Right-Wing Populists and the Politics of Renegotiating International Environmental Agreements

The rise of populists, primarily of the right-wing variety, across developed and developing country democracies has captured the attention of pundits and the media. This is hardly surprising. After all, right-wing populist (RWP) leaders and their parties, including Donald Trump in the United States, the National Rally in France, the Alternative für Deutschland in Germany, Modi’s Bharatiya Janata Party (BJP) in India, and Erdogăn’s Adalet ve Kalkınma (AK) Party in Turkey, have excoriated the global elite, attacked globalization, repealed environmental regulations, and implemented trade restrictions (e.g., Levitsky and Ziblatt 2019; Norris and Inglehart 2019; Rodrik 2021). These parties also challenge the liberal international order, including international trade and international environmental agreements (IEAs) that promote global governance and international cooperation (Böhmelt et al. 2023; Fiorino 2022; Jones 2021). Their antipathy toward international cooperation is alarming, as electoral support for RWP parties across developed and developing country democracies has grown—as illustrated in Figure 1—sharply during the last three decades.

The possibility that the growing popularity of RWP parties may undermine multilateral cooperation has prompted scholars to investigate their impact on trade and environmental policies, immigration policy, and democratic backsliding in primarily developed democracies (Böhmelt 2021; Böhmelt and Ezrow 2023; Colantone and Stanig 2018; Levitsky and Ziblatt 2019; Norris and Inglehart 2019). Yet we know relatively little about how the emergence of RWP parties affects a key component of the liberal international order, namely, the global governance of environmental issues via the design of IEAs. As described herein, this study is to my knowledge among the first to address this lacuna by unpacking the association between RWP parties and the design of international agreements, particularly flexibility provisions in IEAs.

The focus of my study is not accidental, as RWP parties often engage in climate change denial and threaten to withdraw from IEAs (Böhmelt 2021; Fiorino 2022; Huber et al. 2020; Kulin et al. 2021). Yet creators of environmental agreements are cognizant of such threats and thus focus on designing mechanisms to deter defection (e.g., withdrawal) from these agreements by IEA member states without discouraging participation by countries (Bernauer et al. 2010, 2013; Koubi et al. 2020; Marchiori et al. 2017). In this regard, numerous researchers emphasize that flexibility provisions in IEAs, such as “exception, reservation, escape, and withdrawal clauses,”¹ address these trade-offs by allowing “states to withdraw from cooperation, fully or partially in response to new circumstances” (Morin et al. 2022, 19; see also Boockmann and Thurner 2006).

Interestingly, new pooled data on flexibility provisions in IEAs ratified by 189 countries (1980–2017) show that the level of these provisions across a diverse set of IEAs² varies from a minimum of 0 to a maximum of 3 on a 0–3 ordered scale.³ They also reveal another intriguing phenomenon—that democratic states (including those in which RWP parties operate) are more likely than other types of political regimes to self-select, negotiate, and incorporate maximum flexibility provisions in IEAs. Indeed, illustrations drawn from a newly constructed data set on flexibility provisions in IEAs described later reveal that the level of IEA flexibility provisions adopted by democratic IEA member states has increased since the mid- to late 1990s, albeit with some temporal variation (see Figure 2) and cross-sectional variation (Figure 3). The variation illustrated in Figures 2 and 3 raises the following question addressed in this article: when are democracies more likely to pursue and successfully obtain higher levels of flexibility provisions in IEAs?

To answer this question, I build on previous research on how domestic political factors influence the design of IEAs (Bauer and Menrad 2019; Brandi et al. 2019; Colgan et al. 2021). But unlike these and other studies on the effect of right-wing populists on trade policies or democratic backsliding (e.g., Jones 2021; Rodrik 2021), I argue that the growing legislative empowerment of RWP parties in democracies, including those in IEA member states, accounts for higher levels of flexibility provisions in IEAs. This theoretical framework is as follows. To start with, RWP parties in democracies often publicly oppose IEAs by claiming in the domestic political arena that environmental regulations in IEAs that harm the material interests of “ordinary people” are designed by global elites. These parties then employ such populist anti-IEA narratives to further sensitize their constituencies (e.g., those employed in the coal industry) to the costs of stringent IEA-mandated environmental standards.

Building on this, I suggest that if RWP parties hold a majority of the seats in the lower house of the national legislature, they will use this legislative leverage to claim to other IEA member states that, first, their “hands are tied” to the material interests of domestic constituencies supporting them and that, second, they need significant concessions—that is, flexibility provisions—to meet the interests of these constituencies to remain in the IEA agreement that is being negotiated at the international level. Majority representation in the legislature allows these parties to threaten to withdraw from IEAs if they do not obtain maximum flexibility provisions to compensate their constituencies from other member states.

Legislative leverage also allows RWP parties to make credible promises ex ante to other IEA member states that they can successfully ratify IEA agreements—provided they have substantial flexibility provisions—in the national legislature. Credible promises of this sort thus enable RWP parties with sufficiently high legislative representation to obtain more IEA flexibility provisions. This leads to the following hypothesis: RWP parties in democracies that hold a majority of the seats in the national legislature are more likely to obtain maximum flexibility provisions in IEAs.

Because this hypothesis focuses on the choice of flexibility provisions by democracies, I employ a pooled sample of seventy-five democracies (1980–2017) to test the hypothesis. Statistical results support the hypothesis stated earlier. The results that corroborate my hypothesis remain robust not only in samples in which IEAs are the unit of analysis but also across a battery of several robustness tests.

This article contributes to research on the effects of populism and studies on the design of international agreements. It does so by taking an important step forward in unpacking the influence of RWP parties on the liberal international order by providing a new theoretical framework that explores how the legislative empowerment of RWP parties affects the design of international agreements, particularly flexibility provisions in IEAs. Furthermore, I use and present newly developed data on RWP parties’ representation in the national legislatures of developed and developing country democracies. Scholars of environmental politics and governance can use the data to explore how RWP parties influence policies and regulations (e.g., emission standards) that have profound environmental consequences.

Global environmental politics has often centered on the negotiation and ratification of IEAs (Koubi et al. 2020; Marchiori et al. 2017; Marcoux 2009; Roberts et al. 2004; Wangler et al. 2013). Accordingly, researchers have devoted substantial effort to exploring how factors like economic development, democracy, and lobbying by interest groups determine when countries self-select into IEAs (Bättig and Bernauer 2009; Bernauer et al. 2010; Marchiori et al. 2017). Others have examined how export diversification, the presence of NGOs, trade openness, participation in international organizations, and policy learning explain when states ratify IEAs (Bernauer et al. 2010; Roberts et al. 2004).

Furthermore, given the variation in compliance with IEAs, studies explore how third parties and legalization of IEA agreements influence compliance with IEA mandates (Böhmelt 2022; Böhmelt and Butkuté 2018). Examining compliance is important, as the domestic distributional consequences of IEAs are not as stark and broadly based as preferential trading arrangements (PTAs). IEAs are thus not politically salient like PTAs, which often incorporate flexibility provisions to address challenges faced by member states when complying with PTA-mandated trade reforms (Baccini 2019; Pelc 2009). Unpacking and empirically evaluating the links between domestic politics (e.g., the rise of RWP parties) and flexibility provisions in IEAs is therefore a hard case and challenging precisely because IEAs are typically not politically salient or consequential.

Notwithstanding this challenge, extant research on IEAs converges on the view that, like PTAs, flexibility provisions within IEAs facilitate compliance with their mandates (Boockmann and Thurner 2006; Marcoux 2009; Thompson 2010; von Stein 2008). Despite this benefit from IEA flexibility provisions, Figure 3 reveals that some (democratic) member states pursue and obtain a substantial amount of these provisions, while others desist from doing so (Böhmelt 2022; Böhmelt and Butkuté 2018). What accounts for this variation?

Less effort has been expended toward answering this question despite extensive work on flexibility provisions, such as escape clauses in PTAs (e.g., Baccini 2019; Pelc 2009). To be sure, some scholars examine why political elites in Latin America favor flexibility clauses in IEAs (Freire et al. 2021), while others explore why populist incumbents grant nonstate actors access to IEAs (Böhmelt 2022; Böhmelt et al. 2023). Yet research on the determinants of flexibility provisions in IEAs has not received sufficient attention despite the growing levels of IEA flexibility provisions and the rise of RWP parties that publicly seek to defang IEAs.

I thus address the lacuna by arguing herein that RWP parties have political incentives to pursue maximum IEA flexibility provisions but can successfully obtain these flexibility provisions only when they have a majority seat share in the national legislature. The next section presents these theoretical arguments in detail.

The theoretical framework presented in this section focuses on flexibility provisions in IEAs that have not only been signed⁴ but are also ratified and in effect. To this end, the theory first discusses why RWP parties oppose environmental standards and seek IEA flexibility provisions. It then explains why legislative empowerment of RWP parties induces these parties to pursue IEA flexibility provisions and allows them to successfully obtain maximum flexibility provisions from other IEA member states.

RWP parties derive their distinct identity from an ideational composite that combines populist discourse with nativist narratives. Their populist rhetoric emphasizes that society is divided into “ordinary people” pitted against the global elite, who denigrate the aspirations of ordinary people. Their nativist narrative propagates nationalist, xenophobic, and ethnocentric ideals. Consequently, the RWP parties’ populist and nativist narratives have two effects.

First, they drive them to oppose international trade agreements (e.g., PTAs), as such agreements are deemed to promote economic globalization, which—as per the RWP parties’ populist rhetoric—benefits only the elite while imposing substantial costs (e.g., job losses) on ordinary people. RWP parties focus more on opposing trade agreements than IEAs, as the adverse distributional consequences of international trade are widespread across society and thus more salient. Their opposition to trade is distinct from the past—particularly the 1990s—when right-leaning (including right-wing) parties endorsed trade liberalization and market reforms (Caiani and Graziano 2022; Jones 2021; Rodrik 2021). Second, they induce RWP parties to oppose environmental regulations and standards, including those mandated by IEAs.

In fact, RWPs often assert that mainstream parties and international institutions promote multilateral environmental cooperation via IEAs only to benefit the elite, rather than to benefit ordinary families (Huber et al. 2022; Walter 2021). This allows RWP parties to gain votes, especially from voters who have lost their jobs because of the pro-globalization and pro-environmental agendas of mainstream parties and international institutions (Colantone and Stanig 2018; Fiorino 2022; Gidron and Hall 2017; Huber et al. 2020). Apart from the allegations mentioned previously, targeted toward mainstream parties, RWP parties also downplay and even deny the severity of environmental issues (Huber et al. 2020, 2022; Kulin et al. 2021). This appeals to their constituents for two main reasons.

First, implementing antipollution environmental standards substantially increases the costs for heavy polluters, including industries in the primary sector (e.g., natural resources) and heavy-manufacturing industries (Broz et al. 2021; Fiorino 2022; Guriev and Papaioannou 2022). Higher costs from enforcing IEA-mandated antipollution standards lead to firm closures and job losses among heavy-polluting industries. These job losses adversely affect supporters of RWP parties, as they often work in pollution-heavy industries. Second, RWP parties often focus on economic nationalism, which not only involves subsidizing domestic manufacturing industries that produce value-added jobs for their constituents but also generates higher pollution. Accordingly, RWP parties refuse to limit emission levels or implement stringent environmental regulations (Bomberg 2017; Żuk and Szulecki 2020).

Examples from several RWP parties support the preceding claims. For instance, the Modi-led RWP BJP Party in India subsidizes the country’s coal industry and solicits support from voters by promising them jobs in fossil fuel industries (Kashwan 2022; Saravanan 2023). The BJP has also alleged that IEAs like the Kyoto Protocol violate the country’s sovereign choice of promoting job growth via the fossil fuel industries (Saravanan 2023). Poland’s RWP Law and Justice party is a nationalist pro-coal party that obstructs climate policies to appease constituents from the coal industry, and it publicly denounces IEAs (Żuk and Szulecki 2020). The RWP Conservative People’s Party of Estonia denies human-driven global warming and dismisses the Paris Agreement as classic left-wing action. The RWP AK Party in Turkey has focused on job creation from the coal industry for its voters and has rejected ratification of the Paris Agreement from 2015 to 2021, believing that this agreement threatens the livelihood of the party’s constituents (Adaman et al. 2017; Arsel et al. 2021).

Appealing to voters opposing the imposition of environmental standards allows RWP parties like the BJP and the AK to differentiate themselves from non-RWP centrist and antielite left-wing populist parties. Indeed, non-RWP centrist and left-wing populist parties often favor IEAs and compliance with IEA environmental standards, which makes them easy targets for political attacks by RWP parties. Unlike non-RWP parties, RWP parties are more interested in reducing their countries’ obligations concerning IEA-mandated environmental standards.⁵ An important and blunt instrument for reducing these obligations is flexibility provisions, which allow states to withdraw from their IEA-mandated requirements for a certain time period. IEA flexibility provisions also curtail the commitments of RWP parties to implement antipollution policies. I next build on these claims to explain when RWP parties will openly pursue such provisions.

Greater electoral support for RWP parties since the late 1980s has allowed these parties to become the ruling parties and enter government in several democracies (Levitsky and Ziblatt 2019; Norris and Inglehart 2019). In office, these parties are driven to pursue maximum IEA flexibility provisions for two reasons. First, greater electoral support for RWP parties generates substantial bottom-up pressure from their constituents to secure sufficient concessions from other IEA member states when negotiating IEAs. Once a RWP party becomes the ruling party, or at least the party with the highest vote share in the electoral arena, its constituents will expect the RWP party to deliver policies that are compatible with the constituents’ material interests (Guriev and Papaioannou 2022; Thomson et al. 2017).

In the context of negotiating IEAs with other member states, RWP parties favor pursuing maximum flexibility provisions in IEAs to lower the costs of IEA-mandated environmental regulatory standards borne by their constituents. Failure to pursue IEA flexibility provisions can damage RWP parties’ reputation and credibility among their constituents. This jeopardizes their odds in future elections. RWP parties that garner sufficient domestic electoral support will thus have strong political incentives to be responsive to their core constituents’ interests by pursuing maximum IEA flexibility provisions when negotiating with other IEA member states.

Second, note that a higher number of electoral vote shares obtained by RWP parties typically leads to their majority control of the legislature. More specifically, regarding the swing ratio, which captures the “percent in legislative seats associated with a 1 percent change in legislative votes” (Niemi and Fett 1986, 76), a greater vote share obtained by RWP parties translates to a majority seat share held by these parties in the lower house of the national legislature (Gallagher and Mitchell 2005; Niemi and Fett 1986). Controlling a majority of the legislative seats gives RWP parties domestic political leverage, triggering the following cascading effects.

It first encourages RWP parties to exaggerate the adverse impact of IEA-mandated environmental regulatory standards on their constituents to further sensitize them about the alleged possibility that IEAs serve the global elite’s selfish interests rather than those of ordinary people. This provides the rationale for RWP parties to signal to their domestic audience, particularly their core constituents, that they will be tough and uncompromising in their demands for higher IEA flexibility protections that protect ordinary people. Doing so allows RWP parties to credibly reinforce their populist reputations as the sole political organizations that genuinely represent the material interests of ordinary people.

This increases expectations from their constituents that RWP parties—that enjoy a legislative majority—have the will and capacity to deliver on their promises, including those that pertain to IEAs. Such expectations further induce RWP parties to hold themselves accountable for delivering on their promises to obtain maximum IEA flexibility provisions. Accordingly, RWP parties that command a majority legislative seat share will advertise themselves as parties that can extract substantial IEA flexibility provisions and actively pursue these provisions in international negotiations with other IEA member states to placate their constituents.

Numerous examples support these claims. For example, once the BJP obtained a legislative majority in India’s Lok Sabha in 2014, it began to “inform citizens that international agreements like the UN Framework Convention on Climate Change (UNFCCC) infringe upon India’s economic sovereignty” (Saravanan 2023, 94) by restricting the growth of the country’s fossil fuel industries that generate jobs (Kashwan 2022). Given the BJP’s majority seat share in the national legislature between 2014 and 2019 and its legislative clout to implement policies—recognized by the electorate—the party emphasized to voters that it was the only organization that could safeguard the country’s interests and jobs generated from fossil fuel industries, as it has the “political capital to obtain concessions when negotiating environmental agreements with other states” (Saravanan 2023, 97).

After the AK Party gathered a majority of seats in Turkey’s Grand National Assembly in the 2015 general election, it publicly attacked IEAs to sensitize citizens to its allegation that these agreements benefit only the West and hurt Turkey’s vital heavy-manufacturing industries (Adaman et al. 2017; Arsel et al. 2021). Proclaiming itself as the party of the common people, the AK Party then pledged that it would not make compromises on environmental agreements but instead obtain safeguards in these agreements to protect heavy-manufacturing industry jobs in the country (Adaman et al. 2017).

I build on these examples and the preceding discussion to suggest in the following section that when RWP parties control a majority seat share in the national legislature, they will use this legislative heft as bargaining leverage to negotiate and obtain substantial IEA flexibility provisions from other IEA member states.

RWP parties holding a majority of legislative seats will have a substantial number of coparty members (a sizable caucus) in the national legislature. Therefore substantial legislative presence makes it easier for RWP parties to design policies, control the policy-making agenda, amend bills, and implement policies. It also allows these parties and their leaders in IEA member states to credibly signal to other IEA member states that they have a dominating and united front in the national legislature. The credibility of this signal has three key effects.

First, it enables RWP parties and their leaders (who are incumbents) to cultivate their reputations for being strong in the international arena. Such a reputation allows IEA member states ruled by legislative-dominant RWP parties to extract concessions, including flexibility provisions, from other IEA member states while conceding fewer concessions. Second, it allows legislative-dominant RWP parties to convey to other negotiating IEA member states that their win-set is small, given that said parties represent domestic constituents that oppose IEA-mandated environmental standards. A smaller win-set means that member states in which RWP parties control the legislature will avoid offering concessions to other IEA member states to prevent electoral punishment.

Instead, these RWP parties will pursue maximum flexibility provisions from other IEA member states to appease their constituents. If RWP parties with a legislative majority fail to obtain significant flexibility provisions, they may face pressure to defect from IEAs that do not address their supporters’ material concerns. Common knowledge of the risk of defection under immense domestic pressure enhances the bargaining leverage of incumbent RWP parties. Incumbent RWP parties can and will use this leverage to credibly threaten other negotiating IEA member states that they will walk away (i.e., withdraw) from IEAs that are detrimental to the material interests of their constituents.

For instance, during negotiations over the Paris Agreement in 2015–2016, negotiators from India’s External Affairs Ministry—under instructions from the ruling RWP BJP Party, which had a legislative majority in 2015—threatened to formally exit the Paris Agreement unless other member states incorporated flexibility provisions that would curtail India’s obligations to fulfill the agreement’s mandates (Kashwan 2022; Saravanan 2023). These threats were credible, as Indian negotiators emphasized that the BJP’s legislative majority status obliged it to protect constituents and that India would be forced to exit the Paris Agreement if sufficient flexibility provisions were not offered.

Following directions from the Duterte-led RWP PDP-Laban Party (which had a supermajority in the country’s congress by 2016), negotiators from the Philippines indicated that they would be compelled to leave the Paris Agreement unless sufficient flexibility provisions were incorporated into the agreement (Berse 2024). They noted that the ruling PDP-Laban Party’s supermajority status meant that the party had an obligation to obtain flexibility provisions in the Paris Agreement to protect its citizens’ interests (Berse 2024).

Note that RWP incumbent parties do not directly deliver such withdrawal threats to other negotiating IEA member states; rather, they instruct official negotiators from their countries to issue the withdrawal threats (as suggested by the preceding examples) to other IEA member states during the negotiation process. The credibility of these threats will induce other IEA member states to acquiesce to demands for maximum flexibility provisions to dissuade withdrawal from IEAs by states in which RWP parties dominate the national legislature. Nevertheless, RWP parties recognize that the process of formally withdrawing from IEAs can be costly and onerous. Hence, although RWP parties make threats about walking away from IEAs, they do not formally (and fully) exit IEAs. As discussed later, the statistical results from my sample do not reveal associations between the legislative seats held by RWP parties and formal withdrawals from IEAs, which provides added support to these claims.

Third, member states in which RWP parties control a majority of legislative seats can also credibly promise to other IEA member states that they can ratify a given IEA in their domestic legislature as long as the agreement contains substantial (i.e., maximum) flexibility provisions. This promise is credible, because RWP parties that command a legislative majority can use their party representatives in the legislature—the majority party caucus in the legislature—to vote for, approve, and ratify IEAs. A high probability of IEA ratification will induce other IEA member states to offer concessions by agreeing to incorporate maximum IEA flexibility provisions that are sought by the legislatively dominant RWP parties in RWP party–led states. This helps RWP parties obtain maximum flexibility provisions in IEAs. Overall, the theoretical framework presented here leads to the following hypothesis:

I employ a pooled sample of seventy-five democracies during the 1980–2017 period to test H₁, which examines the association between RWP parties in democracies and maximum flexibility provisions in IEAs. While the democracies in my sample meet the criteria for democratic regimes defined by Cheibub et al. (2010),⁶ the results do not alter if democracies are coded using the Polity Index. Furthermore, as reported later, the results remain robust when I use a sample of only IEAs as the unit of analysis.

My pooled sample incorporates a global sample of developed and developing country democracies (1980–2017) from all the following regions of the globe: Africa, Asia, central and eastern Europe, western Europe, North America, and Latin America. Employing a global pooled sample of democracies (the theoretical focus of my study) enhances the generalizability of the statistical results. The dependent variable employed to evaluate H₁ is the level of flexibility provisions in IEAs negotiated and adopted by states in my sample. This dependent variable, denoted as IEA Flexibility, is constructed using Mitchell et al.’s (2020) IEA Database and Morin et al.’s (2022) Flexibility Index, which operationalizes the degree of flexibility provisions in IEAs based on six types of clauses that give members the ability to unilaterally terminate their commitments partially or fully.

Building on this measure, I then match each democracy member state to the IEA Flexibility provision measure in each IEA starting from the year the agreement is signed.⁷ Next, I aggregate the information by calculating the modal level of flexibility in the environmental agreements by country-year. The resulting dependent variable, IEA Flexibility, is a discrete ordinal variable ranging from 0 (minimal flexibility) to 2 (maximum flexibility).

The independent variable in H₁, the share of legislative seats held by RWP parties in the lower house of the national legislature (RWP Seats), is operationalized in three steps. For the first step, I identify all political parties with legislative representation for each country-year in my sample. Next, in this set of all political parties, I identify and code RWP parties for each country-year using the following three criteria. The first, as per Mudde’s (2007) definition, is that the party must be “populist,” meaning that it divides society into the “pure” people versus the corrupt “elite.”

The second is that, unlike left-wing populists, the populist parties must be “right wing.” Following existing research, I define populist parties as right wing if they employ nativism and ethnonationalist xenophobia, accuse “outside” groups (e.g., ethnic or religious minorities) of threatening the national identity or culture, advocate antiminority discrimination, propagate economic nationalism combined with antiglobalization, and encourage “strict adherence to conventional moral norms, and intolerance of multiculturalism” (Norris and Inglehart 2019, 76; see also Mudde 2007).

The third criterion applies Norris and Inglehart’s (2019, 4) definition of RWP parties as parties that endorse an authoritarian “style of governance (that) challenges constitutional checks-and-balances” and seeks to undermine democratic institutions. The list of sources to identify RWP parties in my sample and examples of these parties are presented in the Supplementary Appendix.

For the second step, I gathered and coded the data on the share of legislative seats held by all political parties, including RWP parties, in the lower chamber of the national legislature in every country-year in my sample. For the third step, I employed the aforementioned data to calculate the cumulative seats RWP parties hold in the national legislature for each democratic country-year in my sample. I then divided the cumulative seats held by RWP parties by the total number of legislative seats in each democratic country-year in the sample. Doing so led to the share of legislative seats held by RWP parties, labeled RWP Seats, which is the independent variable. The RWP Seats measure exists for all the countries—across every region of the globe listed earlier—in my sample, as all the countries have had RWP parties. Yet there is some variation in the number of RWP parties and hence variation in the RWP Seats measure, as some countries have had just one RWP party, whereas other sample countries have had as many as five RWP parties.

To test H₁, I assess the effect of RWP Seats on IEA Flexibility in several specifications, including those with minimal controls, and in a set of fully specified models that include economic and political control variables that account for alternative explanations for variations in IEA Flexibility provisions. In the set of fully specified models, I first incorporate the following control variables, which, as per previous research, influence flexibility provisions in international agreements, including IEAs: GDP (logged), GDP Per Capita (logged), GDP Growth, Trade Openness, and a Democratization dummy, which is coded as 1 for IEA member states that have democratized over the past ten years (e.g., Baccini et al. 2015; Mansfield and Milner 2012).

Next, given that the presence of veto players influences the design of international agreements (e.g., Mansfield and Milner 2012), I estimate an additional fully specified model in which I control for veto players, which is drawn from the political constraints data set.⁸ Veto players account for independent branches of government (executive, lower, and upper legislative chambers), the extent to which these branches are controlled by the same party, and preference heterogeneity within each branch. In the said model, I also control for the binary EU variable for European Union member countries due to the active role of the EU in fostering IEA membership for its members, and for oil rents, because countries dependent on exporting fossil fuels (which increases pollution) for revenue may favor IEA flexibility provisions to reduce the burden of meeting IEA-mandated standards.

The PR electoral system dummy is also added as stronger environmental protection, as these states may compel their incumbents to resist flexibility provisions and thus tie their hands more strongly to IEA-mandated environmental standards. I also conduct several specification robustness checks with additional controls (e.g., foreign direct investment, green parties, and national capability) and those with dependent variable(s) that focus on specific components of IEA flexibility. I check whether my main results hold in samples in which IEAs are the unit of analysis and in samples without salient and controversial IEAs. Because the IEA Flexibility dependent variable is an ordered measure, I test H₁ by using ordered logit (OL) models with random effects and assess the econometric robustness of my results using different statistical tests, discussed subsequently.

Models 1–5 in Table 1 present the main results from numerous OL specifications estimated with random effects. Model 1 is a basic specification with minimal controls and the RWP Seats independent variable. The set of fully specified models that include the main control variables discussed earlier is reported in Models 2 and 3, while Models 4 and 5 compose the first set of specification robustness tests that incorporate additional controls.

The statistical association between RWP Seats and IEA Flexibility is positive and highly significant in Models 1–3 and remains positive and significant in the augmented specifications in Models 4 and 5. I turn to reporting and illustrating the substantive effect of RWP Seats on IEA Flexibility drawn from the fully specified models. To start with, this exercise reveals that a one-unit increase in RWP Seats leads, on average, to an increase of 1.564–2.035 in the ordered log-odds (or an increase of 4.777–7.652 in proportional odds ratio) of IEA Flexibility. Figure 4 illustrates the predicted probability of each ordered outcome of IEA Flexibility at varying levels of RWP Seats when the remaining covariates in the fully specified model are held at their sample mean values. Figure 4 first reveals that as RWP Seats increases, the predicted probability of the highest-ordered outcome of IEA Flexibility increases.

Importantly, Figure 4 shows that when RWP Seats crosses the threshold of majority legislative seats in the lower house of the national legislature, the statistical association between RWP Seats and IEA Flexibility is positive and highly significant (but is insignificant below the majority legislative seats threshold). Furthermore, increasing RWP Seats by 1 standard deviation from the sample mean of RWP Seats, which is approximately the majority legislative seat share in the lower house—holding other covariates constant at their mean or mode—increases the predicted probability of the highest-ordered category of IEA Flexibility (i.e., maximum flexibility provisions) by 1.7 percent. By contrast, the probabilities of the lower-level outcomes of 0 (no flexibility provisions) and 1 (minimal flexibility provisions) for IEA Flexibility decrease as the value of RWP Seats increases. Together, these results provide strong statistical and substantive support for H₁. The results for the control variables are mixed in that some controls (e.g., GDP Per Capita, GDP, Democratization, PR, CO₂ Emission) are statistically significant, while the remaining are insignificant.

I conducted several specification, econometric, and sample robustness tests that I briefly summarize, but I report the results from these robustness checks in Tables A.1–A.6 in the Supplementary Appendix to save space. First, in Models 4 and 5 in Table 1, I incorporate four additional controls that, as explained in the Supplementary Appendix, may influence the IEA Flexibility dependent variable: FDI (Foreign Direct Investment) Inflows, FDI Outflows, CO₂ Emission, Fuel Consumption, Urban Growth, and Industry Size (as share of GDP). Second, Models 6–8 in Tables 2 and A.2 (in the Supplementary Appendix) report results from specifications estimated on the sample in which IEAs are the unit of analysis. In the IEA unit-of-analysis models, I control for three IEA-level dummy variables, rationalized in the Supplementary Appendix, that operationalize whether IEA member states are neighboring states and have signed bilateral environmental agreements (BEAs) and free-trade agreements (FTAs). Third, in Models 9 and 10 in Table 2, I incorporate the interaction RWP Seats × Green Seats, where Green Seats operationalizes the seat share of green parties in the national legislature and the separate constitutive components of this interaction term.

In Model 11 (Table 2), I incorporate the RWP Seats × CINC interaction term, where CINC (Composite Indicator of National Capability) operationalizes the country’s national capability and is drawn from Singer (1988), and the separate constitutive components of the interaction term to assess whether RWP parties can influence the design of IEAs when they have greater capacity (i.e., are more powerful) than other IEA member states. Fourth, Model 12 in Table 2 reports the results from a specification in which I exclude the following salient and controversial IEAs from the sample: the Montreal Protocol and those associated with the UNFCCC (including the Kyoto Protocol and the Paris Climate Accords). In Table A.3 of the Supplementary Appendix, I also report additional specifications with controversial agreements omitted. Fifth, I assess the robustness of my results in numerous specifications in Tables 1 and 2 and Tables A.1–A.4 in the Supplementary Appendix, where I control for party system fractionalization, environmental performance, and the following system-level (i.e., international-level) controls described earlier: the EU, FTA, and BEA dummy variables. Sixth, Tables A.5 and A.6 present estimates from models that are estimated separately for the following dependent variables using data from the IEA database: the amendment measure from the IEA database, the IEA withdrawal measure, and reservation and withdrawal clauses.

The effect of RWP Seats on IEA Flexibility remains positive and statistically significant in all the aforementioned models in Tables 1 and 2 and Tables A.1–A.6 that are employed for specification and sample robustness tests. The impact of RWP Seats on both IEA amendment and reservation clauses is positive, as anticipated, but significant only in the case of amendments. The positive but insignificant association between RWP Seats and withdrawal broadly indicates, as suggested by my theory, that RWP parties threaten to—but do not—withdraw from IEAs. Furthermore, I find but do not report to save space that my results hold in Heckman selection-corrected ordered probit models.⁹

A growing body of research explains why flexibility provisions are incorporated into international agreements, particularly preferential trade agreements (Koremenos 2005; Mansfield and Milner 2012; Marcoux 2009; Morin et al. 2022). These studies undoubtedly provide rich insights into why and when flexibility provisions are incorporated into the context of designing preferential trade agreements—a salient issue in international politics—to reduce uncertainty for states and foster compliance with these agreements. But we know less about why or when democracies influence the design of international environmental agreements, which are less salient than trade agreements, as IEAs do not have direct, broad-based distributional consequences (as import-competition does) on society. Thus, unpacking the association between RWP parties and the design of flexibility provisions in IEAs is hard, given that IEAs are less politically consequential relative to trade agreements.

Yet, this study addresses this lacuna by exploring how the rise of RWP parties can account for variation in flexibility provisions in IEAs. The key prediction from my theoretical story is that when the share of legislative seats held by RWP parties in democracies is sufficiently high, they will successfully obtain and incorporate maximum flexibility provisions in IEAs.

My article contributes to the literature in numerous ways. First, my study is among the first to theoretically analyze the association between RWP parties in office and the design of IEAs. Unlike conventional wisdom—which suggests that RWPs renege from international treaties—my theory suggests that RWP parties remain committed to IEAs. Indeed, they obtain more flexibility provisions to remain in these agreements. As a second theoretical contribution, the study demonstrates how flexibility provisions in IEAs allow RWP parties to be “Janus faced”—on one hand, RWP parties threaten to renege from international agreements in the negotiation process; on the other hand, they continue to cooperate with other states when negotiating international agreements under the shadow of the aforementioned threat.

A key empirical contribution of this study is that it presents an original data set of seat shares held by RWP parties in a sample of democracies over the last four decades. Scholars of environmental governance and politics can thus employ this data set to evaluate how politically empowered RWP parties influence a variety of domestic environmental policy outcomes, including regulation of environmental standards. Furthermore, a vital but perhaps unexpected implication of my empirical results is that the rise of RWP parties does not lead to a breakdown of international cooperation; rather, it triggers the renegotiation of international agreements in the context of the global environment.

An important implication that emerges from this study is the potential deleterious impacts that RWP parties in office may have on international agreements, including IEAs. Given that IEAs coexist with various formal international institutions, such as the EU and the UN, the fact that RWP parties attack and undermine IEAs will likely jeopardize the legitimacy of these formal agreements. This raises serious doubts about the survivability of the international liberal order in the near future and beyond (Colgan et al. 2021; Lake et al. 2021; Walter 2021).

Future research can build on the findings presented herein. For instance, this study suggests that RWP parties use disinformation to manipulate domestic audiences in the context of international agreements. It will thus be worthwhile to examine whether voters in democracies are indeed successfully manipulated by RWP parties in the context of foreign policy. Finally, while I discussed the effects of ruling RWP parties on IEAs, the effects of left-wing populist parties on the design of IEAs are also worth exploring.

I am deeply grateful to Bumba Mukherjee and Vineeta Yadav for their invaluable guidance and advice. I also extend my heartfelt thanks to the editors of Global Environmental Politics and the anonymous reviewers for their constructive feedback. An early version of the article was presented at the International Studies Association 2023 annual convention in Montreal. Replication material is available upon request.