Why is there no investor-state dispute settlement in RCEP? bargaining and contestation in the investment regime

The Regional Comprehensive Economic Partnership (RCEP) is one of the most important mega-regional trade agreements signed to date. Yet, it failed to include an Investor-State Dispute Settlement (ISDS) mechanism in its investment chapter. What explains this omission? To unpack this, we examine international negotiations as a two-step process. In the first stage, we theorize that initial preferences towards ISDS are based on countries ’ orientation toward foreign direct investment (FDI), experience with ISDS, and past treaty practice. Second, we theorize that during protracted negotiations, adverse regime developments and domestic politics can have a profound impact on treaty design. To test our framework, we examine the RCEP negotiations. Our analysis shows that mounting cases as well as the eroding norm of ISDS in other treaties lowered support for ISDS as the negotiations progressed. Then, a change of government in Malaysia shifted that country ’ s position dramatically, which tipped the balance against ISDS in the final round of negotiations. Our findings have important implications for the international investment regime. They highlight the factors that determine countries ’ initial preferences while also demonstrating the importance of developments during the negotiations, which can lead to the abandonment of the institutional status quo.


Introduction
The Regional Comprehensive Economic Partnership (RCEP) is one of the most notable mega-regional economic agreements signed in recent years. 1 As the largest agreement of its kind in terms of GDP, proponents anticipated that RCEP member countries would write ambitious, new rules governing trade and investment, with major implications for global supply chains centered on East Asia. 2 Particularly important was the potential for RCEP members to establish new rules in the global investment regime, which lacks a central rulemaking body and is governed by a decentralized network of international investment agreements (IIAs). 3Yet, once unveiled, the investment provisions contained in the final text of the RCEP agreement were notably unambitious.The chapter contains boilerplate provisions on protection, facilitation and liberalization of investment sectors. 4Most notable, however, was that the final agreement lacked formal enforcement provisions, as negotiating partners could not agree on an investor-state dispute settlement system (ISDS) and postponed the decision to a later stage, without a clear timetable.The outcome was surprising, as ISDS was an integral part of the initial negotiation drafts but was eliminated from the final agreement.
Why would RCEP countries fail to include enforcement provisions for investment in this new, landmark treaty?For some, this outcome is unsurprising given the growing backlash against ISDS has led multiple countries to terminate and renegotiate IIAs with ISDS clauses. 5At the same time, much of the attention has centered around reform rather than complete abandonment and most studies focus on Bilateral Investment Treaties (BITs) rather than Preferential Trade Agreements (PTAs), which tend to be more complicated to negotiate and may have more signatories.Moreover, evidence from the RCEP negotiations indicates that ISDS was included in the initial negotiation drafts.Additionally, anchoring ISDS in mega-regional trade agreements in Asia has been accomplished before, as the case of the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) shows. 6In this context, we view RCEP as an important case.The absence of an important treaty design feature, one that was the institutional status quo for many years and was planned for inclusion at the outset of the negotiations, is important because it serves as a test of existing explanations for reorientation in the global investment regime, while also pointing towards new dynamics.
Our unique argument is that understanding why ISDS was not included in RCEP requires accounting for the two-step process which occurs during protracted international negotiations.The first step is to map the preferences of the parties going into the negotiations.Building on existing insights from the literature, we theorize that countries' initial negotiation positions on ISDS are a function of their orientation towards foreign direct investment (FDI), their experience with ISDS, and their past treaty practice.This provides a baseline as to whether ISDS provisions are likely to be supported going into the negotiations.Second, we argue that during the negotiations additional factors become operative, which can have a major impact on the final treaty design.This is especially true in the context of PTA negotiations, where the median negotiation is nearly two years, and some agreements can take as long as fourteen to complete. 7e theoretically account for three factors that are likely to become salient during negotiations.First, power is important during negotiations, as powerful states will attempt to influence the outcome towards their preferred position.Second, we theorize that domestic politics during the negotiations, such as changes in government and the politicization of ISDS domestically, can impact the stability of countries' preferences.Third, we theorize that adverse regime developments, such as the incidence of ISDS cases against signatories as well as the erosion of existing legal norms during the negotiations, can affect positions by altering the actors' perceptions of the costs and benefits. 8Overall, we contend that in the context of protracted negotiations country preferences are less stable and are more likely to soften in response to domestic politics and adverse regime developments.This increases the likelihood that countries switch positions.This insight is important because we emphasize that in the case of a regime like investment, which faces a period of instability, changes during the negotiation period can have a profound impact on treaty design, effectively nudging the parties off the original equilibrium path more easily than would occur in other circumstances.
We test our framework through an analysis of the recently negotiated RCEP agreement.In the first part of the paper, we map each country's orientation towards FDI, experience with ISDS, and past treaty practice to categorize country positions at the beginning of the negotiations.Then, we analyze the RCEP negotiations themselves to empirically evaluate whether the factors we identify affected the design of the final treaty.During this exercise, we pay special attention to Malaysia, which became a pivotal player in the negotiations.Specifically, the Malaysian government switched their position from support of ISDS to strong opposition right before the final round of negotiations in 2019.This pushed the balance of negotiations from the probable adoption of standard ISDS provisions with some country exceptions to the final outcome where it was omitted altogether.
Our findings have important implications for the evolution of the global investment regime.First, they show that states are increasingly willing to abandon the institutional status quo as their preferences evolve.In this sense, our results add to recent work showing how exposure to adverse regime developments can weaken the institutional status quo, but we also show that other factors including past treaty practice and domestic orientation towards FDI matter as well. 9Second, we show that we must also account for developments during the negotiations, which can systematically alter the final outcome.Specifically, this case demonstrates the fragility of negotiating positions to domestic political developments, which can profoundly alter the negotiating dynamics, leading to a loss of support for items that were originally slated for inclusion.We theorize that this is increasingly likely in a regime like investment which has been weakened by rising backlash that has cast doubt on the legitimacy of core principles that were once taken for granted.

The global investment regime in flux
The global investment regime is at a critical juncture.The modern regime began developing in the late 1950s, and today includes at its core a network of thousands of IIAs, including BITs and investment chapters in PTAs, which set the terms for foreign investment. 10Beyond most-favored nation status, national treatment, and protection from expropriation, most IIAs signed since the 1980's include ISDS, a mechanism whereby investors can sue countries at international tribunals if the terms of an agreement are abrogated. 11During the initial expansion of IIAs beginning in the 1960s and lasting until the 1980s, IIAs were viewed as a device that helped solve a time-inconsistency problem between rich country international investors and host countries in the developing world. 12The inclusion of ISDS in IIAs became a standard treaty provision in the 1980's. 13Proponents claimed that strong legal protections conferred through ISDS would help facilitate FDI, especially in country pairs where there was an undersupply of FDI due to concerns about property rights protections. 14fter gaining a toehold in the previous three decades, IIAs with ISDS proliferated rapidly in the 1990s and 2000s.Increasingly, ISDS was regarded as a means to mitigate political risk, depoliticize investment disputes, and promote the rule of law. 15Central to this dynamic was that proponents began to view signing IIAs with ISDS as a norm in the international system, which led to their wide diffusion, even in country pairs where they were unlikely to facilitate a meaningful increase in FDI. 16This norm was reinforced by countries and organizations associated with the "Washington Consensus" as well as the example of several Asian countries whose economic success was said to be a product of attracting FDI. 17 In turn, ISDS began to be included not only in BITs but also in PTAs and other multilateral instruments, which helped expand the scope of ISDS beyond North-South pairs.In addition to diffusion, countries were also more likely to sign IIAs with ISDS in an effort to make up for the perception of weakness in their domestic property rights regimes and in response to worries about regime instability. 18y the mid-2000s, cracks began emerging in the regime.Academic studies found mixed results on whether BITs increased FDI. 19Even more importantly, private investors began to utilize ISDS, which in some cases led to multi-billion-dollar monetary decisions. 20Even in cases where the country won, simply being sued can inflict reputational harm, negatively impacting future FDI inflows. 21For some countries-and many observers-this has cast doubt on the benefits of the current investment regime.Studies have shown, for example, that countries sign fewer IIAs after experiencing litigation 22 and several countries, like India, Indonesia and Pakistan have begun to terminate and renegotiate their agreements to gain greater regulatory space. 23Moreover, these concerns have expanded to developed countries, as arbitration (or the threat thereof) has sparked public backlash.Central to this debate is the notion that IIAs with strong protections for foreign investors impinge on state sovereignty, which has a "chilling effect" on domestic legislation. 24espite widespread concerns and several reform proposals, no new regime has emerged. 25This puts countries negotiating IIAs in a challenging position.Do they include provisions associated with the status quo or do they write new rules?Specifically, do they include stringent or weak ISDS positions or do they switch to some other potential system?Most literature on international institutions proposes that the status quo is sticky, suggesting that the default for new negotiations should be to retain features of the extant regime, which in this case includes some form of institutionalized ISDS. 26But as the perceived costs of ISDS have increased, many states seem to be caught in an "unstable" position which can evolve during multi-year negotiations. 27This provides researchers with a unique opportunity to assess the process by which states engage with an extant yet fragile regime. 28For example, recent scholarship on institutional design suggests that contextual variables, including the current state of a given regime, influence states' positions in negotiations. 29Increased uncertainty about distributional outcomes and domestic political reactions means that the negotiations may take on increased importance.In addition, past work has often seen ISDS as a direct result of the bargaining power of the more developed country.But as backlash has increased, the interests and intensity of preferences of developing countries have received more attention. 30

Theoretical framework
Several important insights have emerged from the literature on ISDS in IIAs.First, preferences for the inclusion of ISDS are a function of the nature of the state's economy, including the role that FDI plays in a country's economic system as well as policymakers' beliefs and the preferences of multinational corporations (MNCs). 31Second, preferences for ISDS are influenced by the perceived costs of the investment regime to various actors, especially in cases where adverse legal rulings have led to a newfound appreciation that ISDS can "bite." 32Third, the inclusion of ISDS (and the stringency of ISDS provisions included) is strongly conditioned by the relative bargaining power of the negotiating countries, with the standard assumption being that capital-exporting countries will generally be in favor of ISDS and that developing countries are more skeptical but will be forced to consent to stringent ISDS when in a weak bargaining position vis-à-vis a powerful capital exporter. 33hese initial factors are help us understand important regularities in IIA design.However, our central contention in this paper is that these factors only tell part of the story, especially as the legitimacy of the investment regime has been challenged in recent years.Here we argue that the inclusion of ISDS provisions will also be affected by additional factors, including shifts in domestic politics and adverse regime developments during the "negotiation phase." In the following paragraphs, we present a theoretical framework that accounts for the two-stage nature of international treaty negotiations, and which integrates insights that we believe help explain why countries may include or exclude ISDS from their final treaty design.Figure 1 demonstrates the process.The left panel of the figure outlines the factors that affect the parties' preferences going into the negotiations, whereas the right part of the figure depicts factors that become operative during the negotiations.Below we elaborate the logic behind each step and then discuss how they interact to affect treaty design.

Preference formation
A first factor that matters during the phase of preference formation is a country's orientation towards FDI.For years IIAs were touted by proponents as a way for developing countries to gain access to global FDI, which helped contribute to their diffusion. 34In recent years, this rosy view has dimmed, especially as the costs of ISDS have come into focus. 35Regardless, there still exist numerous proponents of IIAs, who continue to tout them as part of a broader strategy that states can use to attract and promote FDI. 36mportantly, pro-ISDS actors comprise an influential epistemic community occupying roles in the bureaucracy, in industry groups, and intergovernmental organizations. 37Moreover, policymakers in some countries have incentives to sign IIAs to overcome regime instability or to improve the "investment climate," which can help them maintain power. 38Thus, even though there may not be a clear, mechanical linkage between ISDS and FDI, there are pro-ISDS constituencies in many countries, that view it as a way to attract, and maintain, FDI.
There are also important reasons to believe that capital-exporting interests advocate for ISDS as well.The United States Trade Representative (USTR) and other public agencies in many capital-exporting 34 Elkins et al. (2006); Jandhyala et al. (2011); Moehlecke and Wellhausen (2022).
35 Poulsen and Aisbett (2013); Thompson et al. (2019).36 Calvert and Tienhaara (2023).37 Yackee (2012); Poulsen and Aisbett (2016).countries have historically touted ISDS as a way to mitigate risks for investors. 39Moreover, recent scholarship suggests that firms have an outsized role in foreign economic policymaking, and studies have begun to document the numerous linkages between large firms and governments. 40Kim, for example, shows that firm lobbying influences when BITs are negotiated, with the strongest effects observed for lobbying by larger firms. 41Following a similar logic, Betz, Pond, and Yin show that firms strategically locate subsidiaries in countries with large networks of IIAs to expand their legal coverage. 42aken together, this suggests that a country's orientation towards inward and outward FDI is an important factor that helps explain initial preferences for ISDS.Developing countries with a strong policy orientation toward inward FDI are likely to see the potential advantages to ISDS, which may be especially true for countries enacting widespread market reforms or attempting to overcome a reputation for poor domestic property rights protections or government instability.Moreover, if the country has already engaged in extensive reforms to help foster property rights protections, the potential costs of exposure to ISDS are likely to be lower than for countries that have yet to enact such reforms.At the same time, countries seeking to promote outward FDI are also likely to have preferences for ISDS as a way to provide an additional layer of protection for investors.This is likely to be especially true in cases where capital-exporting firms have influence over policymaking.
A second factor that matters is a country's experience with ISDS.The underlying idea is that governments update their assessment of international rules depending on their calculations of the costs and benefits.In the investment regime, disputes are high-impact but low-probability risks that governments are increasingly concerned with. 43Litigation affects governments both as a home state and host state.
For host states, being hit by a legal claim has several consequences.First, the costs, including legal and tribunal fees, average about $13 million combined for the parties, and can be much higher for complicated cases. 44More importantly, the average award size is over $400 million, and many claims are higher. 45For some developing countries these sums can be a major burden.Second, the signal that a negative judgment sends to international investors can have negative consequences on FDI inflows. 46inally, states may also pay a significant "sovereignty cost" by exposing themselves to the regime. 47isputes, or the threat thereof, send a signal to the host state government and societal groups that the pursuit of legitimate policy objectives is constrained by international obligations.Given that recent ISDS cases have drawn heightened attention and led to public discontent, governments feel obliged to act as voters demand change. 48Whatever the outcomes of the cases, governments do not want to be seen as breaching international legal commitments or having to limit their "regulatory space."Following this logic, we expect that countries that have faced investment claims are more likely to be defensive in investment negotiations.Countries become risk-averse, and this shows in the treaty negotiations.Defensive interests translate into lower ambition for investor protection rights, more demands for explicit rights for states to regulate, and less support for legalized dispute settlement systems.
A third factor that matters is a country's past treaty practice.One dynamic that has been prominently observed both in the context of BITs, but also in PTAs more generally, is that some countries push their own model treaty language. 49Moreover, existing literature shows how the design of pre-existing templates of chapters in BITs, more often than not, are diffused to newer agreements with other parties. 50Countries engage in copy-pasting for several reasons.One explanation is powerbased and suggests that powerful states use their leverage in negotiations to insert preferred provisions into new treaties.A second explanation is that states copy-paste to minimize transaction costs, as they can rely on existing templates to rationalize negotiations.Taken together, this suggests that states should value establishing coherence in their treaty networks which can help reduce negotiation costs in the future as well as ensure that focal treaties contain their favored provisions. 51Moreover, this may have the added effect of decreasing uncertainty in the larger regime, which systemically important countries may value.Castle shows, for example, that countries often engage in "strategic sequencing" of trade agreements in order to establish model rules that they can carry forward to future negotiations. 52hese dynamics of copy-pasting mean that countries will show a tendency to rely on the provisions included in their most recent IIAs going into the negotiations.This type of path dependency likely contributes to a status quo bias or what we call legal consistency.At the same time, and as we discuss in more detail below, the focality of past treaty language is sensitive to adverse developments that occur during the negotiations.

Multi-party negotiations
Once negotiations begin, a new set of factors becomes operative.Of primary importance is the distribution of power between the signatories.There is robust literature suggesting that powerful countries systematically steer design outcomes towards their preferences, which has been shown in the context of PTAs, ISDS provisions in BITs, and BIT renegotiations. 53The distribution of power is likely to be stable throughout the negotiations, as power typically does not fluctuate enough from year to year to cause a major reorientation.Thus, power is often incorporated into the initial scoping exercise at the outset of the treaty.If powerful countries have strong preferences for (or against) a treaty provision, then other countries understand what this means and negotiate accordingly.In the context of multiparty agreements with multiple issue areas, powerful countries are likely to engage in various strategies, such as issue linkage, side payments, and threats of exit that can help them obtain their preferred outcomes. 54It should be noted, however, that power should decline in relevance with the intensity of preferences.We see this dynamic in cases like the initial TPP, where the US model BIT was largely copy-pasted directly into the agreement. 55However, when there is uncertainty and changing preferences as there is at the present moment in the global investment regime, power may be less important.
A second factor that influences country positions during the negotiations is what we call adverse regime developments, which are events that impact perceptions of the benefits and costs of the regime for states.Principal among these are new legal developments such as additional ISDS cases brought against the parties to the agreement as well as rulings in pre-existing cases.Cues about being sued (costs) weigh more heavily during negotiations.The idea here is that governments engage in "experiential learning" which has been applied in earlier studies that analyze trade negotiations in the multilateral system.One example of experiential learning was during the Uruguay Round negotiations.The GATT dispute settlement experience was pivotal for explaining the shift of the European Union to embrace a more legalized system. 56The concept has also been applied to study the spillover from one type of trade forum (the WTO) to another (PTAs), e.g., negative experiences in WTO dispute settlement translate to country positions in PTA negotiations leading to less enforceable and more flexible commitments. 57The literature on BITs also focuses on the costs of ISDS cases.Not only do some cases involve large financial sums a government would need to pay, but high-profile cases also have political costs, as governments worry about their policy space as well as public backlash.While governments representing home states of investors may see such cases as positive, being a defendant has a much higher impact on governments positions.In addition, loss aversion may lead to a rethinking of initial positive positions.
An additional adverse regime development that impacts the parties' positions during the negotiations occurs as countries actively begin to re-evaluate their legal approach in other treaties.In this sense, we view past treaty practice as an important predictor of initial preferences going into the negotiations, but we also stress that in some cases, treaty practice may be in flux or be in a state of active re-evaluation, which can further erode the status quo.The field of investment has witnessed potential changes when focusing on overall policies to support certain BIT models or by starting to move away from existing models by either terminating BITs or renegotiating BITs.Besides adjustments related to existing BITs, governments might start questioning ISDS if such shifts also appear in related trade or investment negotiations (e.g., CPTPP), especially if those negotiations are overlapping.Thus, we expect that changes in investment protection practices by the parties can lead to more uncertain preferences, which can affect treaty design.This is likely to be exacerbated if parties to the negotiations participate in other treaty negotiations where the norm of ISDS has eroded or where active renegotiation has taken place, which has the effect of weakening the focality of ISDS.
A final factor important during protracted negotiations is domestic politics.When it comes to the political economy, governments represent different economic interests in society and, as a consequence, have different orientations towards open markets, trade in goods and services, and FDI.Many left and progressive parties are more skeptical of the impacts of market liberalization due to concerns from labor, whereas right and conservative parties are often more market-friendly due to the influence of business interests.This may also be witnessed concerning the salience of the costs imposed by ISDS, with left-leaning politicians and their supporters more sensitive to what they view as a mechanism which empowers MNCs at the expense of broader social concerns, such as domestic regulation in the name of public health, the environment, and labor. 58Moreover, domestic regimes vary in how the interests of domestic actors impact the policymaking process.In some countries, pro-ISDS actors such as MNCs may be better able to organize and lobby policymakers, whereas in other countries labor or environmental groups may hold more sway. 59ur view is that domestic political developments are likely to become particularly consequential during the negotiations, which are often protracted in the case of PTAs. 60Two possibilities are important.First, government turnover within a given country can lead to the empowerment of governments that are more skeptical (or supportive) of ISDS, which can shift the balance of the negotiations.For example, there are numerous examples of left-leaning governments skeptical of ISDS coming to power in Latin America, as well as in Europe, Australia, and New Zealand, forcing major shifts in government policy. 61Calvert and Tienhaara suggest these shifts are likely to be more pronounced in established democracies with strong parties as opposed to political systems where officials are less experienced and elite capture is more likely. 62Regardless, the core point is that government turnover can lead to the empowerment of a different winning coalition comprised of voters and other domestic actors who are more or less supportive of ISDS.
Second, we posit that the politicization of ISDS is particularly likely during the negotiations themselves, as this raises the saliency of the issue at home, providing anti-ISDS actors a political 57 Wüthrich and Elsig (2021).58 Hufbauer (2016).59 Kim (2023).60 Lechner and Wüthrich (2018).61 Helfant (2023).opportunity to mobilize.We have seen numerous examples of negotiations for ISDS being politicized, such as in the case of the Transatlantic Trade and Investment Partnership (TTIP) and during debates over the TPP and later the CPTPP.Some governments are more likely to be sensitive to this politicization, which can lead to a loss of support after the negotiations have started.Moreover, adverse regime developments that occur during the negotiations, such as an initiation of a new claim or a decision in an outstanding dispute, might magnify domestic opposition, creating more pressure on governments to alter the status quo.In sum, we argue that domestic political developments, such as changes in government from right to left and progressive parties as well as mounting public backlash that puts pressure on policymakers, can alter the initial status quo.

Final treaty design
Our framework takes into account how initial preferences interact with dynamics that occur during the negotiations to jointly determine whether ISDS is likely to be included in a given IIA.Importantly, ISDS is likely to be included in an IIA if preferences are aligned in favor of ISDS among the parties and/or if the most powerful parties to the negotiations have a strong preference for ISDS.This helps explain why we observe relatively strict ISDS provisions in instances where there is significant bargaining asymmetry or when there is a preference by all parties for strong ISDS provisions. 63Moreover, the probability of including ISDS, as well as the stringency of the provisions, will also increase as powerful parties to the negotiations bring to the table treaty language used in previous agreements, which they hope will contribute to consistency in their treaty networks and establish focal legal language that can be recycled into future agreements. 64At the same time, countries are unlikely to include ISDS if their preferences are aligned against inclusion and/or if the most powerful countries in the negotiations strongly oppose it, which is likely to be a function of their negative experience with ISDS. 65ur primary innovation is to argue that additional variables become operative during the negotiations, which can lead to major changes in the final outcome.First, adverse regime developments, including new ISDS cases and negative rulings, can swing the balance against ISDS during the fragile negotiation period.Moreover, these cases are likely to be weighed more heavily by the participants.Similarly, an erosion of ISDS in parallel treaties during the negotiations may also lead states to re-evaluate the focality of ISDS in their treaty practice, which can weaken the strength of the previous status quo.Second, changes in domestic politics during the negotiations can alter treaty design dramatically.In particular, a shift in preferences, due to a change in government or due to governmental responsiveness to increased domestic politicization of ISDS can also lead to a weakening of the positions of the parties.

Preferences, negotiations, and the design of RCEP
In this section, we empirically evaluate our theoretical framework in the case of the recent RCEP agreement.First, we map the countries' initial positions by examining their orientation towards FDI, their experience with ISDS, and their past treaty practice.Then, in the second part, we analyze the negotiations themselves.For this, we rely on participant observation as well as interviews with key officials.

FDI orientation
We start by focusing on initial support for ISDS at the onset of negotiations.We assume that countries have preferences towards ISDS as a function of their orientation towards FDI.As noted above, the classic view is that capital exporters are largely in favor of ISDS, whereas capital importers are thought 63 Allee and Peinhardt (2010, 2014); Allee and Lugg (2016); Alschner and Skourevskiy (2016); Huikuri (2023); Elkins et al.  (2006); Jandhalya et al. (2011).to be more skeptical.Yet, there are also a variety of reasons to expect that these preferences are not so clearly delineated, as some countries are likely to want to use ISDS as a credible commitment device or in response to other domestic developments, and pro-FDI constituencies may view ISDS as desirable.Moreover, some countries' preferences change over time as they move from being capital importers to capital exporters, such as in the cases of South Korea and China. 66As a result, we focus both on inward investment and outward investment and also on policies beyond BITs that signal certain policy orientations towards FDI.To capture this complexity, we consider actual FDI flows as well as regulatory policies, and we also separate preferences by home state and host state.
For inward investment and host state attitudes, we use the OECD's FDI Restrictiveness Index as a proxy. 67The index builds on four aspects of the FDI climate within a country, namely foreign equity restrictions, screening and prior approval requirements for FDI, rules for key personnel, and other restrictions affecting the operation of foreign enterprises. 68The index covers 22 sectors and averages the total scores to derive the total FDI index, which ranges from 0 to 1. Figure 2 maps this index for all RCEP countries going into the negotiations.
Figure 2 demonstrates that China, Myanmar and the Philippines are the three RCEP countries with the highest FDI restrictiveness.They are followed by Vietnam, Thailand, and Indonesia.In the middle are New Zealand, Malaysia and Laos.On the other end of the spectrum, we find that Singapore, Japan and Cambodia have the least restrictions on FDI.These countries have invested in their domestic infrastructure to best accommodate foreign investors.This may also correlate positively with the existence of ISDS to credibly signal that the interests and rights of foreign investors will be protected. 69inally, Australia, Korea and Brunei are generally open to FDI and, hence are likely to be in favor of ISDS as a means to attract FDI.
Next, we examine outward FDI to help determine preferences from the home state perspective.Since there is no comparative and reliable data on FDI export policies to actively promote FDI, we look at the role FDI plays in relation to overall GDP.This allows us to capture how domestic firms' investments in foreign markets might support the inclusion of ISDS in trade agreements.Figure 3 summarizes the percentage of outward FDI and stocks in relation to overall GDP across RCEP countries.
Assuming that countries with a higher ratio are more inclined to protect their firms' outward investments, we see that Singapore and Malaysia exhibit the highest FDI outflow in relation to GDP, ranging between 5 to 7 percent.South Korea and Thailand as also above average.On the other side of the spectrum, we find that New Zealand, Laos and Cambodia have relatively little outward FDI, therefore firms from these countries are less likely to attempt to influence policymaking.Also, Australia, Indonesia and Vietnam show relatively low numbers of 0.5 to 0.7 percent.In the upper end of the middle, we find countries such as China, the Philippines and Japan with percentages from 1.03 to 1.95.Importantly, China and Japan are among the top sources of outward FDI in the world. 70From this, we can deduce that Singapore and Malaysia are the countries most likely to support ISDS as a means to support their firms' investments overseas, followed by South Korea and Thailand, and thereafter China, Philippines and Japan.Conversely, countries like New Zealand, Australia, Laos, Indonesia, Vietnam and Cambodia are less likely to support ISDS on this basis.

Experience with ISDS
Second, we assume that countries have preferences towards ISDS as a function of their experience with litigation prior to the negotiations.Table 1 shows the distribution of the 17 cases brought against RCEP members between 1994 to 2012.The countries hit with the most complaints were the Philippines (4), Indonesia (3), Malaysia (3) and Vietnam (3).Australia, China, Myanmar and Thailand each faced one ISDS case.Singapore, Japan, South Korea, Brunei, Cambodia, Laos, and New Zealand all had zero cases against them prior to the negotiations.The most notable case during this time was Phillip Morris v. Australia, which was filed in June 2011 under the Australia Hong Kong BIT and challenged Australian plain packaging regulations on tobacco.This case contributed to significant politicization of the issue domestically in Australia, which, as we outline in more detail below, influenced countries' positions during the negotiations. 71inally, we examine whether firms from RCEP countries were users of the regime.We assume that governments follow cases where their firms are participants and expect adequate compensation in case of other governments' wrongdoings.Therefore, we expect that governments see such cases as testimony on the desirability of ISDS, which in turn informs their position at the beginning of negotiations.Table 2 provides an overview.There are only four RCEP countries whose firms used ISDS during the period from 1994 to 2012.Firms from Malaysia (3), Singapore (2) and China (2) were the most frequent users with multiple cases brought against other governments.One Australian firm used ISDS.From the investors (and home states) the outcomes of all cases were positive as investors' claims were supported or settlements with foreign governments were reached (see Appendix A1).There were no cases from Brunei, Cambodia, Indonesia, Japan, Laos, Myanmar, New Zealand, Philippines, South Korea, Thailand, or Vietnam during this period.
Ultimately, RCEP countries' experiences with ISDS prior to the negotiations are neither overwhelmingly positive nor negative.Eight of the thirteen original parties to the negotiations experienced at least one claim  as a host country, indicating that most of the countries were aware of legal exposure associated with the regime.Moreover, the Philippines, Indonesia, and Malaysia were hit by multiple claims, indicating that they were likely the parties most aware of the potential costs.In contrast, four RCEP countries had firms that had utilized ISDS during the pre-negotiation period.Our reading is that this does not provide strong evidence that countries were supportive simply as a function of usage.At the same time, prior to the negotiations, the costs of ISDS were not so severe as to prevent the parties from including standard ISDS provisions either.Malaysia's experience with ISDS prior to the negotiations is illustrative.In Malaysian Historical Salvors Sdn Bhd v Malaysia, which was decided in May 2007, a Tribunal ruled in favor of the government, but the decision was annulled thereafter and Malaysia had to pay the costs of the proceedings. 72This led to some internal concerns within the government, which resulted in the initiation of consultations between the government, academics, civil society, and Malaysian companies as well as a reorganization within the government. 73Ultimately, cases brought against Malaysia led to a newfound appreciation of the costs, which led to a recognition by some in the government of the need to explore potential safeguards in future treaties.Yet, at the same time, the government was also responsive to Malaysian companies, such as Telekom Malaysia and MTD Capital, who had initiated claims.In sum, the Malaysian government was becoming more cognizant of the potential costs of ISDS but was still reluctant to fundamentally question ISDS, instead embarking on internal reforms to minimize the costs.

Past treaty practice
We now examine country preferences based on past treaty practice.Here, we postulate that countries that signed more treaties including ISDS before the negotiations were more likely to support the inclusion of ISDS at the beginning of the negotiations.To measure this, we look at both PTAs and BITs signed before 2012 that are currently still in force and include a general ISDS clause.This is because some countries attempt to push their own models, both in the context of BIT negotiations, but also in PTAs. 74We also include whether countries have an existing published Model BIT, as countries tend to rely on these templates in future negotiations. 75This also signals less flexibility to adjust one's existing model.
From the results in Table 3, we identify China, Malaysia, Singapore, South Korea and Vietnam as countries with a high number of BITs and PTAs with ISDS provisions (over 40 treaties each), which made them most likely to include ISDS in future treaties at the time.We then observe countries such as Australia, Indonesia, Japan, Laos, the Philippines, Vietnam and Cambodia each with more than 20 treaties with ISDS provisions.All others have fewer than twenty treaties in force when negotiations start.Overall, this demonstrates that many RCEP parties had robust BIT programs, and had extensively included ISDS in past treaties, indicating that ISDS was the status quo option leading up to the negotiations for several of the parties, even if the costs were being re-evaluated as ISDS cases began to increase.
We categorize each RCEP country with respect to their position on ISDS going into the negotiations in Table 4. Group 1 are countries that were very supportive of ISDS.This includes Japan, South Korea and Singapore.Group 2 were moderately supportive; this includes Malaysia and Brunei.Countries in Group 3 were neither overtly positive nor overtly negative vis-à-vis ISDS, hence their positions were most likely to be unstable.These countries include China, New Zealand, Laos, Thailand, Australia and Cambodia.The fourth group was skeptical towards ISDS determined by their overall attitude towards FDI, past dispute experience and active involvement in reforming BITs.This group includes Indonesia, Myanmar, the Philippines and Vietnam.Table 4 provides an overview of the country groups based on our initial mapping.For FDI attitude and past cases, we differentiate between host and home state as outlined above.

The negotiations
This section examines the negotiations to better understand which factors contributed to the final treaty text, which excluded ISDS.While ISDS is explicitly listed as part of an in-build agenda for further negotiations in years to come in the work program, 77 it is clear that with the requirement of consent of all the member states for amending RCEP, achieving a later inclusion of ISDS is an arduous task.To understand this important outcome, we highlight how the initial preferences of the parties interacted with dynamics that occurred during the negotiations.We use primary and secondary sources and rely on interviews with negotiators.Importantly, this allows us to assess whether country positions conform to our conjectures about their basic preferences, while also helping us understand which combination of factors best explain why the parties decided to omit ISDS after initially favoring including standard ISDS provisions.
RCEP negotiations began in 2012 among 16 parties. 78Figure 4 provides an overview of major developments that occurred during the negotiations.The consensus visible in early drafts was that the treaty would include a standard ISDS mechanism allowing foreign investors to sue host states in international arbitration.Importantly, the early draft was prepared based on the ISDS provision in the Association of Southeast Asian Nations (ASEAN) Comprehensive Investment Agreement (ACIA), which came into force in 2012. 79Based on interviews with Malaysian and Thai negotiators, there was little discussion on the ISDS provision at this time since it was prepared by the ASEAN Secretariat, and 76 South Korea has a Model BIT but does not publish it publicly, see Qc and Triantailou (2017).

77
The agreement obliges parties to start discussions on an investment-related dispute settlement mechanism no later than two years after the date of entry into force of the Agreement, subject to the agreement of the Parties (RCEP, Article 10.18).

78
Initial negotiations included countries from ASEAN (Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, and Vietnam) and Australia, China, India, Japan, New Zealand and South Korea 79 ASEAN Secretariat (2016).many expected it to be included. 80Moreover, countries like Singapore and Japan viewed ISDS favorably, as both countries had never faced investor-state arbitration. 81etween 2012 and 2016, there were eight rounds of negotiations.These were conducted in manners unique to ASEAN.For example, before each round, ASEAN members would convene a day before for internal discussions.Only after these internal discussions would the entire RCEP group meet for negotiations.Further, in each negotiation round there would be two co-leads; one from an ASEAN member state and the other representing the six other countries, which were referred to as ASEAN Foreign Partners (AFPs). 82These eight rounds were mostly focused on ironing out schedules suited for each country, including tariffs, services, and investment.During these early stages, countries closely followed the RCEP Principles Negotiating Guidelines, 83 which included a short paragraph on investment that pointed towards the inclusion of ISDS.Ultimately, during these rounds of negotiations, there was no obvious expression of dissatisfaction or opposition to ISDS, especially given that the parties were relying on the ACIA as a blueprint.
It was not until 2016 that countries in the "skeptical" group started to more forcefully voice concerns about ISDS.Smaller economies such as Myanmar and Laos argued that ISDS provisions could negatively impact their domestic policy space and courts as they were slowly opening up to FDI. 84 Laos grew more skeptical because of several ISDS cases that were brought against the country, where they incurred expensive legal costs and gained a reputation of being an unfavorable investment destination.Myanmar officials, for their part, were worried that their early-stage work on a new investment framework could expose them to ISDS claims under the fair and equitable treatment provision.Indonesia and the Philippines, already skeptical going into negotiations, advocated for a watered-down version of ISDS.They suggested that international arbitration or conciliation be subject to a separate written agreement or consent between the disputing parties as required under their existing laws and policies. 85Vietnam, for its part, was openly questioning ISDS, as one of its negotiators stated in the same draft that "The fact that Indonesia and the Philippines require a separate written agreement does not mean that this Agreement shall constitute the consent of [Viet Nam] [any other Party] to conciliation or arbitration" (Version dated 6 December 2016).
Simultaneously, several countries in the "unstable" group started to become more negative towards ISDS.Both Australia and New Zealand had recent and unique experiences that negatively colored public perception towards ISDS. 86In 2015 Australia prevailed in its dispute over tobacco plain packing against Phillip Morris, but it spent around US$40 million in legal fees, and the case generated significant domestic political backlash, much of which was not felt until after the decision was announced. 87This case was observed closely in New Zealand and led to increased public interest there as well.In 2017, New Zealand under the Ardern Administration renounced ISDS, stating that no future PTAs would include it.Therefore, when countries in the skeptical group started voicing their aversion to ISDS, Australia and New Zealand were quick to follow suit, as confirmed by Malaysian negotiators. 88n summary, leading up to more intense negotiations in 2019, critical voices had increased among RCEP countries.The negotiations became more contentious on ISDS in 2019. 89  India ultimately withdrew from the agreement altogether, citing its concerns in the text such as tariff barriers, intellectual property rights, and the ISDS mechanism itself, Wang and Sharma (2021).member countries, there was a major shift in position from one of the least expected of all RCEP members going into the negotiations: Malaysia.The government initially represented a position that was moderately supportive of ISDS, as illustrated above, but based on insights from interviews, it shifted its position sharply during the negotiations, much more than any other RCEP country.Malaysia's sudden opposition to ISDS provided space for other negotiating parties to voice their pre-existing concerns, which helped turn the tide against the inclusion of ISDS.This shift drew support from Australia, India (when it was still a Party to the negotiations), Indonesia, New Zealand and Thailand. 90oreover, several other countries' positions had also softened considerably.Ultimately, the parties agreed to water down the ISDS provision into a work program, replacing ISDS with a State-to-State dispute settlement system.Even pro-ISDS classified states like Singapore or Japan did not try to stop the development. 91e now assess in more detail whether the factors we outlined in our theoretical frameworkbargaining power, adverse regime developments, and domestic politics-can help explain this outcome.We pay special attention to Malaysia given that its shift in position was the most remarkable from the outset of the negotiations.

Power
One common explanation for the design of PTA provisions is bargaining power.Initially, several of the most influential countries in the region, namely Japan and South Korea, had pro-ISDS preferences.China was more lukewarm, but still supportive.This suggests that it may have been possible for these countries to push through ISDS had they had a strong preference for doing so even against mounting pressures.However, evidence from the negotiations suggests that they did not leverage their collective bargaining power in an effort to include ISDS in the final agreement.
In the case of China, preferences for ISDS were not that salient.China has concluded a large number of investment agreements which include "restrictive ISDS clauses" i.e., that considerably limit the type of investment claim that can be filed before ISDS.This indicates that China is supportive of ISDS in general, but that their preferences are not overly strong. 92Moreover, China strategically adopted a more collaborative and consensus-building role in the negotiations.Overall, China's involvement in the RCEP negotiations translates into a strategic intent to achieve consensus among the diverse group of participating economies. 93imilarly, Korea and Japan demonstrated a cautious approach, or a weak preference, towards ISDS.In Korea, public debate intensified with the Korea-US FTA negotiations in 2006, leading to a more critical look at investment agreements. 94In Japan there were debates against ISDS in parliament, with criticisms often reflecting broader political views rather than specific issues with ISDS. 95These internal debates limited the projection of power for Korea and Japan in the negotiations.
Ultimately, it appears that there was no coalition of powerful states that were willing to push strongly for ISDS.The lack of strong preferences, coupled with domestic opposition in several of the more democratic countries, led to a negotiating dynamic whereby countries that could have done more to push for ISDS were unwilling to do so.At the same time, there is no evidence to suggest that a powerful coalition of anti-ISDS states emerged who actively pushed against inclusion.Instead, as we explore in more detail below, we see a general weakening of preferences in several of the countries in response to ongoing changes occurring in the regime, as well as government shifts towards more skepticism in several of the parties.This suggests that power remains an important factor, but only if it is coupled with strong preferences.Moreover, because RCEP was in the context of a larger, multilateral 90 Interview with Malaysia negotiators, 6.8.2021; interview with Thai RCEP investment negotiator, 2.4.2022.91 Bath and Nottage (2021).92 Du (2022).

Business and Politics
negotiation with multiple issue areas, rather than a bilateral negotiation over investment only, countries may have been less willing to expend political capital on this increasingly contentious issue and instead choose to pursue other goals.

Adverse regime developments
One factor that seems to have played a role in the case of RCEP is adverse regime developments occurring during the negotiation phase.As we theorized above, adverse regime developments can be thought of in two interrelated ways.First, ISDS cases have the potential to change the parties' perceptions of the costs of ISDS, which can lead to a shift in preferences.Second, an erosion of the ISDS norm in other treaties and among the parties can further weaken support.We address each in turn.
Table 5 identifies ISDS cases filed against RCEP member countries during the negotiation period.Between 1994 and 2012 when the negotiations started there were a total of 17 cases brought against RCEP states, but there were 31 cases brought during the negotiations period from 2012 to 2020.Several countries experienced multiple cases, including China, South Korea, Vietnam, the Philippines and Indonesia.Moreover, several cases, including Phillip Morris v. Australia which was active during the negotiation period, as well as two cases launched in 2011 against Indonesia (Al Warraq v. Indonesia and Rafat v. Indonesia), and one case against the Philippines and Vietnam each, led to significant politicization of the issue.Importantly, even though Australia eventually prevailed in the Phillip Morris case, concerns about regulatory chill spilt over into public debate in other democratic states including New Zealand and South Korea, making ISDS a harder issue to negotiate.
In addition to an uptick in ISDS cases, we also see during the negotiation period a weakening of treaty practice with respect to ISDS among several of the parties, which had the effect of undermining the focality of ISDS as a norm.Table 6 provides an overview of the number of agreements negotiated as well as terminated and renegotiated by each RCEP country during the negotiations.We provide a relative number to capture the degree of change, by taking the percentage of terminations and renegotiations in relation to the total number of treaties negotiated.The table shows that Indonesia and Australia terminated or pushed for renegotiations in a large portion of their agreements.Several other parties were starting to explore changes at relatively low levels of under 5% (Cambodia, Japan, the Philippines) and Malaysia and Thailand at levels below 10% of existing BITs.Brunei, China, Laos, Singapore, South Korea and Vietnam have more elevated activities, whilst Myanmar, prior to the military takeover had introduced domestic investment policy reform which included negotiations of BITs with the EU.This suggests that treaty practice toward including standard investment provisions, especially with respect to ISDS, was changing among several of the parties during the negotiations.Perhaps equally important, was what was happening with respect to the other mega-regional which was being negotiated during the same time: CPTPP.When the US was still part of the (then) TPP negotiations, ISDS was non-negotiable, and analysis of the treaty text indicates that the investment chapter from past US agreements was used as a template. 96But when the US dropped out in 2017, the remaining 11 countries chose to revisit the ISDS mechanism, instituting several key changes. 97mportantly, New Zealand signed five side letters with Australia, Peru, Brunei, Malaysia, and Vietnam that excluded the application of the ISDS. 98volving standards towards ISDS in the CPTPP after the US' departure highlighted the potential for controversy and discord over ISDS provisions, leading RCEP negotiators to adopt a more cautious approach.Moreover, other skeptical countries, observing New Zealand's maneuvers in CPTPP, might have felt emboldened to adopt a similar approach in RCEP.This is particularly relevant to the countries that were part of both the CPTPP and RCEP negotiations, as reflected in Figure 5.The CPTPP was officially signed in 2018, further affecting the observed shifts in the RCEP negotiations in 2019. 99n sum, a noticeable increase in ISDS cases during the negotiations, as well as changing treaty practice in the form of renegotiations and opt-outs from ISDS, including in the regionally significant 96 Allee and Lugg (2016).
99 Yunus (2019).See Appendix A3 for a timeline of event for the CPTPP and RCEP negotiations.
CPTPP, influenced country positions as the RCEP negotiations progressed.The net effect was that as the potential costs of ISDS became more salient to the parties, the strength of ISDS as the status quo option was being weakened.

Domestic politics
A third important factor during the negotiations is domestic political developments.We theorized above that two developments are likely to have an important impact during the negotiations, government turnover and domestic politicization of ISDS.Our analysis indicates that these factors impacted several of the countries during the negotiation period, including Australia and New Zealand.However, the most notable instance was in Malaysia, where a change in government led to a complete shift in that government's stance in the negotiations.We first discuss the countries where domestic politics had a smaller impact on the negotiations, and then turn to Malaysia where the impact was more pronounced.
In Australia and New Zealand domestic politicization of ISDS as well as government turnover meant that both countries never fully supported ISDS in RCEP.In Australia, the prominent Phillip Morris case, coupled with activism from domestic groups and the Green party, led to a major rethinking of ISDS around the time that the RCEP negotiations were about to get underway.The newly elected, center-left Gillard government announced in April 2011 that it would not include ISDS in any trade agreements.This changed when the conservative Liberal-National Coalition gained power in 2013.It committed to ISDS on a case-by-case basis and signaled that it was willing to conclude negotiations that had dragged on for years, including a PTA with South Korea. 100Ultimately, in Australia, domestic politics complicated the government's ability to strongly advocate for ISDS in RCEP.In New Zealand, domestic politicization of ISDS during the TPP negotiation process as well as concerns related to the Phillip Morris case in Australia, led to a repudiation of ISDS altogether by the new labor government elected in 2017. 101As such, New Zealand went from being a country that was indecisive on ISDS to one that was staunchly against it as the negotiations entered their critical phase.Domestic political developments had a particularly profound effect in the case of Malaysia.Since the start of the TPP negotiations, civil society groups had formed coalitions across countries to express their concerns regarding ISDS. 102In July 2016, when information leaked that the RCEP's investment chapter would include ISDS, some 80 participants from 18 countries met in the capital of Malaysia and drafted an open letter calling for "TPP rules" in RCEP to be rejected. 103The letter specifically points to the leaked RCEP investment chapter, and highlights that "apart from a few safeguards, all the TPP's  Capling and Ravenhill (2015).101 Kawharu (2016).
substantive and main investor-to-state dispute settlement (ISDS) provisions have been proposed." 104Yet, a large part of the business sector continued to support ISDS.Malaysian investors including a few state-owned enterprises (SOE) investing abroad deemed protection through ISDS necessary for their investments in certain RCEP member countries.This position was also reflected by the positive experience of Malaysian firms acting as claimants.A representative of Telekom Malaysian stated that ISDS was useful as a last resort avenue to settle their claim against Ghana. 105There was also the oftenshared view that the experience with ISDS made businesses aware of the full benefits and scope of IIAs, including ISDS. 106Going into the RCEP negotiations, Malaysia's treaty practice was largely consistent.Yet, internal concerns led to a slow adaptation and increasing politicization created a more contested status quo during the RCEP negotiations.
A second major development that occurred was the 2018 Malaysian general election.Following the 1MDB scandal, the Pakatan Harapan party won elections based on a nationalistic and anti-corruption campaign.It focused on increasing welfare and strengthening the power of Malaysian companies, workers, farmers and trade unions.While the party was pro-investment and business, it proclaimed to increase domestic productivity and not to be over-reliant on foreign investment.In other words, the new governing party saw foreign investment as a means for Malaysian companies to access greater markets.In light of the corruption scandal, the party also greatly stressed the need to empower civil society groups, by contributing to the growth of these organizations and increasing transparency to allow their collective voices to be heard.As such, the new Minister of International Trade and Industry, Darell Leiking, was more open to arguments advocated by civil society organizations opposing ISDS. 107urther, Minister Leiking was also regularly provided with critical briefings on ISDS by civil society members of the Trade Advisory Council (TAC). 108This, combined with some pre-existing skepticism within the Minister's office and the bureaucracy vis-à-vis ISDS due to past cases, led the Minister to officially withdraw support for the ISDS provision.This was announced right before the final round of negotiations in September 2019. 109n Malaysia, these two domestic factors were deeply intertwined.The incumbent government was displaced by a new government following a controversial scandal regarding the misappropriation of investment funds.The new government, in turn, was more receptive to the views of domestic groups who were increasingly voicing skepticism of the status quo in terms of trade and investment policies.With continued consultations with these groups through direct channels such as briefings and roundtables, the increased skepticism towards ISDS was made apparent.This compounded other skeptical voices inside policymaking circles, which had been seeking to insulate Malaysia from exposure to ISDS in future economic agreements.Ultimately, the Pakatan Harapan switched the government's long-standing acceptance of ISDS, coming out against its inclusion in RCEP.

Conclusion
In this paper, we advanced a framework for understanding sources of continuity and change in the investment regime, with a specific focus on the newly negotiated RCEP agreement, where ISDS was excluded during the final negotiations.Our unique argument is that international treaty-making is a dynamic process that unfolds in two stages.First, countries determine their initial preferences towards ISDS based on: (1) orientation towards FDI; (2) experience with ISDS; and (3) past treaty practice.
Once negotiations begin, another set of factors become operative.Here we argue that there are two that have been overlooked in the literature.First, we argue that adverse regime developments that occur 104 The full text of the open letter can be found here: https://igj.or.id/wp-content/uploads/2016/12/RCEP-CSO-sign-onletter.pdf105 Discussion of author with business groups in a special TPP roundtable in 2013.
106 Ayub (2018).during the negotiations, such as new ISDS claims and the erosion of ISDS as a legal norm in parallel treaties, can soften the parties' positions.A country's experience with the regime, especially with disputes before negotiations is important and has been documented extensively in the literature. 110However, we argue that adverse developments during the negotiations are likely to exert a particularly strong impact on actors' perceptions of the costs and benefits.Second, we argue that negotiations over ISDS are sensitive to domestic political developments, which can also serve to magnify concerns with ISDS.These include government turnover and domestic politicization, which can shift government policies by empowering actors who are more (or less) skeptical of ISDS.
Our analysis of the RCEP negotiations provides support for our approach and helps contextualize the exclusion of ISDS in this case.Originally there was a weak coalition in favor of ISDS, which was included in the initial negotiating drafts.However, over the course of the negotiations, mounting ISDS cases, including most notably the Phillips Morris case in Australia, as well as the politicization of the TPP and the weakening of the ISDS norm there, led to increased skepticism among the parties.This boiled over when Malaysia abruptly changed positions to be in strong opposition to ISDS after the election of a new government there.Malaysia's shift during the negotiations galvanized opposition and led to a strong push for abandoning the idea of ISDS in the final rounds of the negotiations.We theorize that this type of shift is likely to be more consequential in regimes that are already in flux, such as the investment regime, where backlash has weakened the status quo.
Our research has important implications.It sheds light on the importance of developments during multilateral negotiations, and the profound impact this can have on the design of international treaties generally, and the investment regime specifically.This is important given that treaty negotiations can stretch on for many years.Hence, our framework captures this dynamic process by mapping the prenegotiation conditions, capturing the changes that occur during negotiations, and demonstrating how these can be decisive in shaping the final outcome of treaties.This approach recognizes the dynamic interplay between international and domestic factors in shaping the global investment regime.This opens new avenues for future studies to examine how these factors interact in other settings as well as apply them to other treaty design dimensions.

Figure 1 .
Figure 1.Preference formation, negotiations, and treaty design in the investment regime.

Figure 2 .
Figure 2. RCEP signatories by restrictiveness to FDI.Source: OECD FDI Restrictiveness Index, data from 2012, for countries lacking data in 2012 we use data available in following years.

Figure 3 .
Figure 3. Outward FDI and stocks (as % of GDP).Source: UNCTADStat, data from 2012.Brunei Darussalam and Myanmar are missing due to data unavailability.
Among the remaining 80 Interview with Malaysian negotiators on 6 August 2021, interview with Thai negotiator on 2 April 2022.81 Bath and Nottage (2021); Chaisse and Olaoye (2020); Lewis (2021).With respect to Japan, however, the Financial Times broke the story in March 2021 that the Hong Kong-based Shift Energy fund had filed what may be considered the very first investment treaty arbitration claims against Japan.82 Elms (2021); Rillo et al. (2022).83 Guiding Principles and Objectives for Negotiating the Regional Comprehensive Economic Partnership.https://asean.org/wpcontent/uploads/2012/05/RCEP-Guiding-Principles-public-copy.pdf84 Based on face-to-face discussions and email exchanges with Lao DR and Myanmar RCEP negotiators in 2016 and 2017, especially during one author's advisory work in the context of drafting the Myanmar Investment Law 2016 and the Lao PDR Investment Promotion Law 2016.85 Negotiation draft dated 6 December 2016 (on file with authors), 86 Chaisse and Nottage (2017).

107
Interview withMalaysian negotiators, 6.8.2021.108Murugiah (2019); civil society organizations had more direct access to the Minister through their membership in the TAC, which was set-up by the Pakatan Harapan Government.

Table 1 .
Case experiences as host states(1994-2012) Source: UNCTAD Investor-State Dispute Settlement Navigator.Note: We exclude India which has been part of the final negotiations, more detailed information about the cases can be found in Appendix A1.

Table 4 .
Summary of country groupings, from skeptical to supportive Notes: we use a five-scale indicator for each of the five factors highlighted: ++ very supportive for ISDS; + supportive; +/neutral; -skeptical;very skeptical; NA: not available.We aggregate the total for mapping RCEP members into 4 categories.Figure 4. Timeline of events in CPTPP and RCEP.

Table 5 .
ISDS cases against RCEP countries during negotiations (2013-2020) UNCTAD Investor-State Dispute Settlement Navigator, see Appendix A2 for a full list. Source: