by host States.⁵ Nevertheless, investment tribunals’ decisions regarding how to delineate the acceptable margin of changes are far from consistent.⁶ The uncertainties brought by the obligation present a significant challenge to states’ implementation of domestic policies. Renewable energy transition (RET) policies, that is, policies that aim to facilitate the gradual transition from traditional fossil fuel to renewable energy-based production and consumption modes, are particularly prone to arbitration claims on the basis of legal stability. In the power sector, RET policies typically involve the employment of policy tools such as feed-in tariffs and premium payments that provide financial incentives to electricity producers.⁷ The determination of the degree and form of incentives is highly complex, involving consideration of a broad range of stakeholders and collaboration with different governmental departments. It also needs to consider the development of technology and market conditions that are constantly changing. In a word, RET policies inherently display volatility and require periodic adjustments throughout their implementation phase.

The unstable nature of RET policies has exposed states to significant litigation risks: in the past decade, several countries were sued by foreign investors in front of ISDS tribunals for violation of treaty obligations,⁸ leading to more than 80 ISDS cases in total.⁹ A typical example is Spain, which acted as the respondent state in more than 50 cases and in the vast majority of these cases, investment tribunals found the country's modification to the renewable energy remuneration regimes to have violated the obligation of stability and FET in the Energy Charter Treaty. There are also disputes arising under bilateral investment treaties, for example, the Hong Kong – Japan Bilateral Investment Treaty (BIT) and the Argentina–Spain BIT.¹⁰ The massive number of ISDS cases spurred states’ dissatisfaction with the legal regime protecting foreign investments in the energy sector and as Section 4 below will discuss, the European Commission has initiated a coordinated withdrawal from the ECT. It is therefore important at this time point to systematically and critically review the notion of the obligation of ‘stability’ in the context of RET and discuss the avenues of treaty reform.

This paper critically analyzes investment tribunals’ interpretation of the obligation of stability in ISDS disputes. It begins with the Spanish cases as they offer a straightforward example of how investment tribunals interpret the same obligation inconsistently regardless of mostly identical facts. A close examination of awards shows that investment tribunals’ interpretations of stability mainly revolves around two dimensions, namely the protection of legitimate expectations and States’ right to regulate for public purposes. The term ‘dimension’ here thus refers to the analytical framework, or perspective, adopted by tribunals to unfold the notion of stability. Particularly, investment tribunals employed various techniques to ‘balance’ the protection of legitimate expectations against regulatory rights, leading to inconsistent interpretations of the obligation. Against this backdrop, this paper probes into the legal bases and contents of the two dimensions. It shows that they have different sources and represent heterogeneous approaches to the obligation of stability: the dimension of legitimate expectations gauges the degree of regulatory changes by their deviation from investors’ expectations; and the dimension of regulatory rights presupposes states’ inherent right to regulate and examines – against standards of substantive review – whether the regulatory changes are reasonable and proportionate.

Examining the two dimensions separately – rather than treating them as ‘factors’ to be balanced against each other in one single analytical framework – not only elucidates the notion of stability but also reveals the problems of some tribunals’ interpretive approaches. Particularly, regarding the dimension of legitimate expectations, this paper argues that the linkage between legitimate expectations and legal stability originates from the earlier ISDS jurisprudence that was underdeveloped and overreaching; thus, the interpretation of stability should be detached from the notion of legitimate expectations. As to the dimension of regulation for public purposes, tribunals are ill-equipped to employ fully fledged proportionality analysis and ‘second guess’ states’ policy choices, especially given the polycentricity of energy policies. These all lead to the same conclusion: an intrusive interpretation of stability lacks legal and/or institutional bases in the context of investment arbitration, and greater deference should be granted to host states regarding renewable energy policies. These arguments also have important implications for delineating the notion of stability: it should be understood as a requirement for host states to maintain a relatively stable regulatory environment, primarily assessed – with high degrees of deference – through the reasonableness and proportionality of regulatory changes per se, rather than being tied to investors’ legitimate expectations.

The remaining Sections unfold as follows: Section 1 briefly introduces the features of RET policies and the background of related ISDS disputes; Section 2 reviews how tribunals’ interpretation of stability has evolved in the ISDS jurisprudence; Section 3 zooms in on the cases against Spain, and critically canvasses the two dimensions through which investment tribunals interpret the obligation of stability; finally, Section 4 discusses possible avenues for States to reinforce their regulatory autonomy regarding RET policies in future international law-making.

2. Background

Renewable energy transition, or clean energy transition, refers to the global trend to replace fossil fuel-based energy sources in power generation and consumption with zero-carbon energy sources such as solar, wind, hydropower, biomass, and geothermal.¹¹ It has played an important role in the global efforts to combat climate change since the conclusion of the 1997 Kyoto Protocol.¹² It also contributes to mitigating domestic energy security risks by reducing a State's reliance on the import of coal, oil, and gas.¹³ To boost RET, governments have implemented various financial and tax incentive schemes for electricity producers using renewable energy sources. A prominent one is Feed-in-Tariff (FiT), which has been implemented in over 75 countries¹⁴ and typically offers ‘long-term contracts that guarantee a price to be paid to a producer of a pre-determined source of electricity per kWh fed into the electricity grid’.¹⁵ FiT aims to ensure the profitability of eligible power plants given the capital-intensive nature of the sector (especially the massive amount of initial investment).

The incentive schemes prove effective in attracting investments in renewables.¹⁶ However, after years of implementation, some States found it necessary to water down or even repeal the remuneration schemes. The reasons are manifold. To start with, the global levelized cost of electricity, especially that generated through solar and wind, has dropped dramatically in the last decade.¹⁷ This gives rise to the question of whether it is still necessary to provide excessive subsidies to relevant investors. In addition, the rapid increase of investments in renewable energy plants imposes significant financial burdens on some governments, and in some countries causes an increase in electricity prices for consumers. Some countries, for example China, also encountered technical issues: due to insufficient power grid transmission capacity, some of the electricity generated from renewable energy sources cannot be delivered, which causes a waste of power as well as the corresponding financial input.¹⁸

The changes in the remuneration schemes led to a decrease in investors’ revenues and consequently gave rise to a large number of ISDS claims. Up to now, as mentioned above, Spain alone is facing around 50 ISDS cases as a result of its modification to the incentive schemes for electricity producers using renewable energy sources. In addition, the Czech Republic, Romania, Italy, Germany, Ukraine, Bulgaria, Japan, and Argentina also face ISDS cases for similar reasons.¹⁹ The disputes concern a variety of measures, including, for example, the implementation of a tax or solar levy that requires PV investors to pay a percentage of revenue from FiT, repeal of the income tax exemption for renewable energy producers, adjustments to methods to calculate remunerations, repeal of the FiT scheme, and reduction of ‘green certificates’ issued to renewable energy electricity producers.²⁰

The vast majority of the claims are brought under the Energy Charter Treaty, and a key ground relied upon by the investors is Article 10(1), which requires host States to ‘encourage and create stable, equitable, favourable and transparent conditions’ and ‘accord at all times to Investments of Investors of other Contracting Parties fair and equitable treatment’.²¹ The investors allege that the host State's changes to the remuneration regimes violate the obligation of FET and stability. In arbitral practice, investment tribunals have developed different approaches to parse the content and sources of the obligations. Section 3 below will examine the evolution of relevant jurisprudence.

3. The Evolution of the Stability Requirement in ISDS Jurisprudence

Overall, jurisprudence relating to the issue of legal stability is developing towards stricter and contextual interpretations of the term. Notably, except for the ECT, IIAs generally do not make explicit commitments on stability in FET clauses. Nevertheless, as the subsections below will show, tribunals repeatedly assert that the obligation of stability is subsumed in the FET standard.²²

3.1 The Broad Interpretation

In some earlier cases, the requirement of regulatory stability was found by tribunals to be associated with the protection of investors’ legitimate expectations. The Tribunal in Metalclad decided that Mexico violated FET because it ‘failed to ensure a transparent and predictable framework for Metalclad's business planning and investment’.²³ The Tribunal highlighted that, according to NAFTA Article 102, transparency is a key objective of the agreement, and thus the host State bears the obligation to avoid uncertainty in regulatory regimes. Moreover, the investor was entitled to rely on the federal government's representation that it could continue the construction of the landfill. Therefore, the Municipality's later refusal to issue the construction permit breached FET.

Similarly, in the later case Tecmed, the Tribunal held that FET under the Spain–Mexico BIT protects the ‘basic expectation’ taken into account by the investor when making the investment, and the investor expected the host State to act in a consistent manner with regards to policy goals and regulatory practice.²⁴ This interpretation of FET exerted a profound influence on the development of jurisprudence relating to legitimate expectations. After Tecmed, it has been a common practice for tribunals to presume that the notion of legal stability is closely related to investors’ legitimate expectations.²⁵ There are also tribunals contending that stability is inherent in the notion of FET because the aims of BITs are to protect legal certainty and foster foreign investment.²⁶

The Tribunal in Occidental v. Ecuador referred to another ground that it believed to justify the link between stability and FET, that is, the Preamble of the applicable treaty (i.e., the US–Ecuador BIT), which states that FET ‘is desirable in order to maintain a stable framework for investment and maximum effective utilization of economic resources’.²⁷ This justification was adopted in a number of later cases against Argentina, where the country made a series of radical changes to regulatory frameworks in response to its economic crisis.²⁸ A notable exception is the Tribunal in Continental, which viewed stability as a ‘precondition’ of the object of promoting investment flow, rather than the treaty's object per se. In the view of the Tribunal, it would be ‘unconscionable’ for a State to promise not to change its laws; as such, treating stability as a treaty object would be contrary to an effective interpretation of the term.²⁹

The context of the ECT is notably different from the above-mentioned BITs in that it prescribes the requirement to ‘encourage and create stable, equitable, favourable, and transparent conditions’ and the FET obligation side by side.³⁰ Although tribunals repeatedly emphasize that this does not mean that stability is a stand-alone or absolute requirement in ECT, they consider the juxtaposition an important textual basis for their interpretation that stability is subsumed in FET.³¹ The Tribunal in Petrobart, for example, concluded that the paragraph ‘in its entirety is intended to ensure a fair and equitable treatment of investments’.³²

3.2 Balancing Stability against other Factors

Saluka is a landmark case that seeks to balance the protection of legitimate expectations against the host State's regulatory interests. It stresses that the content of legitimate expectations cannot be solely determined by investors’ subjective motivations and considerations, and that investors should expect the host State's bona fide conduct that is ‘reasonably justifiable by public policies’ and that does not ‘manifestly violate the requirements of consistency, transparency, even-handedness and nondiscrimination’.³³ Siding with Saluka's approach, the Tribunal in Cargill v. Poland accentuated that ‘[t]he protection of the investors’ expectations has its limits’.³⁴

The Tribunal in Parkerings conducted a more contextual analysis of whether the investor could have any legitimate expectations of legal stability. It examined whether the host State made any explicit or implicit representations at the time of the investment, the circumstances surrounding the conclusion of the investment agreement, as well as whether the investor exercised due diligence.³⁵ It found that, although the host State made various modifications to its laws, the investor should have anticipated the risk given that the country was in transition from being part of the Soviet Union to a candidate for the European Union membership.³⁶

In line with the development of jurisprudence towards contextual analysis, two factors have crystallized as key considerations of tribunals to parse the notion of stability under FET, namely the existence of specific commitments and the reasonableness of regulatory changes. As the Tribunal in El Paso summarized, an investor's legitimate expectations ‘can only be examined by having due regard to the general proposition that the State should not unreasonably modify the legal framework or modify it in contradiction with a specific commitment not to do so’.³⁷ The Tribunal further defined that the commitment is specific when it is made directly to the investor or is in the form of a statement whose ‘precise object was to give a real guarantee of stability to the investor’.³⁸ It emphasized that a general legislative or political statement itself cannot be regarded as a specific commitment, otherwise it may be tantamount to requiring the host State to immobilize its legal order.³⁹ Likewise, the Tribunal in Philip Morris v. Uruguay stressed that ‘[p]rovisions of general legislation applicable to a plurality of persons or of category of persons, do not create legitimate expectations that there will be no change in the law’.⁴⁰ By contrast, the Tribunal in Antaris and Göde found that several statements by regulatory authorities of the Czech Republic (e.g., bill to Parliament, speech by the former Ministry of Environment in a newspaper, and government reports) – which depict stability as a key goal of the legal regime – qualify as commitments that give rise to legitimate expectations.⁴¹

As to the issue of reasonableness, tribunals generally underscore the proportionality of regulatory changes. As the tribunal in Blusun famously held,

In the absence of a specific commitment, the state has no obligation to grant subsidies such as feed-in tariffs, or to maintain them unchanged once granted. But if they are lawfully granted, and if it becomes necessary to modify them, this should be done in a manner which is not disproportionate to the aim of the legislative amendment, and should have due regard to the reasonable reliance interests of recipients who may have committed substantial resources on the basis of the earlier regime.⁴²

Overall, later tribunals are generally aware that stability is not an absolute obligation and must be balanced against other factors such as states’ regulatory demands and the nature of commitments made. However, tribunals’ interpretation of the meaning and importance of the factors vary among each other: as the review of cases above shows, some tribunals adopt the restrictive understanding that legitimate expectations can only arise from specific commitments, while others consider that general legislative statements can also give rise to legitimate expectations.

Moreover, investment tribunals have adopted different analytical frameworks to balance these factors. Some revolve around the issue of what the investor could have expected, and subsume the consideration of states’ regulatory demand in the delineation of the scope of legitimate expectations (e.g., whether the regulatory changes are so fundamental and unproportionate that investors could not foresee at the time of investment).⁴³ Some tribunals acknowledge that states enjoy the autonomy to make regulatory changes. Nevertheless, they stress that the changes must be within an acceptable ‘margin’ and/or take due regard to the reasonable reliance interests of investors, otherwise it would violate investors’ legitimate expectations as well as the obligation of stability.⁴⁴ There are also tribunals examining the breach of legitimate expectations and the proportionality of changes separately.⁴⁵ The heterogeneity of analytical frameworks exacerbates the uncertainties in investment arbitration concerning the obligation of stability. Particularly, intertwining the analysis of legitimate expectations and regulatory rights fuels confusion regarding the obligation of stability and opens the door for arbitrary interpretation of the obligation. Section 4 will elaborate on this argument further.

3.3 The Restrictive Approach

It is worth noting that the development of jurisprudence under NAFTA is different from that under other treaties due to the influence of the Notes of Interpretation issued by the NAFTA Free Trade Commission.⁴⁶ The Notes limit the scope of FET to customary international law minimum standard of treatment and exerted a profound influence on some tribunals’ interpretation of FET under NAFTA.⁴⁷ In Cargill v. Mexico, for example, the Tribunal ruled that no evidence suggested that there is a requirement for stability and predictability in NAFTA or customary international law.⁴⁸ Similarly, in Mobil and Murphy, the Tribunal stressed that FET does not require the host State to maintain a stable environment for investment, and it only protects investors from regulatory changes that are ‘arbitrary or grossly unfair or discriminatory, or otherwise inconsistent with the customary international law standard’.⁴⁹ This interpretation raises the bar for FET violation as the disputed measures not only need to be improper but also need to reach a certain degree of severity (e.g., ‘grossly’ unfair). However, the jurisprudence under NAFTA is hardly consistent despite the existence of the Notes. For example, the Tribunal in Eli Lilly acknowledges that a claimant may succeed on its claims under NAFTA Article 1105 if it demonstrates ‘a dramatic change in the law’.⁵⁰

To sum up, ISDS jurisprudence relating to the notion of stability is evolving towards more balanced and contextual analyses. The genesis of the obligation of stability is closely related to an overly broad interpretation of investors’ legitimate expectations by some earlier investment tribunals. Later tribunals sought to construct a more balanced framework by lending weight to States’ right to regulatory changes, the existence of specific commitments and investors’ due diligence. However, the specific approaches adopted by tribunals to delineate scope and content of the obligation of stability are hardly consistent, and such inconsistencies visibly intensify in the cases against Spain where the applicable law is the same (i.e., the ECT) and the regulatory measures in dispute are largely identical.

4. The Three Dimensions of Stability: Taking Stock of the Cases against Spain

This section first introduces the regulatory changes made by Spain regarding remuneration schemes for electricity producers using renewable energy sources. It then critically examines investment tribunals’ interpretation of stability in these cases by canvassing the legal bases and contents of the two dimensions.

4.1 Spain's Renewable Energy Regimes for Electricity Production

Spain's Law 54/1997 introduced a special regime that treated electricity producers primarily using renewable energy sources differently from those using traditional energy sources.⁵¹ Under the special regime, registered clean energy producers enjoy the right, inter alia, to incorporate excess energy into the electricity system.⁵² In the following years, the government issued a series of decrees that provided more explanations for the calculation of remuneration under the special regime.⁵³ Particularly, Royal Decree 436/2004 specifies that producers under the special regime may either choose to transfer the electricity to electricity distribution companies with a sale price in the form of a ‘regulated tariff’ (i.e., feed-in-tariff, a guaranteed above-market price⁵⁴) or sell the electricity in the market with a market-based price plus an incentive or premium.⁵⁵ The later implemented Royal Decree 661/2007, which repealed Royal Decree 436/2004, basically kept this two-fold remuneration scheme and provided more details on the calculation of tariffs and premiums for different categories of energy producers.⁵⁶ It also specified that the tariffs and premiums for certain categories of producers (including those using solar energy, wind, hydropower, or biomass) would be updated annually, taking into account the increase in the Consumer Price Index and other factors.⁵⁷

The financial incentives greatly boosted investments in electricity production based on renewable energy sources. Nevertheless, at the same time, Spain suffered from outstanding electricity tariff deficits due to the rapid increase in regulated costs, and a significant portion of these costs was constituted by the financial incentives under the special regime.⁵⁸ As a result, from 2010 onward, Spain implemented a series of decrees to dilute the incentives. For example, the Royal Decree 1565/2010 shortened the application of the solar energy FiT from 30 years to 25 years;⁵⁹ the Royal Decree-Law 14/2010 imposed the ‘equivalent reference hours of operation’ – that is, the maximum number of yearly hours to enjoy FiT – to photovoltaic solar technology installations;⁶⁰ and the Law 15/2012 levied a new tax on electric energy production.⁶¹

In 2013, Spain made the most fundamental changes to the remuneration scheme for renewable energy-based electricity producers. The Royal Decree-Law 9/2013 and Law 24/2013 abandoned the special regime and implemented a new regime that offers a market-based ‘specific remuneration’ that is said to allow producers to obtain adequate profitability.⁶² Factors to be considered for the calculation of the remuneration include revenues from the sale of the energy in the market, operating costs, and the value of the initial investment of the installation.⁶³ The later implemented Royal Decree 413/2014 listed various parameters to calculate the remuneration, and Order IET/1045/2014 added more details on calculation methods (for example, setting the reasonable return to be 7.398%).⁶⁴ The new regime caused a significant drop in revenue received by renewable energy investors and consequently triggered a large number of investment arbitrations against the government.

4.2 The Obligation of Stability versus the Right to Regulation

To date, approximately 50 cases have been initiated against Spain's energy reforms. The first tribunal to issue an award on merits was Charanne. In the opinion of the Tribunal, the FET obligation under ECT Article 10(1) ‘is included in the more general obligation to create stable, equitable, favourable, and transparent conditions’.⁶⁵ Nevertheless, the Tribunal stressed that, lacking specific commitments, the investor could not expect the legal frameworks to be frozen, as long as the regulatory changes ‘are not capricious or unnecessary and do not amount to suddenly and unpredictably eliminate the essential characteristics of the existing regulatory framework’.⁶⁶ It consequently found that the 2010 rules did not violate the obligation of stability. Charanne is a notable exception in this series of disputes because the 2013 and 2014 rules – which fundamentally changed the regulatory framework and were found by later tribunals to violate the obligation of stability – were excluded from the Tribunal's jurisdiction.⁶⁷

In later disputes where the 2013 and 2014 rules are addressed, tribunals’ interpretations are hardly consistent with regard to the status and content of the obligation of stability under ECT. Overall, tribunals have canvassed the obligation of stability from two dimensions, namely investors’ expectations and states’ regulatory rights. As Section 3.2 above demonstrated, tribunals frequently analyze them under a single framework; particularly, they discuss whether the regulatory changes have exceeded an acceptable margin in the analysis of legitimacy expectations. Sections below argue that these dimensions imply different approaches to the interpretation of the obligation of stability. Probing into their legal bases and contents reveals the problems of some existing interpretive approaches.

4.2.1 Investors’ expectations

Under this dimension, whether regulatory changes have exceeded the acceptable margin is gauged by their deviation from the investor's legitimate expectations. Essentially, it treats stability as an obligation that protects investors’ property rights, and its analysis relies heavily upon the long-standing albeit controversial doctrine of legitimate expectations in the ISDS jurisprudence. Accordingly, the analytical scheme revolves around two themes, namely what the investor could have expected at the time of investment, and whether the regulatory regimes have been ‘radically altered’ in a way that ‘deprive investors who invested in reliance on those regimes of their investment's value’.⁶⁸

With regard to the issue of what the investors can expect, as discussed in Section 2 above, tribunals have adopted different approaches. The broader interpretation is that investors can legitimately expect an overall stable regulatory regime. In the Spanish cases, tribunals siding with this approach considered that the FiT regime as promised in Spanish laws (especially RD 661/2007) and government statements was the very reason the foreign investors made their investments; therefore, the investors should reasonably expect the FiT regime to be maintained for a fixed length of time.⁶⁹ Accordingly, they found that the modifications to the FiT regime violated the obligation of legal stability because the changes to the remuneration scheme are fundamental and caused a sharp fall in investors’ revenues from the levels anticipated under the 2007 rules.⁷⁰

In contrast, some tribunals adopt a more restrictive approach, that is, legitimate expectations can only arise from more specific and individualized commitments.⁷¹ As the majority in PV Investors v. Spain noted, ‘expectations which are purported to be founded on general legislation have been treated with caution in a number of recent decisions’, given that the protection of legitimate expectations must be balanced against ‘State's sovereign prerogative to adapt the regulatory framework to changed circumstances’.⁷² Accordingly, the majority concluded that ‘the regulatory framework, including RD 661/2007, did not provide for a stabilization guarantee according to which investors would enjoy’; instead, what the investors could legitimately expect is a reasonable return on their investments.⁷³ In Masdar, the Tribunal relied upon the letters sent to the investor by the Ministry of Industry, Tourism and Business (guaranteeing FiT for the plants’ ‘operational lifetime’) to find the existence of legitimate expectations.⁷⁴

Another important factor in the assessment of legitimate expectations is whether the investors conducted due diligence. The majority in Stadtwerke v. Spain attached significant weight to this factor: it highlighted elements in Spanish laws and court decisions that indicate that the remuneration regime will not remain unchanged; it further concluded that ‘a prudent investor, having conducted an appropriate due diligence, would not have reasonably formed an expectation of a legally stable income stream for the life of [its plant]’.⁷⁵ Likewise, in Isolux Netherlands, the Tribunal emphasized that an investor cannot reasonably expect regulatory stability if his information allowed him to foresee the unfavorable evolution of the regulatory framework.⁷⁶

Admittedly, the requirement of the protection of legitimate expectations and the obligation of stability impose similar constraints on states’ exercise of regulatory autonomy – States do not enjoy unfettered discretion to implement regulatory changes. Therefore, it is unsurprising that some tribunals treat the two obligations as intertwined and discuss them together.⁷⁷ Nevertheless, the understanding that legal stability forms part of an investor's legitimate expectations is problematic from several aspects.

To start with, the linkage between the obligation of stability and legitimate expectations lacks legal basis from the perspective of treaty interpretation.⁷⁸ This interpretation is rooted in the earlier ISDS jurisprudence (e.g., Tecmed as mentioned above) which made the general statement that investors should legitimately expect a consistent and predictable regulatory environment. Subsequent tribunals simply followed this interpretation in their assessment of stability without sufficient engagement with the rule of treaty interpretation as provided by the Vienna Convention on the Law of Treaties (VCLT).⁷⁹ Although some referred to the preamble or the objective of investment treaties to highlight the importance of a stable regulatory framework in the protection of foreign investment, no plain treaty text suggests that States intended to treat stability as part of investor's expectations to be protected (not to mention that the vast majority of IIAs do not refer to the term legitimate expectations in FET clauses). Even in the context of ECT, as the Tribunal in RWE Innogy correctly highlights, if legitimate expectations can be understood to be able to generate from domestic law (rather than specific commitments), ‘the FET standard would in practical terms start to approximate an overarching stabilization clause, elevating each change in a domestic legal regime to a source of potential breach of international law’, which could not be the intention of ECT Contracting Parties.⁸⁰ Overall, the overreaching linkage between legitimate expectations and stability appears to be just an invention of investment tribunals that lacks justification.

Secondly, subsuming the obligation of stability in the notion of legitimate expectations undermines the widely shared understanding that legitimate expectations should be analyzed in concreto.⁸¹ Whether investors can reasonably expect a stable regulatory framework should be decided case by case, taking into account factors such as the existence of specific commitments, host states’ past practice, and the nature of the public matter concerned. Presuming that investors can expect legal stability imposes an unreasonable burden on host States. In the Spanish cases, as Professor Christian Tomuschat highlighted in his Partial Dissenting Opinion, since the implementation of RDL 661/2007, the government had taken various measures to adjust the FiT regime, and the Preamble of RD 6/2009 explicitly pointed out the problem of tariff deficit facing the government.⁸² Relying solely on legislation as the basis of legitimate expectations risks overlooking factors that indicate the possibility of changes in the regulatory framework. Furthermore as explained in the introduction section, RET policies are inherently fluid and complex, thus a reasonable investor should anticipate constant regulatory changes in the design of reimbursement schemes.⁸³

For the above-discussed reasons, it seems better to detach the interpretation of stability from the notion of legitimate expectations. As Professor Philippe Sands emphasized in his dissenting opinion in RENERGY, the stability obligation in the ECT ‘has a different legal foundation than the obligation to respect investors’ legitimate expectations’: the latter is based on investors’ expectations arising from host States’ commitments, while the former is based on the text of ECT Article 10(1) and ‘operates independently of any such expectations’.⁸⁴Accordingly, he argued that the proper approach to interpreting the ECT stability obligation is to ‘treat it as an additional element of the FET standard, and one which is distinct from the established doctrine of legitimate expectations’.⁸⁵

Moreover, the subject of the stability obligation is public regulations, and whether the regulations have violated the obligation should be primarily determined by the characteristics of the measures per se rather than upon an external benchmark (e.g. investors’ expectations). In this sense, the second dimension, which will be discussed in the next sub-section, provides a more proper analytical framework to interpret stability.

4.2.2 Regulation for Public Purposes

The second dimension of the stability obligation concerns States’ right to regulate for public purposes. Under this dimension, the starting point of analysis is that States inherently enjoy the power to modify domestic laws; in other words, they enjoy a margin of appreciation. Nevertheless, the exercise of regulatory power must be reasonable and proportional.⁸⁶ Tribunals have adopted different standards of review: some applied a more deferential standard, assessing the good faith and rationality of the regulatory measures; others opted for a stricter proportionality analysis.⁸⁷

In the Spanish cases, a number of tribunals employed proportionality analysis. A typical proportionality analysis involves consideration of the legitimacy of the objective, the suitability and necessity of the measure to achieve the objective, and proportionality stricto sensu (i.e., balancing between the impact of the measure on private rights and the importance of the public interest).⁸⁸ In RWE, the Tribunal considered Spain's 2013 measures suitable and necessary given the gravity of the tariff deficit problem.⁸⁹ Nevertheless, it found that the measures imposed an excessive and disproportionate burden on the claimant as some of its plants are estimated to receive significantly lower actual returns than the reasonable rate set by the government.⁹⁰ The Tribunal in PV Investors, in contrast, found no elements of disproportionateness in Spain's disputed measures after reviewing the possible alternative measures.⁹¹

There are also tribunals attempting to enumerate the factors that are to be considered and balanced in the review of regulatory changes. The Tribunal in RENERGY, for example, referred to seven factors, namely the magnitude of the change, economic impact on investors, abruptness of change, external circumstances (e.g., changes triggered by circumstances beyond the government's control), public interests involved, prior legislative practice, and the host State's assurances regarding the stability of the regulatory framework.⁹² These factors concern the rights of both the host State and the investor, and thus appear to constitute a more balanced analytical framework. Nevertheless, in its conclusion, the Tribunal attached significant weight to the fact that the changes to the remuneration regime were fundamental, and consequently found that they were impropriate despite their legitimate purposes.⁹³

Therefore, although approaches like proportionality analysis appear to balance the protection of investment against States’ regulatory autonomy, in practice they do not necessarily lead to more deferential findings, given that tribunals may still attach greater weight to the impact of the measures on the investors as opposed to the public purposes. Moreover, it has long been controversial whether it is appropriate for investment tribunals to carry out proportionality analysis.⁹⁴ As the Tribunal in RREEF carefully noted:

the Respondent enjoys a margin of appreciation in conducting its economic policy; therefore, [the Tribunal] will not substitute its own views either on the appropriateness of the measures at stake or on the characterization of the situation which prompted them; in particular, the Tribunal will abstain to take any position on the issue of the existence of other or more appropriate possible measures to face this situation.⁹⁵

Investment tribunals are arguably ill-equipped to engage in proportionality analysis given their lack of embeddedness within the host state polity.⁹⁶ The institutional settings of investment arbitration are also significantly different from that of European law or domestic laws where proportionality analysis is commonly used. For example, as pointed out by Sornarajah, ‘[t]he European Court can be invoked only after domestic remedies have been exhausted. No such requirement exists in investment arbitration’.⁹⁷ Furthermore, applying proportionality analysis in the context of investment arbitration entails the risk of ‘unwarranted judicial law-making’ if the balancing undermines States’ original intent.⁹⁸

Moreover, investment tribunals lack the institutional capacity to tackle highly technical and complex issues such as designing remuneration schemes for RET.⁹⁹ RET policies fall exactly into the scope of what Fuller calls a ‘polycentric issue’ which he analogizes to a ‘many-centered’ spider web where ‘[a] pull on one strand will distribute tensions after a complicated pattern throughout the web as a whole’ and ‘each crossing of strands is a distinct centre for distributing tensions’.¹⁰⁰ The design of RET policies involves the reconciliation of various factors, for example, the affordability of electricity at the consumer's end, the development of technology relating to solar PV materials, the capacity of the electrical grid system, energy supply, and the government's fiscal burdens.¹⁰¹ Investment arbitration, where a small number of ad hoc arbitrators make decisions on the basis of arguments and evidence presented by the disputing parties, lacks informational and institutional settings for addressing such a highly polycentric issue. In this sense, tribunals are not in a better position than States to decide the optimal measures to achieve a public purpose.¹⁰²

Therefore, adopting a more deferential standard of review (instead of carrying out fully fledged proportionality analysis) and eschewing second guessing the host State's policy choice appears to be more proper in assessing the limit of the margin of appreciation.¹⁰³ This does not mean leaving the host State's conduct unscrutinized. Good faith review, which examines whether the host States’ measures were carried out in good faith and reasonably¹⁰⁴ – guarantees a minimum level of protection to foreign investors. Regulatory measures may also be examined against fundamental principles of the rule of law, for example, the non-retroactive application of law.¹⁰⁵

In addition, as Schill pointed out, investment tribunals may ‘compensate the broader degree of deference granted to domestic institutions with a stricter approach to assessing procedural propriety’.¹⁰⁶ Fair and equitable treatment standard entails several requirements on the procedural aspects pertaining to regulatory changes, for example, non-discrimination, transparency, and administrative due process.¹⁰⁷ These grounds have also been debated – albeit less extensively – in the Spanish cases under FET. For example, in RENERGY, the investor alleged that Spain's law-making process was ‘highly irregular’ and thus fell short of transparency and due process;¹⁰⁸ in FREIF, the investor claimed that Spain breached the duty of transparency and good faith because it did not consult the industry and inform the investor before enacting the new regime;¹⁰⁹ in Sevilla, the investors argued that the new regime, particularly its calculation of the incentives, is opaque and unpredictable, thus lacking transparency.¹¹⁰ These procedure-based claims are generally not supported by investment tribunals either for lack of evidence or that the reforms were within Spain's regulatory autonomy.¹¹¹ Additionally, some tribunals highlighted that the regulatory reform was announced sufficiently in advance and that the government did consult the industry.¹¹²

To sum up, this Section critically examines tribunals’ analysis of legitimate expectations and regulatory rights in their interpretation of stability. Analysing and ‘balancing’ them in one single framework not only obfuscates their different legal bases but also – to borrow Waibel's succinct words – ‘risks a drift into a treacherous slippery slope of double standards and inconsistency’,¹¹³ as tribunals may allocate different weights to these factors according to their own preferences. The arbitral awards are frequently accompanied by a separate dissenting opinion; Spain also actively sought to annul some of the awards, arguing that, inter alia, the tribunals failed to state reasons in respect of the analysis of legitimate expectations, which thus constituted a ground of annulments according to Article 52(1)(e) of the ICSID Convention.¹¹⁴ By ‘disentangling’ the two dimensions and probing into their argumentative bases separately, this Section argues that, first, the linkage between the notion of stability and legitimate expectations is overreaching; second, regarding substantive review, tribunals are ill-equipped to conduct a fully fledged review of regulatory policies. The two arguments, albeit unfolded under separate frameworks, point to the same conclusion, that is, deference should, to a greater extent, be guaranteed in the interpretation and application of stability in these RET cases.

Teleologically, a more deferential interpretation of stability is consistent with the goal of the ECT which seeks to reach a balance between the promotion of the development of an efficient energy market for the flow of investments on the one hand and the respect for States’ sovereignty over energy resources on the other.¹¹⁵ A broad interpretation and application of stability, for example, finding a violation of the obligation on the sole basis that the degree of changes is fundamental, imposes an unrealistic burden on the host State. As the Tribunal in Stadtwerke correctly stated, Spain's RET policies exemplify what the sociologist Robert Merton called ‘the law of unintended consequences’ which frequently triggers repercussions beyond the government's control.¹¹⁶

5. The Way Forward: Towards Greater Deference to RET Regulatory Changes

Discussion in previous sessions focuses on the ECT because the vast majority of cases against renewable energy incentive schemes were brought under the treaty. Probing into the legal bases of the notion of stability in the ECT has profound implications for understanding FET in general: as explained above, the FET clause in the ECT is notably different from that in other IIAs in that it juxtaposes the terms ‘stable’ and ‘fair and equitable treatment’. Such an explicit reference to stability is generally absent in the FET clauses of other IIAs, while some investment tribunals still presume that investors are entitled to expect a stable legal framework. If – following the analysis under the first dimension – the notion of stability is to be detached from the notion of legitimate expectations, it may well be contested whether, or to what extent, the obligation of stability should be guaranteed in the context of non-ECT treaties in the absence of stabilization clauses.

The series of ISDS cases against host States’ adjustments to RET policies also showcases the dilemma facing states in devising legal frameworks to achieve the goal of limiting temperature increase in the Paris Agreement.¹¹⁷ On the one hand, a stable regulatory framework and a decent level of legal protection, especially that guaranteed in investment treaties, boost investors’ confidence in investing in renewable energy projects; as such, it is unwise to entirely abandon the existing international investment protection regime. On the other hand, as explained above, the design and implementation of RET policies are inherently unstable and it is necessary for states to preserve regulatory space.

One way to mitigate this tension is to create more leeway for regulatory autonomy in investment treaties. Revising the investment protection provisions to ensure host States’ right to regulate is also a core pillar of the ECT modernization project.¹¹⁸ In the EU's proposal, it suggested adding a ‘regulatory measures’ article in the investment protection section which reaffirms host States’ right to regulate for public purposes such as environmental, social, or consumer protection.¹¹⁹ Particularly, the proposed article emphasizes that ‘a Contracting Party's decision not to issue, renew or maintain a subsidy … in the absence of any specific commitment under law or contract to issue, renew, or maintain that subsidy … shall not constitute a breach of the provisions of Part III of the Treaty’.¹²⁰ This seems to be specifically designed in response to those RET-related ISDS cases. Moreover, in the EU's proposal, it deleted the reference to ‘stable, equitable, favourable and transparent conditions’ in Article 10 and listed the elements of FET (which is consistent with its recent treaty practice); it also adopted the term ‘specific representation’ when delineating legitimate expectations.¹²¹ These proposed modifications were largely adopted in the final Agreement in Principle concluded at the Ad hoc Energy Charter Conference meeting on 24 June 2022.¹²²

The revised ECT FET clause should be able to substantively narrow the scope of FET – especially legitimate expectations – and thus increase host States’ regulatory space.¹²³ Nevertheless, the ECT reform appears to be late and insufficient – the EU has embarked on a coordinated withdrawal from the treaty.¹²⁴ Such a massive backlash indicates that the ECT has deviated too far from the States’ shared understandings and the sustainability of the system is at risk given the weight of the EU members among the ECT Contracting Parties. Since all the intra-EU bilateral investment treaties have been terminated in 2020 following the Achmea judgment,¹²⁵ withdrawal from the ECT means that intra-EU energy disputes will be settled by an amicable ‘facilitator’ or adjudicated by national courts.¹²⁶

Although the outlook for successful modernization of ECT is not promising, the proposed revisions discussed above shed light on the revision of other investment treaties. As reiterated above, existing investment agreements generally do not incorporate the obligation of stability in their FET clauses, while a significant number of them recognize the importance of providing a ‘stable environment’ in their preambles. Based on the review of cases in previous sections of this paper, is it reasonable to anticipate that investment tribunals will continue interpreting legitimate expectations and FET broadly as including an obligation of legal stability. To mitigate this risk, apart from modifying treaty clauses, States might issue joint interpretations (which constitute subsequent agreements according to Article 31 of the Vienna Convention on the Law of Treaties¹²⁷) that limit the scope of legitimate expectations to those arising from specific commitments made by the host State governments. States may also design a new multilateral treaty to protect investments in renewable energy sectors. The new treaty should emphasize their right to regulation for environmental protection or other public purposes, narrow down the scope of FET (e.g., explicitly excluding the obligation of stability or the protection of legitimate expectations), or even exclude certain disputes from the jurisdiction of ISDS tribunals.

6. Conclusion

In conclusion, this article critically examined – with a specific focus on the recent cases against Spain's modifications to its RET remuneration schemes – the two dimensions of the obligation of stability through which investment tribunals scrutinize host States’ regulatory changes, namely investors’ expectations and regulation for public purposes. By probing into their legal bases separately, this article argues that the interpretation of stability should be detached from the notion of legitimate expectations and that more deferential standards should be adopted in the substantive review of RET-related regulatory changes. As such, an intrusive approach that interprets stability broadly lacks legal and institutional support in the context of ISDS. In addition, to further safeguard States’ RET regulatory autonomy, the existing investment treaty framework relating to energy regulation must be extensively modified to narrow the scope and degree of investment protection.