The Rise of Sustainability Provisions in International Investment Agreements Free

There is a lively debate about whether international investment agreements (IIAs)—the central instruments governing foreign direct investment—are compatible with sustainable development objectives. One critical argument is that investment treaty provisions limit governments’ policy space to achieve transitions toward sustainability and may lead to costly legal disputes. Against this background, governments, international organizations, and other stakeholders have begun to consider options for modernizing international investment law by including provisions related to sustainable development in the treaties. In this article, we examine to what extent and why states include sustainability references in IIAs. Our automated text analysis of 2,083 bilateral investment treaties and 374 other treaties with investment provisions shows that the number of sustainability references has risen sharply recently. Drawing on the literature on issue linkage in economic agreements, we test different explanations for the changes in the treaty language. We find that the level of democracy among treaty parties and the experience of being sued in the past are strong drivers of the described trend.

The relationship between foreign direct investment (FDI) and sustainable development is a widely debated topic today. There is widespread agreement that the transformation toward sustainable development can only succeed with the help of enormous sums of private capital (Mann 2013). At the same time, the United Nations Conference on Trade and Development (UNCTAD) noticed an annual investment gap of US$ 4 trillion to achieve the Sustainable Development Goals (SDGs) in developing countries (UNCTAD 2023b). Parts of the investment policy community believe that international investment agreements (IIAs), the central instruments governing FDI, can promote sustainable investments, such as renewable energy investments, by granting foreign investors legal protection of their property. The headlines, however, are dominated by concerns about the risks that IIAs pose for sustainable development measures.

In this view, private capital might primarily be placed into industries incompatible with sustainable development objectives, such as oil and gas projects, which account for large parts of global greenhouse gas emissions. There is furthermore a growing concern that IIAs constrain the national policy space and thus hinder regulations needed to achieve key goals associated with sustainable development (Johnson et al. 2019). Most importantly, most IIAs include investor–state dispute settlement (ISDS) provisions, which allow foreign investors to sue states in international arbitration. The mechanism is heavily criticized for allowing investors to claim compensation for government measures in the public interest. The critics’ concerns are not unfounded. In fact, the number of ISDS cases involving sustainable development issues is growing rapidly. According to a 2022 UNCTAD report, 175 ISDS cases, about 15 percent of all known cases, have been brought against environmental protection measures (UNCTAD 2022, 2). Policymakers have taken note of these cases. The governments of Denmark and New Zealand have admitted that the risk of being sued by investors has discouraged them from taking more ambitious climate action.¹

In 2012, then United Nations (UN) Secretary-General Ban Ki-moon stated, “It is necessary to strengthen the development dimension of international investment agreements, manage their complexity, and balance the rights and obligations of States and investors” (UNCTAD 2012, iii). The growing literature on IIAs (Elkins et al. 2006; Haftel and Thompson 2018; Thompson et al. 2019), however, has not yet studied the extent to which the treaties have changed to meet sustainable development objectives in light of the described challenges. We ask to what extent IIAs incorporate sustainable development references and what explains the varying levels of references across treaties.

To answer these research questions, we used methods of automated text analysis to analyze the language of 2,083 bilateral investment treaties (BITs) and 374 other international treaties with investment provisions (TIPs) that were concluded between 1955 and 2021. We compiled an original dictionary of terms covering relevant dimensions of sustainability and counted the number of sustainability references in each treaty. We analyzed the content of BITs, the most important devices that exclusively deal with investment protection. However, we also included TIPs in our analysis to take into account that investor protection is increasingly regulated in the context of broader agreements, such as the Comprehensive Economic and Trade Agreement (CETA) or the Comprehensive and Progressive Agreement for Trans-Pacific Partnership, that have become hotly debated in recent years.

We show that references to sustainability in IIAs have sharply increased, particularly in treaties signed in the last decade. Some states or country groups, including the European Union (EU) and Canada, have a leading role in this development, as their treaties tend to have exceptionally high numbers of sustainability references. Beyond these exploratory findings, we analyze factors influencing the number of sustainability references in IIAs. We find that democratic states tend to include more sustainability references and that involvement in ISDS proceedings leads to more references in subsequent treaties of the country that got sued.

Our findings contribute to two literatures. First, scholars of the investment treaty regime have shown that governments include new provisions and legal language that help them to protect their policy space and restrict the discretion of investment tribunals in new or renegotiated IIAs (Haftel and Thompson 2018; Spears 2010; Thompson et al. 2019). Our study contributes to this literature by focusing on the potential change toward more sustainable investment treaties in response to new demands for sustainable investments and the unintended consequences of international investment protection.

Second, our study contributes to a larger literature on issue linkage in international economic agreements (Milewicz et al. 2018). An important strand in this literature focuses on the so-called trade–environment nexus, where scholars have noticed a significant rise in environmental provisions in new trade agreements (Blümer et al. 2020; Jinnah and Lindsay 2016; Lechner and Spilker 2022; Morin et al. 2018). This rich literature on issue linkage in the trade regime stands in contrast to the other major field of global commerce, international investment, where such debates are only beginning to emerge. Our study advances the debate about the relationship between foreign investment and sustainable development by empirically demonstrating the evolution of sustainability standards in IIAs and thus widens the focus of the issue linkage literature.

An IIA is a legal instrument signed between two or more states to reciprocally facilitate and protect investments in the respective countries. Most of the existing more than 3,000 IIAs are BITs, but the number of free trade agreements (FTAs) containing BIT-like investment protection chapters is growing fast. The typical IIA stipulates a standard set of investment protection standards, including clauses protecting investors against uncompensated expropriation and provisions that require “fair and equitable treatment” by host governments. Probably the most important and controversial features in IIAs are ISDS provisions that provide foreign investors with a direct legal standing against states (Allee and Peinhardt 2010; Simmons 2014). If foreign investors feel that a government has breached an investment treaty, they can file a case against the state in international arbitration. The disputes are resolved by three arbitrators who are not tenured judges but usually lawyers appointed by the investor and the state to resolve the specific case. Arbitral awards are final and binding to the disputing parties. The remedy available in investment arbitration is compensation to be paid to the foreign investor. In many cases, investment tribunals have ordered states to pay spectacularly large amounts of compensation, showing that the commitment to IIAs can have significant financial consequences (Wellhausen 2016, 132–134).

A common view underlying the investment treaty regime is that the ratification of IIAs leads to increased investment flows. Accordingly, supporters of IIAs believe that modernized IIAs might also channel investments into projects contributing to sustainable development. Numerous studies have attempted to answer the question of whether IIAs have the expected positive effects on FDI flows. While some studies showed that IIAs can increase investment flows (Neumayer and Spess 2005; Tobin and Rose-Ackerman 2011), other studies found that IIAs have little to no effect (Brada et al. 2021; Hallward-Driemeier 2003; Yackee 2008). Given these ambiguous findings, the treaties’ capacity to promote sustainable investments seems uncertain. Notably, in this context, a recent study notes that IIAs play a minor role in renewable energy companies’ investment decisions (Mehranvar and Sasmal 2022).

In contrast, critical observers and, increasingly, government officials maintain that the broad set of protections offered to foreign investors would constrain the state’s room for action to realize measures targeting sustainable development. To understand the growing concerns about investment treaties and ISDS, it is helpful to look at some recent developments and cases. The wider literature on the investment treaty regime has discussed the possibility of reduced policy space as the risk of a “chilling effect,” meaning that governments might give up policies or refrain from taking action in response to an investor claim or in anticipation of it (Moehlecke and Wellhausen 2022). Several scholars and other commentators have discussed these risks in the context of climate action, stating that IIAs could hinder effective and necessary measures to combat climate change (Maljean-Dubois et al. 2022; Tienhaara 2018). Related to that last point, a recent study estimates that the transition from fossil fuels to renewables could cost US$ 340 billion if fossil fuel companies sue states in investment arbitration based on these measures (Tienhaara et al. 2022). The view that investment treaties pose obstacles to climate action is also mentioned in a 2022 Intergovernmental Panel on Climate Change report (IPCC 2022, 1593–94). Indeed, it is documented that fossil fuel companies have historically been the most frequent claimants in ISDS, although not all claims are directly related to environmental protection measures (Di Salvatore 2021; UNCTAD 2022). The ISDS case Rockhopper v. Italy attracted considerable public attention in this regard.² Rockhopper, a UK oil company, claimed compensation after Italy reintroduced a ban on oil and gas exploration within twelve miles of the Italian coastline, which was invoked to deny the company a license to exploit off-shore oil fields. An arbitration tribunal confirmed that Italy breached obligations under the Energy Charter Treaty (ECT) and ordered the government to pay Rockhopper € 190 million in compensation. At the same time, renewable energy investments have caused a number of costly ISDS cases. A wave of arbitrations against Spain stands out (Reynoso 2019). Foreign investors launched more than fifty cases, claiming compensation for the country’s energy reforms that included subsidy cuts for renewables. The following quote from Hamed El-Kady, a UNCTAD representative, shows that relevant stakeholders have realized the need for action: “[g]overnments and the international investment community should intensify their efforts to reform investment treaties in support of the energy transition and to minimize risks of expensive legal disputes” (UNCTAD 2023a).

There are more examples of conflicts between international investment law and sustainable development objectives. Widely noticed has been a legal dispute in which the tobacco company Philip Morris sued Australia for its plain packing laws, raising the question about the implications of investment law for public health measures.³ Many arbitrations in the mining sector involve measures of environmental protection or concerns of Indigenous people living in the region and affected by the projects (Puig 2019). Other arbitrations have had consequences for critical sectors like water supply and thus have strong links to basic human needs and human rights (Marisi 2022). The list could be continued.

While “old” investment treaties were mostly silent on social issues, focusing on creating favorable conditions for foreign investors, more and more new IIAs include some form of sustainability provisions to respond to the described challenges and enhance the legitimacy of IIAs.

The Brundtland Report famously defined the concept of sustainable development as development that “meets the needs of the present without compromising the ability of future generations to meet their own needs” (World Commission on Environment and Development 1987, 16). Several international conferences and declarations further affirmed the objective of sustainable development. The 1992 Rio Summit made sustainable development the guiding principle of international environmental and development policy. The Rio Declaration (United Nations [UN] 1992b) was ratified by 178 states and defined twenty-seven principles signifying the commitment of the international community to pursuing a sustainable development agenda. Crucially, the related Agenda 21 underpinned the importance of foreign investment to achieving these goals, especially in developing countries. It reads in part,

As a next major step, the UN member states adopted the eight Millennium Development Goals (MDGs) in September 2000. The MDGs are generally considered to reflect a broad international consensus concerning key development objectives, ranging from reducing extreme poverty and hunger and fighting the most severe diseases to ensuring environmental sustainability. After the target date of 2015, the SDGs replaced the MDGs. They defined new milestones for global development policy until 2030 that align economic with social and environmental objectives.

IIAs do not seem to be made for the new inclusive approach to development. It is argued that the investment treaty regime strongly adheres to a traditional development paradigm that prioritizes economic growth and prosperity. However, this practice is changing. In 2012, UNCTAD published its Investment Policy Framework for Sustainable Development. The framework proposed a menu of concrete policy options to assist policymakers in operationalizing sustainable development objectives in IIAs, ranging from adjusting existing provisions to make them more compatible with sustainable development and adding new provisions to better balance investor rights and responsibilities to introducing special and differential treatment clauses that reduce the classical obligations under IIAs for less developed states. UNCTAD published an updated version of its Investment Policy Framework in 2015 and constantly reports on the progress toward sustainable investment treaty making.

There is still little research on sustainability provisions in IIAs. A few studies review existing IIAs and map the different provisions in the treaties (e.g., Chi 2019; Gazzini 2014). More policy-orientated scholars have proposed a list of characteristics that could assist policymakers and other stakeholders in identifying “sustainable investments” (Sauvant and Mann 2019), and others have offered recommendations on how to improve IIAs in a way that they align with sustainable development objectives (Johnson et al. 2019). The existing empirical studies on sustainability provisions in IIAs have focused on certain types of provisions, such as environmental provisions (Su and Shen 2023) or anticorruption provisions (Yan 2020), or have a geographical focus. Analyzing 271 BITs concluded by the Association of Southeast Asian Nations (ASEAN), McLaughlin (2022) finds that references to sustainability are generally rare. However, the study also points to a positive trend of increasing use of sustainability references in the IIAs of some ASEAN member states.

The wider literature on the politics and impacts of IIAs is further instructive to advancing our understanding of the relationship between the investment treaty regime and sustainable development. Our study relates to existing research focusing on the design of IIAs. This scholarship investigates to what extent and why the content of investment treaties has changed. Scholars working on these questions take up debates about a backlash against international investment law and suggest that changes in treaty making appear as responses to growing legitimacy concerns and the unintended consequences resulting from the commitment to IIAs. From this perspective, new treaty practices can be interpreted as an attempt to better balance private and public interests in the investment regime. Recent empirical studies generally find a trend of more balanced and increasingly precise IIAs, confirming that many governments now seek to better protect their policy space and restrict the discretion of investment tribunals (Haftel and Thompson 2018; Manger and Peinhardt 2017; Thompson et al. 2019). In this view, including sustainability provisions is an attempt to protect the state’s regulatory space, necessary to achieve sustainable development objectives.

In short, the existing literature shows that sustainability issues are now part of the policy discourse and treaty making in international investment law. However, the extent to which references to sustainable development are incorporated into investment treaties remains unclear. Our article builds on previous research to provide an overview of the state of the use of sustainability provisions in the global network of IIAs and to analyze the mechanisms explaining how these provisions spread across treaties.

Including sustainability provisions in IIAs seems counterintuitive from an economic point of view. The main motivation of host countries to sign IIAs is to attract foreign investors (Elkins et al. 2006). By including sustainability provisions in IIAs, the contracting states, in theory, are trying to impose obligations on investors or control which investments are made on their territory. We would thus principally expect a decrease rather than an increase in investment flows if an IIA includes sustainability provisions. Why would states do that? We build on a rich literature on issue linkage to study the relationship between IIAs and sustainability provisions. When “linking” issues, states seek to achieve two (or more) goals with the same instrument. It is well known that trade agreements have become “deeper” and more complex, dealing with a range of nontrade issues (Baccini et al. 2015; Milewicz et al. 2018). Especially the spread of environmental provisions in trade agreements has been widely studied. This development is commonly explained by increasing awareness of environmental problems, growing public demands to protect the environment, and states’ international commitments (Morin et al. 2018). We argue that a similar development is ongoing in the investment treaty regime and discuss different explanations for states’ motivation to include sustainability provisions in IIAs.

We argue that two parallel developments are likely to push the inclusion of sustainability provisions in IIAs. First, as a normative principle, sustainability gains importance in various aspects of our social life and influences political debates and decision-making at all levels (local, national, international, transnational). Consequently, we can expect IIA negotiators to become increasingly aware of questions related to sustainable development. Such processes of norm diffusion are typically initiated by specific entrepreneurs or pioneering states, but once the issue has gained sufficient attention and has found enough supporters, the diffusion of the norm accelerates (Finnemore and Sikkink 1998).

At the same time, the investment treaty regime faces a legitimacy crisis (Bonnitcha et al. 2017, chap. 9; Dietz et al. 2019; Waibel et al. 2010). It is clear that more and more people seem unwilling to accept economic agreements that threaten noneconomic public interests. The result is growing protests against agreements like the Transatlantic Trade and Investment Partnership. Policymakers are now responding with measures to reconcile economic goals with broader interests. Thus integrating sustainability provisions in IIAs may reflect a larger trend of reorientation in the international investment policy community that is looking for ways to balance the objective of protecting foreign investments and the state’s right to regulate. Hence our main expectation is that the rise of sustainability as a norm guiding local, national, and international policymaking processes and the quest for greater legitimacy in the investment treaty regime lead to more sustainability references in IIAs over time:

Research on environmental provisions in trade agreements suggests that democracies push for balancing economic and noneconomic interests in international treaties (Morin et al. 2018). As Milewicz et al. (2018, 746) note, the trend of nontrade issues is, in many respects, motivated by the need “to appease domestic audiences and to avoid competition on regulatory standards.” Thus we can expect that electoral pressures and high levels of domestic regulation motivate democracies to strive for sustainability clauses in IIAs. On one hand, we assume that democratic states have a greater interest in negotiating sustainable IIAs that allow them to pursue an ambitious policy agenda domestically. Indeed, some work measuring the extent to which states achieve the SDGs suggests that democratic states perform better than other political regimes. For instance, the top ten countries in the Bertelsmann SDG Index are highly democratic.⁴ The SDG Index ranking is led by Sweden, Denmark, and Finland, which are countries that are consistently ranked highly in democracy rankings. Moreover, we assume that democracies are more responsive to the domestic demands of civil society groups and the electorate than are nondemocratic states. In international investment law, domestic groups, such as environmental nongovernmental organizations or labor unions, may push for stricter environmental rules or labor standards, thereby influencing negotiations. After the conclusion of a treaty, these domestic groups may also serve as agents in observing state compliance with these norms (Simmons 2009). Thus we propose the following hypothesis:

Next, we consider the effect of power. Previous research on IIAs showed that bargaining power is a key factor in explaining the diffusion of treaties, especially the variation in the content of treaties (Allee and Peinhardt 2010; Simmons 2014). We maintain that power asymmetries between the treaty partners may also affect how frequently contracting states incorporate sustainability aspects into their IIAs. Powerful states might use their bargaining power to push weaker states toward stronger rules for sustainability because they believe that advancing the sustainability agenda is normatively desirable. An alternative argument holds that rich, industrialized countries may seek to protect domestic industries that are threatened by competitors in the Global South. One strategy for protecting domestic companies would therefore be to force weaker countries in the Global South to adopt stricter environmental or other social measures via specific provisions in economic agreements (Bhagwati 2003; Lechner 2016). In this view, including sustainability provisions is primarily a form of protectionism. Some of the most powerful actors in the investment treaty regime, such as the EU or the United States, have included issues related to sustainability, such as references to the environment, health, or labor, in their recent treaties or model agreements. We would expect that these states have a vital interest that their rules and principles define the standard of modern IIAs. Weaker states are often left with the decision either to accept the conditions of the more powerful partner to gain economically or to forgo the potential benefits of the partnership if they are not willing to do so:

Existing research on the diffusion of environmental clauses in trade agreements has shown that states with high levels of domestic environmental protection tend to commit to more and stricter environmental provisions in international treaties (Morin et al. 2018). A first perspective emphasizes the low adaptation costs for states that already have high national standards (Milewicz et al. 2018). For these states, only minor direct consequences follow from the obligations under international law. They do not need to radically change their policies to bring them into compliance with specific treaty provisions since their national measures are already sufficient for compliance or even go beyond the obligations deriving from the treaty. We expect a similar development in international investment law. The adjustment costs for implementing more sustainable IIAs are lower for more developed states, as they are already more likely to align their policies with sustainability principles. Developed countries should worry less about sustainability provisions in international treaties than developing countries, which may prioritize rapid industrialization to realize economic growth. Besides, due to their high level of prosperity, highly developed countries can afford an ambitious sustainability agenda more easily than developing and emerging economies can. Thus we hypothesize the following:

Finally, our theory considers the effect of frequent exposure to ISDS. In the last two decades, the number of ISDS cases increased significantly, and many states experienced claims in which foreign companies challenged regulations related to sustainable development objectives. Previous studies suggest that states learn from the experience of costly ISDS claims and adjust their IIAs in response to these disputes (Haftel and Thompson 2018; Manger and Peinhardt 2017; Thompson et al. 2019). Building on this research, we expect states that have experienced multiple investment disputes to be more likely to integrate sustainability provisions when negotiating a new investment treaty to prevent future disputes:

UNCTAD’s IIA database lists 2,828 BITs and 451 TIPs.⁵ This massive number of existing treaties and the volume of text pose a challenge for researchers interested in the content and evolution of IIAs. We use automated text analysis methods that allow the “systematic analysis of large-scale text collections” (Grimmer and Stewart 2013, 268; see also Alschner et al. 2017).

We downloaded electronic copies of all BITs and TIPs available in English from UNCTAD’s database. Although our sample is not representative of the whole IIA network, it is sufficiently comprehensive and diverse to draw meaningful conclusions regarding the presence of sustainability standards.⁶ Our data set includes the full texts of 2,083 original BITs plus 374 TIPs. We compiled an original dictionary to measure our theoretical concept of sustainable IIAs. In addition to direct references to “sustainability” or “sustainable development,” we included related keywords that refer to important dimensions of sustainable development, such as the objectives of public health, environmental protection, labor rights, and good living conditions (see the  Appendix). In our analyses, we consider all these keywords to be analytically equivalent. The keywords were selected based on a review of the key provisions and policy options recommended by UNCTAD’s Investment Policy Framework for Sustainable Development (UNCTAD 2012) and a review of academic articles on sustainability standards in IIAs (Chi 2019; Gazzini 2014; Johnson et al. 2019). The dictionary was further refined after we scanned the relevant provisions of a few randomly selected IIAs that were signed after the publication of the UNCTAD report.

For our analysis, we used the individual IIA as the level of analysis and measured the frequency at which keywords included in our dictionary occurred. We believe that the absolute number of references is highly meaningful and superior to other methods. For example, one might consider alternatively measuring the “density” of sustainability references by relating the number of references to the length of the treaty text. However, we believe that this could be misleading. The text portion that constitutes sustainability references may decrease simply because of expanding text or new and more precise standards in other parts of the treaty. However, this alone says little about the importance of sustainability aspects for international investment policies. In other words, the value of individual sustainability provisions is unrelated to the length of the treaty text.

Our hypotheses suggest that several factors influence the number of sustainability references in IIAs: time, power asymmetries, regime type, development status, and the experience of investment disputes. The first independent variable, TIME, measures the evolutionary character of sustainability standards. With each additional year, we expect the number of sustainability references to increase. In our model, we used the year of signature instead of the year of ratification because some treaties that were signed but never ratified might nevertheless influence subsequent negotiations. Our second variable, DEMOCRACY, captures the average quality of democracy in the contracting states when the treaty is signed. We used the Varieties of Democracy Project’s Liberal Democracy Index as our data source (Coppedge et al. 2019). GDP DIFFERENCES is a continuous variable that measures power asymmetries between the contracting states. We calculated the difference by subtracting the total GDP (thousands in current US dollars) of the weaker country in the treaty from the GDP of the more prosperous country, using the year the treaty was concluded. For treaties with more than two parties, we calculated power asymmetries using the difference between the state with the highest and the state with the second highest GDP. We obtained the data from the World Bank’s database.⁷ We used the Human Development Index (HDI) to account for the DEVELOPMENT STATUS of the contracting states. Again, we calculated the mean HDI score of all treaty parties for this variable. Finally, we included the variable ISDS RESPONDENT in our model to test the effect of exposure to investment disputes. We calculated the sum of cases in which the contracting parties were involved as respondents. ISDS cases initiated after the conclusion of a specific IIA cannot influence the design of that treaty. Thus we used the year of treaty signing and only counted those cases initiated before the conclusion of the particular treaty. We used UNCTAD’s Investment Dispute Settlement Navigator⁸ to collect information on ISDS, noting that arbitrations can be kept confidential, so the actual number of cases is likely to be higher (Hafner-Burton et al. 2016).

The text analysis clearly shows that sustainability references in IIAs have sharply increased. Figure 1 provides an overview of this trend by showing the number of sustainability references in the preambles and main texts of BITs and TIPs per year. Looking first at the development of BITs, we see that references to sustainability aspects increased for the first time at the beginning of the 1990s. This first wave is best explained by the high number of BITs signed during that time. However, Figure 1 also indicates that the most remarkable increase occurred when the number of annually signed BITs had already declined. We generally observe a similar trend for TIPs: the number of sustainability references started to increase in the 1990s and has turned up since then. We find that references made in the treaty’s preamble cover a good portion of the total numbers. The preamble of the EU–Vietnam FTA⁹ is a good example illustrating the new treaty language. It reads in part,

Although not legally binding, the preamble indicates the spirit of the treaty and thus provides host states, foreign investors, and, most importantly, arbitrators guidance on how to interpret the legal provisions of the treaty (Chi 2019, 20). However, it is also evident that sustainability references are spread over the whole treaty and may thus also influence the substantive character of investment protection.

Figure 2 presents the rise of sustainability references in more detail by breaking down the trends in BITs and TIPs for different periods. About half of the total number of BITs and a considerable number of TIPs do not have any sustainability reference at all. These are mostly first-generation investment treaties that were signed between 1959 and 1990. Between 2001 and 2005, sustainability references slightly increased for the first time, and several outstanding treaties included significantly more references. Particularly noteworthy here are treaties concluded between the Belgium–Luxembourg Economic Union (BLEU) and partner countries, including Libya (2004), Mauritius (2005), Guatemala (2005), and Sudan (2005), for each of which thirty-eight references were counted. Since 2005, we observe a steady increase in sustainability references as a general trend in investment treaty making. The 2010s then mark the beginning of a new generation of BITs. We observe a significant increase in the median level of sustainability references for the first time, indicating that new treaty language diffused across BITs. The trend accelerated in the most recent BITs (2016–2020). Figure 2 further shows that the level of sustainability references in TIPs is not equally increasing as in the case of BITs, although it also shows that the number of some exceptionally progressive treaties is constantly increasing. It should be mentioned that trade agreements signed by the United Kingdom after Brexit frequently reference sustainability issues, making the United Kingdom a leader in promoting and codifying sustainability in international economic agreements.

Looking closer at individual treaties or countries, Table 1 shows that the most references to sustainability have been made in the 2016 Morocco–Nigeria BIT (60). An expert noted that the treaty “contains several innovative provisions that recalibrate the legal protection of the interests of all stakeholders and can be expected to enhance the chances for economically, socially and environmentally sustainable investments.”¹⁰ This qualitative assessment confirms the observation that emerges from our text analysis. Next comes the 2016 Chile–Hong Kong BIT (44) and several BITs signed by Canada and various partner countries. The role of Canada as a driving force in incorporating sustainability into the investment treaty regime is further evidenced by the fact that the CETA with the EU includes by far the most references to sustainability (950) in the TIP category (see Table 2). It is important to note that Canada released an updated model BIT in 2004 that specified the government’s priorities for negotiating new investment agreements (Newcombe 2005). Our analysis indicates that Canada has been quite successful in getting its ideas about sustainability issues accepted in subsequent negotiations with different partners.

Next, we look at the distribution of keywords in the treaties. Figure 3 shows the most frequently used keywords grouped into ten categories covering different aspects of sustainable development. It becomes clear that direct references to “sustainability,” “sustainable development,” and related terms are relatively rare in BITs, while they are more frequently used in TIPs. Beyond these direct mentions, many IIAs reference specific dimensions related to the concept of sustainable development, most important, labor, health, and environmental aspects, such as climate change or biodiversity, or include provisions that define the relationship between the regulatory autonomy of the state and investment protection. The latter dimension is particularly pronounced in BITs and likely indicates the spread of nonderogation clauses and provisions stipulating the state’s right to regulate. Nonderogation clauses can be used to preclude host states from lowering or relaxing their standards, for example, in environmental protection or labor rights, to become more attractive to foreign investors, and thus used as defensive measures to prevent a regulatory “race-to-the-bottom” undermining sustainable development (Chi 2019, 20). “Right to regulate” provisions address the concern of shrinking policy space that may hinder new regulation necessary for sustainability transformations. We find that existing IIAs rarely include references to corporate social responsibility and anticorruption measures, which confirms the findings of prior research. It is important to clarify that these provisions are mostly soft law instruments that do not result in investor obligations, meaning that their effectiveness hangs on voluntary compliance (Yan 2020). Furthermore, most IIAs were silent about human rights, which we consider an important legal foundation of the sustainable development agenda. Finally, regarding social development aspects, such as education and gender equality, we see a mixed pattern, with a relatively high level of references in TIPs but only marginal numbers in BITs.

Which factors drive the diffusion of sustainability references in IIAs? Table 3 presents the results of our multiple linear regression analysis (OLS), testing the impact of our independent variables on the number of sustainability references in IIAs. We ran the analysis for two subsamples. The first subsample includes only BITs signed between 1990 and 2021 (column 1), while the second subsample includes only TIPs signed in the same period (column 2). Table 3 also shows the effects for the total sample (column 3).

To recall, we expected in H₁ that the number of sustainability references would be influenced by the time the treaty was signed, as states learn from each other and the norm of sustainability diffuses gradually. Our analysis confirms that TIME positively affects the number of sustainability references. We find that additional years increase the number of references in BITs and TIPs, but the effect was statistically significant only in the subsample of BITs.

DEMOCRACY and ISDS RESPONDENT are strong predictors of the number of sustainability references in BITs and TIPs alike, and the direction of the effect was just as predicted in H₂ and H₅. Treaties concluded between democratic states contain more sustainability references than do treaties concluded between less democratic states. We also find that the number of investment disputes a country has experienced in the past drives the number of sustainability references. After getting sued, countries included more references to sustainable development in subsequent IIAs. The effects of both variables, the level of democracy and ISDS cases, on the BIT and TIP design were statistically significant. These findings, taken together, suggest that governments, to a considerable extent, pursue defensive intentions with sustainability provisions. They want to preserve policy space and are responsive to domestic pressures and legitimacy concerns.

GDP DIFFERENCE also has a positive and statistically significant effect on the number of sustainability references in BITs, meaning that power asymmetries between the contracting states increase the number of references, as expected by H₃. We interpret this finding as evidence that powerful states successfully use their bargaining power to rewrite the rules governing the investment regime. We must leave open whether the underlying interest is a progressive move to advance the sustainable development agenda or more of a protectionist posture in that it precludes weaker countries from unsustainable rise, for example, by seeking to attract fossil fuel investments or maintaining competitive advantages through low labor standards. Surprisingly, power asymmetries do not significantly affect the number of sustainability references in TIPs. One possible interpretation is that wealthy countries are not prepared to forgo potential trade benefits that they expect from these more comprehensive agreements for the sake of noneconomic values and therefore accept lower standards in their partner countries.

Finally, we find no support for H₄, that the DEVELOPMENT STATUS of the treaty parties influences the number of sustainability references. To recall, we theoretically expected that a higher level of development would lead to more sustainability provisions, but our analysis, surprisingly, showed lower numbers for these treaties. The effect, however, was not statistically significant. One possible explanation for this finding is the considerable influence that international organizations, such as UNCTAD, might have on developing countries’ treaty practices. UNCTAD is leading the debate about reforming IIAs in support of sustainable development objectives (UNCTAD 2012, 2015), assisting developing countries in implementing the reforms, and facilitating exchange by hosting events like the World Investment Forum. Policymakers from developing countries have appreciated UNCTAD’s assistance and guidance. For example, a Nigerian policymaker said, “UNCTAD’s Investment Policy Framework for Sustainable Development has been immensely beneficial in guiding our consideration of specific clauses for review.”¹¹ A Mongolian policymaker stated “that UNCTAD is the institution that is most suited to take the lead in devising such guidelines based on best practices in IIA reforms.”¹²

The central finding of our text analysis of 2,083 BITs and 374 TIPs is that the number of sustainability references in IIAs has significantly increased in the last decade. The process of modernizing investment treaties is ongoing: policy recommendations, such as those provided by UNCTAD, are being implemented in the treaty-making process. New treaties today incorporate legal provisions that regulate certain aspects of the investment–sustainability nexus, and they do so to an increasing extent. It seems that policymakers are slowly moving away from the purely economic paradigm of strong investment protection toward a broader development model that emphasizes the need for more regulation of economic relationships (Hindelang and Krajewski 2016; Schill et al. 2015). Beyond this general trend in the content of the treaties, our study provides further insights into bargaining dynamics in international treaty making. The changes in IIAs have accelerated at the time that advanced democracies realized the downsides of the investment regime. They have experienced a growing number of lawsuits and believe that their policy space is shrinking under IIAs. In line with previous research, our analysis showed that democratic states drive innovations in treaty making. This finding supports the argument that growing pressure from potential voters pushes governments to design economic agreements in a way that is compatible with wider public interests (Milewicz et al. 2018, 746). In this view, the rise of sustainability in IIAs shows governments’ efforts to enhance the legitimacy of the investment regime that has come under pressure from the wider public. Whether the inclusion of sustainability references will help regain public support for economic agreements remains to be seen.

Our results further indicate that the attempt to preserve regulatory space is a key motivation of states in treaty negotiations. We found an increase in sustainability references in IIAs into which states entered after being sued by an investor and interpret this response as a learning process in which host states recalculate the risks and benefits of IIAs and thus aim to draft more complete treaties. This empirical finding is in line with previous research on the redesign of investment treaties that generally showed that the involvement in ISDS influences the content of upcoming treaties (Haftel and Thompson 2018; Thompson et al. 2019).

In summary, we thus view the growing number of sustainability references in IIAs primarily as a defensive move by states reacting to domestic pressures and seeking to protect themselves against costly ISDS claims in the future. Despite a positive overall trend, the development is not even. Many old investment treaties that lack sustainability provisions of any kind remain in force, and not all new treaties contain strong sustainability provisions. As long as these treaties exist, they clearly constitute a risk for the transition toward a more sustainable economy. Apart from that, doubts remain about whether the new treaty language and provisions can bring real change and advance sustainable development. The decisive questions in this context are whether sustainable IIAs shield host states against investor claims and whether arbitrators consider sustainability aspects in ISDS proceedings. There are some indications that the modernization of investment treaties does little to change arbitration decisions. In the ISDS case Eco Oro v. Colombia, for example, the arbitrators held that a general exception clause in the investment treaty on which the claim was based did not preclude the payment of compensation.¹³ We are not blind to these challenges but underline that inserting sustainability provisions into IIAs is a nascent development, evidenced by our empirical study. The verdict on this development is still out. Future research must assess the real effects of the new treaty practice.

In our view, the departure from standard investment treaties is irreversible, but there is certainly room for improvement in the emerging generation of investment treaties. One concrete option for even stricter treaties discussed in the literature is a “carve-out” for climate policy, meaning that issues related to climate are excluded from the application of the treaty’s requirements (Paine and Sheargold 2023). The most far-reaching option, of course, is the partial or complete exit from the investment treaty regime. The options range from abolishing ISDS clauses to terminating some or all of a country’s investment treaties. In a remarkable step, several EU member states have announced their withdrawal from the ECT amid a reform process, stating that the treaty is incompatible with their green transition policies. As a consequence of these unilateral actions, the EU Commission (2023) proposed a coordinated EU withdrawal from the ECT.

Earlier versions of this article were presented at the 2021 annual convention of the International Studies Association and the 2022 International Political Economy Conference of the Austrian, German, and Swiss Political Science Associations. We thank the participants for the critical discussion. We also thank the three anonymous reviewers and the editors for their helpful comments and feedback. We are grateful for the support of the research project “Transformations and Sustainability Governance in South American Bioeconomies (SABIO)” at the University of Münster, which is funded by the German Federal Ministry of Food and Agriculture. L.B.-F. acknowledges funding provided by the German Federal Ministry of Education and Research as part of the project “At the Science Policy Interface: LANd Use SYNergies and CONflicts Within the Framework of the 2030 Agenda LANUSYNCON” (grant 01UU2002) and by the Collaborative Research Center Future Rural Africa (CRC228) funded by the German Research Foundation.