Why populism may facilitate non-state actors’ access to international environmental institutions

ABSTRACT

This article examines populists’ relationship with non-state actors in international environmental agreements (IEAs). We ask whether populist governments favor transnational non-state actors’ access to these institutions. Using data on the design of IEAs since the 1970s, evidence is reported suggesting that populists seek to institutionalize non-state actors’ access to IEAs. To explain this relationship, we argue that populist governments likely want to reduce and undermine the influence of established elites, also in international institutions. To this end, they may want to institutionalize access of their own constituents within IEAs. The empirical implication of this claim is that institutionalized access of non-state actors in IEAs is more likely when populist governments are involved. The empirical results provide strong and robust support for our argument, and these findings contribute to our knowledge in a number of areas of environmental politics and political science in general.

KEYWORDS:

• Populism
• international environmental agreements
• treaty design
• Non-State actors

There is increasingly consistent evidence that political populism harms the environment (Beeson 2019, Huber et al. 2021). Existing research on this issue shows that populist leadership (Böhmelt 2021, Jahn 2021) as well as attitudes (Huber 2020, Huber et al. 2020) and ideology (Lockwood 2018) can undermine environmental policy output and performance (see also Lockwood and Lockwood 2022). Case-based, qualitative, and anecdotal evidence supports this conclusion too: for example, Brazil’s President Jair Bolsonaro, who is widely described as a right-wing populist, ‘worked relentlessly and unapologetically to roll back enforcement of Brazil’s once-strict environmental protections.’¹ It was also under Bolsonaro’s leadership that Brazil withdrew from hosting the UNFCCC’s 2019 Conference of the Parties (COP) and, instead, opted for a significantly reduced delegation in Madrid where the annual meeting eventually took place – although Brazil is South America’s largest economy and a key emitter of greenhouse gases.² Interestingly, despite the reduced delegation size at the Madrid COP compared to previous years, the Brazilian delegation included a large number of members from various non-governmental organizations, among others, the Brazilian National Confederation of Industry and the Brazilian Petroleum Corporation (Petrobras) – non-state actors that are not directly affiliated with the government, but arguably have close ties to Bolsonaro’s administration.³

This article focuses on an underlying, more general institutional foundation of this link between populism and non-state actors’ access to international environmental agreements (IEAs): we examine whether and why populists – both left- and right-wing – may favor and institutionalize transnational non-state actors’ access (Tallberg et al. 2013, 2014, 2018, Green 2018) when designing IEAs. For defining these groups, we follow Tallberg et al. (2013, p. 1): the term ‘transnational actors’ thus encompasses non-governmental organizations, civil society organizations, social movements, philanthropic foundations, business associations, and multinational corporations. These groups can mainly be located domestically, but they must have some transnational interest or activity (broadly defined). Political populism commonly ‘considers society to be ultimately separated into two homogeneous and antagonistic groups, “the pure people” versus “the corrupt elite”, and […] argues that politics should be an expression of the volonté générale (general will) of the people’ (Mudde 2004, p. 543). This ideational⁴ perspective points to anti-elitism and threatened people’s sovereignty as the two essential traits of this ‘thin-centered ideology’ (Hawkins and Kaltwasser 2018; see also Akkerman et al. 2014). To this end, populism portrays elites as a predatory class detached from and opposed to a ‘morally pure and fully unified’ people, subscribes to a Manichean worldview in which the ‘people’ is good and elites are ‘evil’ (Hawkins 2003, Mudde 2004, Huber and Schimpf 2016); and, as it is people-centric, it also asserts that the people should make significant political decisions – not elites (Huber and Schimpf 2017, Hawkins and Kaltwasser 2018, Busby et al. 2019). Populists demand that the people ‘be given their voices back and that people should therefore have more influence on the political decision-making process’ (Rooduijn 2014, p. 576). Hence, the role of elites is, according to populists, not to use their electoral mandates to make independent decisions or to guide public debates, but ‘to listen to the people and translate what they find important into policy’ (Roodujin and Akkerman 2017, p. 194).

We refer to populism in general, i.e., both right- and left-wing populism, and contend that this view and these aforementioned features of populism have implications for international governance, including IEAs and their design as well as the inclusion of transnational actors therein. International environmental institutions, as constituted by IEAs, comprise staff and policymakers from what is usually seen as ‘the elite’ and these organizations and decision forums can affect people’s sovereignty – two aspects that populists directly oppose. Indeed, populist leaders often criticize international institutions as being elitist and insulated from the public, technocratic, and as interfering with the internal affairs of states. As a result, they commonly try to protect national sovereignty against external interference that is authorized and legitimized by international institutions (Börzel and Zürn 2021, Pevehouse 2020, Copelovitch and Pevehouse 2019; see also Jahn 2018, Lake et al. 2021). While differences exist across IEAs and factors such as state power certainly play a role, negotiations leading to the design and formation of an environmental regime are based on unanimous decision-making: each party has a veto right and, hence, only a design that all states can agree upon will make it into the final text. If not withdrawing or abstaining from such international fora completely,⁵ populist leaders could seek to design them in such a way that – or only join such institutions when – their interests are preserved, for example, by advocating more veto opportunities within these institutions (Pevehouse 2020, Börzel and Zürn 2021). By doing so, they aim at restoring their ‘people’s voice’ and state sovereignty. This implies that populist leaders will want to institutionalize non-state actors’ access in IEAs. On one hand, this increases the influence of their own constituents within such institutions, while lowering the access and power of the old elites they oppose. On the other hand, at the same time, this could serve as an indication that populists follow through on their promises that ‘nothing should constrain the will of the true people of a country’ (Kyle and Gultchin 2018, p. 3).

We empirically test this argument with quantitative data and find that IEAs are more likely to have institutionalized non-state actor access when countries that negotiated and ratified these agreements are governed by more populist executives. This result is robust to a series of changes in research design, variable specifications, and estimation procedures. The article contributes to four streams in the literature. First, and in our view most importantly, we add to the growing literature on environmental politics and populism (e.g., Lockwood 2018, Beeson 2019, Huber 2020, Huber et al. 2020, 2021, Böhmelt 2021, Buzogány and Mohamad-Klotzbach 2021, Jahn 2021, Lockwood and Lockwood 2022). Specifically, we provide a framework for deriving expectations about the implications of populism for global environmental governance, and, by focusing on the design of IEAs, we present empirical evidence that contributes to a better understanding of the phenomenon. Second, while the empirical findings reported here are relevant particularly to scholars of international environmental politics, they also contribute to the more general literature on the design of international institutions and the institutionalization of non-state actors’ influence in them (Tallberg et al. 2013, 2014, 2016, 2018, Böhmelt et al. 2014, Tallberg and Jönsson 2010; see also Koremenos et al. 2001, Mitchell et al. 2020). Our argument, in principle, applies to all policy fields in global governance and, thus, has implications for how international institutions generally are designed.

Third, most of the existing literature on populism focuses on national, i.e., domestic-level politics, policies, and outcomes (see Rooduijn 2019), such as polarization (Rooduijn et al. 2016), liberal democracy (Huber and Schimpf 2016), immigration (Akkerman 2012), the economy (Rodrik 2018), or environmental quality (Böhmelt 2021, Jahn 2021, Lockwood and Lockwood 2022). Only a small body of research examines the effects of populism and populist leadership beyond the nation state (e.g., Börzel and Zürn 2021, Copelovitch and Pevehouse 2019; see also Verbeek and Zaslove 2017, Jahn 2018, Lake et al. 2021). Our understanding of the impact of populism on political outcomes at the international level remains limited as a result. We contribute to addressing this shortcoming by studying one particularly important outcome at the international level: the design of IEAs and, specifically, the inclusion of non-state actors in these global governance institutions. Populist leaders may perceive foreign policy as a continuation of domestic affairs and, hence, criticize international institutions as being elitist and depriving their states of sovereignty. We show that this perception does indeed seem to apply.

Finally, we add to studies on transnational, non-state, and civil-society actors. There is a large literature on this subject in the context of global governance generally, as well as environmental politics (e.g., Tallberg et al. 2013, 2014, 2016, 2018, Böhmelt 2013a, 2013b, Böhmelt et al. 2014, Green 2018). While non-state actors’ access and influence are widely regarded as ‘normatively positive’ as they can promote policy outcomes or contribute to international institutions being seen as more legitimate (Bernauer et al. 2020), we shed light on a normatively negative aspect of non-state access to IEA decision-making, namely when this is shaped and driven by populist interests in the first place.

Why populism may be conducive to access of non-state actors in IEAs

Many of the most urgent environmental challenges facing humanity, including climate change, require cooperation and collaboration across countries (see Bernauer 2013, Campbell et al. 2019, Mitchell et al. 2020). States commonly establish international environmental agreements comprising multiple states to this end, presumably following a ‘rational design’ approach. That is, countries negotiate over how to design a particular IEA with a view toward achieving the most effective outcome at the lowest possible cost, focusing on key design aspects such as membership rules, obligations, dispute-resolution mechanisms, or financial provisions (Koremenos et al. 2001). Building on this framework, various studies examine the determinants of agreement designs in the first place and, in turn, how they affect states’ implementation and compliance costs as well as their willingness to formally join a particular treaty (e.g., Bernauer et al. 2010, 2013a, Böhmelt and Spilker 2016, Spilker and Koubi 2016, Böhmelt and Butkutė 2018).

Negotiations associated with the treaty design phase and, subsequently, considerations about whether to become a formal party to an IEA are usually executive matters. This means that the incumbent government first negotiates the specific design of an agreement, signs it, and then submits it to the national legislature for approval. Putnam’s (1988) ‘two-level game’ highlights the importance of the preferences of domestic audiences in this context of the bargaining process among state leaders and their agents (e.g., diplomats, ministers) in international fora. Political leaders, knowing that any agreement reached internationally requires backing by certain segments of the domestic audience ‘at home,’ have strong incentives to commit to agreements only if they are likely to survive the domestic approval stage without threatening their own political survival (Putnam 1988, Koubi et al. 2020). Especially populist leaders seek to mobilize ‘groups outside the system and attach new weight to their problems and particular issues’ (Huber and Schimpf 2016, p. 874, Mudde and Kaltwasser 2012, p. 21) to please domestic audiences. However, the design of IEAs does not only enable domestic audiences to draw inferences about the sincerity of their leader, i.e., whether their government pursues their interests internationally or not; it also allows leaders to push for an agreement whose design bolsters their own beliefs and, in the case of populists, their anti-elitist and sovereignty-preserving ideology.

International affairs are traditionally dominated by nation states and their governments. Yet, the increasing participation of transnational actors suggests that governance beyond the nationstate advances also in the sense of a pluralization of the type of actors involved (see Tallberg et al. 2013, 2014, 2016, 2018). Transnational non-state groups are now granted access to most of the major international institutions and governance fora, and they participate in the political decision-making processes of IEAs more and more (Raustiala 1997, Tallberg and Jönsson 2010, Tallberg et al. 2014, 2016, 2018). In line with this observation, recent research shows that treaty provisions allowing for greater non-state actor access to institutions’ meetings increase the likelihood of countries ratifying these agreements (Koubi et al. 2020). These developments may reflect efforts to address a widely claimed democracy and legitimacy deficit in global governance (Bernauer et al. 2020). The potential of non-state, civilsociety actors to mitigate this deficit lies partly in the perceived ability of such groups to act as watchdogs and better represent otherwise marginalized voices (Buntaine 2015). Empirical research also finds that civil society involvement positively affects public support for global governance (see Bernauer et al. 2020).

States allow non-state actors’ participation in IEAs and other institutions when it is functionally effective or cost-efficient to do so (Böhmelt 2013a, 2013b, Tallberg et al. 2013, 2014, 2016, 2018, Raustiala 1997). For example, a stronger involvement of democratic states in international organizations results in more non-state actors’ participation (Tallberg et al. 2016, p. 63), as granting access is not a ‘radical step’ and ‘the same procedural standards to all levels of political organization’ are simply applied. Moreover, building on rational-choice institutionalism, non-state actors are granted influence as they provide services to states in the form of information, resources, and expertise (Raustiala 1997, Tallberg et al. 2013, 2014). The potential for such transnational actors to play a supportive role in implementing IEAs is particularly beneficial in issue areas that are complex and require local execution. Here, states can choose to incorporate non-state actors to further their own regulatory powers, as non-state participation ‘provides policy advice, helps monitor commitments and delegations, minimizes ratification risk, and facilitates signaling between governments and constituents’ (Raustiala 1997, p. 720). In sum, granting and institutionalizing participation rights to non-state actors when negotiating the design of IEAs is the result of rational decisions by states based on considerations of functional effectiveness and cost efficiency.

Our argument focusing on populism and non-state actors’ access follows the rationale in Tallberg et al. (2016), as we contend that populist leaders’ domestic anti-elitism, people-centrism, and sovereignty concerns are likely to dovetail with a disregard for elitist international institutions they presume to be detached from the ‘real people.’ That is, populists are anti-elite at their core and may have the common perception that global affairs merely mirror elites’ interests, distant from the will of the people and ‘ordinary’ citizens (see also Börzel and Zürn 2021, Lake et al. 2021). Populist policymakers are thus likely to be skeptical of international arrangements to begin with, and this particularly applies to environmental institutions due to the complex and transnational nature often with the involvement of technocrats/scientists and the exclusion of the public (Huber 2020; see also Lockwood 2018, Böhmelt 2021, Huber et al. 2021; see also Buzogány and Mohamad-Klotzbach 2021, Jahn 2021). This presumably leads to a lack of responsiveness between elites and citizens’ demands, which populists exploit by claiming to be the only genuine representative of the people (Huber 2020, Huber et al. 2020, 2021). Busby et al. (2019) refer to this as ‘dispositional blame attribution.’⁶ If considering being part of IEAs at all, populists may want to only join them if the old elites are somehow undermined, national sovereignty is preserved, and what they see as the popular will is reflected.

Populist leaders are usually elected on an anti-establishment mandate. They oppose the policies of existing elites during their electoral campaign, and they pledge their allegiance to the ‘people’ (Hawkins and Kaltwasser 2018; see also Akkerman et al. 2014). Consequently, also with a view toward increasing the chances of staying in power given their constituent base, they are critical of the perceived cosmopolitan, globalist ideology of elites dominating IEAs and international institutions in general. In fact, populists’ political survival depends on signaling to their followers that they are catering first and foremost to their interests, which makes them try to curb global elites’ influence and secure more access of ‘the people’ in international fora. In addition, populist leaders perceive IEAs and their unelected overseers as depriving their country of the much-valued sovereignty, which contradicts populists’ fundamental claim, the ‘popular will’ (Hawkins 2003, Mudde 2004, Hawkins and Kaltwasser 2018; see also Huber and Schimpf 2016).

Populists, viewing IEAs as incurably corrupt and captured by distant elites, likely seek to advance the interests of their followers, the ‘real people,’ by weakening these institutions and the ‘old’ elites involved in them. If not deciding to withdraw or abstain from an agreement, we claim that this can motivate populist leaders to embrace more comprehensive non-state actors’ access to IEAs, even though they probably will use such institutionalized access for their own ‘anti-elite’ constituents including business and industry groups to circumvent old elites, including epistemic communities. Consider, for example, Brazil and Russia at the 2021 UNFCCC Conference of the Parties⁷: the former’s delegation included representatives of agribusiness, industry lobby organizations, and general business groups. The latter’s delegation comprised members leading banks like Sberbank and VEB as well as major corporations like Rosatom, Gazprom, Severstal (metallurgy), Inter RAO (electricity), and Sibur (coal).

Note that it is difficult, if not impossible, to find observable evidence that populists openly ‘confess’ that they may want to exploit the system and justify their stance on non-state actors by including ‘their’ own groups. However, even if less direct, some observable patterns allow for conclusions that are consistent with this behavior and our expectations. The evidence presented here and below shows that those groups close to populists benefit most from increased access for non-state actors when populists have negotiated the terms. In other countries, where populists are not in power, it seems that pro-environmental groups (such as Greenpeace) benefit most from increased transnational actor access.

The cases of Brazil and Russia clearly pertain to right-wing populists, but left-wing populist governments are equally likely to approach ‘their’ allies when it is about non-state actor access: Greece under SYRIZA included a number of left-wing academics, advisors, and policy experts in its delegation to the UNFCCC Conference of the Parties in 2018.⁸ Increasing access of non-state actors can weaken normal processes of governance through the addition of new voices and preferences, which may further exacerbate the collective action problems plaguing global governance (Pevehouse 2020, Börzel and Zürn 2021; see also Bernauer et al. 2013a). Consequently, the influence of global elites is lowered, while preserving national sovereignty. Those arguments point in the same direction. Hence, our empirical expectation is that populist government involvement in IEAs makes the institutionalized access of non-state actors more likely.

Research design

The IEA data used for our empirical analysis are taken from Koubi et al. (2020), who compiled time-series cross-sectional information on 178 international environmental agreements and their design characteristics in 1950–2011. Due to missing values of the explanatory variables, our final sample eventually comprises 88 agreements in 1970–2011. This data set is an extension of Spilker and Koubi (2016) who coded data on IEAs’ designs based on information from treaty documents in the International Environmental Agreements database (Mitchell et al. 2020).⁹ International treaties are included in our data if they were open for ratification globally and primarily deal with environmental issues. As the design of treaties does not change over time in the data by Koubi et al. (2020), we use a cross-sectional structure with the treaty-negotiating country as the unit of analysis. That is, for the following analysis, each treaty is paired with each country that participated in the negotiations of that agreement.

In addition, our identification strategy requires that a negotiating country must have ratified the treaty subsequently. There are several selection mechanisms at play: a few countries decide to negotiate an IEA. If they choose to participate in the negotiations, states may advocate for more or less non-state actor inclusion. Depending on the negotiation’s outcome, states decide to ratify the treaty or not. We cannot observe or test for all these selection processes, but instead have opted for an identification strategy that rules out the most common selection effects and allows us to get to the effect of populist negotiators and non-state actors’ access to international environmental agreements. While states with a populist leadership can influence treaty design when participating in the negotiations, this executive may still reject what has been agreed upon and not formally join an IEA in turn. By the same token, a succeeding government could object to what a populist executive has negotiated before and, also, not join a treaty.

Consider, for example, the UNFCCC: multiple countries negotiated this institution. States including the US, Brazil, or Spain not only were part of the negotiations, but also formally joined it since. Our data set comprises observations for those treaty-country combinations: UNFCCC-US, UNFCCC-Brazil, etc. If a country did not participate in the original negotiations leading to an agreement, or did not formally join it by 2011, it is not paired with a treaty and, hence, is not included in the analysis. Furthermore, we do not include repeated treaty-negotiating country observations over time and any time-variant covariate information is based on the year a treaty has been ratified. As a result, we focus on those treaties that (a) states negotiated and (b) they eventually ratified. We omit cases where country governments (including populists) were not involved negotiating or did not choose to ratify in the end. Ultimately, our identification strategy and data setup allow us to estimate the impact of populism on non-state actor access in IEAs with precision, as we circumvent problematic selection issues. At the same time, the empirical scope of our results is more limited as our findings apply to the set of countries that both negotiated and eventually ratified a treaty only. We return to this issue in the conclusion.

The outcome concept of interest pertains to the access of non-state actors to IEAs. Koubi et al. (2020) provide several variables for this. For the main analyses presented below, we opt for a dependent variable that codes in a binary fashion whether non-state, transnational, civil-society actors are granted institutionalized access to a treaty’s meetings (1) or not (0). As Tallberg et al. (2014) remind us, the access of non-state actors to international institutions comprises both quantitative and qualitative aspects. The online appendix thus also provides analyses based on alternative dependent variables that capture the extent to which a treaty institutionalizes non-state actor access. In our sample of 408 treaty-country observations (our unit of analysis), about 35% of them are coded as 1, i.e., there is institutionalized non-state actor access. Due to the binary nature of our dependent variable, we employ logistic regression models and cluster the standard errors by country to control for intra-group correlations.

The main explanatory variable, Populism Score, is taken from the Parties Variety and Organization (V-Party) project (Lührmann et al. 2020). These data code populism as closely as possible to our definition introduced above as there is a focus on anti-elitism and people-centrism (Mudde 2004). Lührmann et al. (2020) use expert surveys to assess parties’ populist positions. Each expert’s coding is based on party documents, e.g., election manifestos, press releases, official speeches, or interviews. For the main analysis below, we use the degree of populism as the core explanatory variable. This item is an index comprising the two factors of anti-elitism and people-centrism.

According to the V-Party codebook (Lührmann et al. 2020, p. 24), elites are defined as ‘relatively small groups that have a greater say in society than others, for instance due to their political power, wealth or societal standing. The specific groups considered to be the elite may vary by country and even from party to party within the same country as do the terms used to describe them. In some cases, “elites” can also refer to an international elite.’ The item codes ‘how important is anti-elite rhetoric for this party?’ on a scale from 0 to 4, with higher values standing for more anti-elitism. The variable on people-centrism can also receive values between 0 and 4 (higher values signify more people-centrism) and is based on: ‘[m]any parties and leaders make reference to the “people”, but only some party leaders describe the ordinary people specifically as a homogenous group and emphasize/claim that they are part of this group and represent it. This means that they do not acknowledge the existence of divergent interests and values in society, but rather suggest that the “people” have a unified political will, which should guide all political action. Often this group is glorified and romanticized, describing an ideal-typical ordinary person/commoner, who embodies the national ideal.’

On one hand, the focus on states’ executive heads is likely valid for parliamentary democracies as well. While, for example, the foreign secretary or environmental minister may not be from a party other than the prime minister’s platform, the latter usually is responsible for foreign-policy decision-making and ultimate decisions in this context. However, we also added a robustness check to the appendix where we distinguish between the populism score of major governmental partners (either the party of a single-party government or the Head of Government belongs to this party) and junior partners (average of the smaller parties the Head of Government does not belong to, but one or more cabinet ministers do, or the party is not represented in government at all).

Our final variable combines the two factors in an index and receives values between 0 and 1, with higher values standing for more populist positions of parties. Given our theoretical focus on state executives, leaders, and governments, we concentrate only on the party to which the head of government belongs to.¹⁰ In our data set, the Australian government party in 1977 had the least populist position (0.034), while Venezuela under Hugo Chávez receives the highest populist score (0.993).

We include several control variables at either the country or treaty level. First, we consider political power, which is captured by a country’s GDP (logged), which is directly taken from Koubi et al. (2020). The higher the GDP, the more influence a state has politically and, in turn, the more likely it can shape treaty negotiations according to its own interests (Bernauer et al. 2010, 2013b, Spilker and Koubi 2016). Second, wealth is measured by a state’s GDP per capita and is also taken from Koubi et al. (2020). We merely control for an income effect and, thus, do not seek to model the actual functional form of the influence of GDP per capita, e.g., along the lines of the Environmental Kuznets Curve (see Dasgupta et al. 2002). We return to this issue in the appendix, however. Third, regime type data are taken from the Polity V data and code countries on a − 10 to 10 scale in terms of their degree of democracy: higher values stand for more democratic countries. Böhmelt and Butkutė (2018) demonstrate that democratic regimes may be reluctant to implement hard-law treaty designs in environmental politics. And Tallberg et al. (2016) show that democratic states favor greater non-state actors’ participation in global governance. Fourth, trade openness is measured via the logarithm of the sum of the absolute shares of imports and exports to GDP from the Penn World Tables. States strongly embedded in the global trade network are more dependent on trade and industry production, making them to oppose more stringent environmental regulations also in international institutions (Bernauer et al. 2010). This could affect how they see and whether they favor institutionalized non-state actors’ access to IEAs (Tallberg et al. 2013, 2014, 2016). Finally, we control for an agreement’s environmental problem type to address the possibility that an institution’s underlying problem structure influences its design (e.g., Böhmelt and Pilster 2010, Mitchell et al. 2020). To this end, we include a dummy variable that indicates whether a treaty addresses (1) air pollution or not (0).¹¹ Air pollution is one of the most frequent issue areas the IEAs in our data deal with, while they are commonly affected by free-riding and other collective-action problems (Campbell et al. 2019). The descriptive statistics of the variables discussed so far are presented in Table 1.

Table 1. Descriptive statistics.

	Obs.	Mean	SD	Min	Max
Non-State Access	408	0.353	0.478	0	1
Populism Score	408	0.257	0.178	0.034	0.993
GDP	408	1.914	1.946	−3.675	6.384
GDP per capita	408	0.744	1.139	−2.613	2.477
Polity Score	408	6.765	5.274	−10	10
Trade Openness	408	−0.929	0.791	−3.262	1.551
Pollution	408	0.181	0.386	0	1

Empirical results

Table 2 presents our three main models. Model 1 focuses on the Populism Score, and we omit any control variables. Model 2 includes the controls, but we do not include the main explanatory variable. Model 3 is our full model that comprises all explanatory variables discussed in the previous section. The table entries are coefficients, which allow for a direct reading of their direction and statistical significance. In Figures 1–3, we present substantive quantities of interest to assess each variable’s substantive effect: Figure 1 shows predicted probabilities, Figure 2 summarizes simulated predicted probabilities for different values of the Populism Score, and Figure 3 presents first difference estimates for the control covariates.

Figure 1. Substantive effects of populism score.

Notes. Graph displays predicted probabilities of Non-State Access (based on Model 3). The dashed lines are 90% confidence intervals.

Figure 2. Simulated predicted probabilities.

Notes. Graph displays distribution of simulated predicted probabilities of Non-State Access (N = 1,000 simulations). Horizontal bars capture simulated probabilities’ point estimates and 90% confidence intervals. Values are calculated for Model 3 while holding all other covariates constant at medians.

Figure 3. First difference estimates.

Notes. Graph displays first differences and 90% confidence intervals. Calculations based on Model 3 and done when changing a variable from its 25^th percentile to its 75^th percentile (minimum to maximum for Pollution). All other variables held constant at medians. First difference of 0 marked by red vertical line.

Table 2. Institutionalized non-state actors’ access to IEAs and populism.

	Model 1	Model 2	Model 3
Populism Score	1.375**		1.235**
	(0.606)		(0.626)
GDP		−0.150*	−0.149**
		(0.077)	(0.076)
GDP per capita		0.068	0.099
		(0.158)	(0.156)
Polity Score		0.025	0.015
		(0.024)	(0.024)
Trade Openness		−0.336*	−0.260
		(0.174)	(0.170)
Pollution		0.446	0.486*
		(0.293)	(0.296)
Constant	−0.966***	−0.943***	−1.162***
	(0.171)	(0.293)	(0.257)
Observations	408	408	408
Log Pseudolikelihood	−262.042	−260.329	−258.310
Prob > χ^2	0.023	0.106	0.011

Table entries are coefficients; robust standard errors clustered at the country level in parentheses.

* p < 0.10; ** p < 0.05; *** p < 0.01 (two-tailed).

Focusing on the core variable of interest, Table 2 shows that Populism Score is positively signed and statistically significant at conventional levels. This effect estimate is robust independent from model specifications. Hence, more populist leaders involved in IEAs are indeed more likely to be linked to institutionalized non-state access in IEA designs. In substantive terms (Figure 1), a value of 0 for Populism Score suggests that a negotiating country is hardly populist at all – the probability to see institutionalized non-state actors’ access in an IEA such a state has ratified is only around 28%. However, this likelihood to see non-state actor access in a treaty’s design increases to almost 60% when Populism Score is set to 1, i.e., there was an extremely populist government participating in the negotiations.

In Figure 2, we simulate the predicted probability of Non-State Access equaling 1 for two scenarios: a populist government having participated in the negotiations and a non-populist government negotiating an agreement. This allows us to define with greater precision the probability of institutionalized non-state actors’ access in an IEA considering the populism score of governments participating in negotiations. The probabilities are simulated quantities of interest and are based on 1,000 draws from a normal distribution. In this graph, the simulated predicted probabilities of the two scenarios do not overlap: the confidence intervals of the point estimates at the bottom of the graph are distinct from each other. This mirrors the statistically significant coefficient estimate in Table 2. The simulated probabilities of Non-State Access for a non-populist government then converge to about 0.28, which mirrors the estimate from Figure 1. For populist governments, this probability increases to about 0.57, which is again similar to what we have calculated above. Figure 2 also shows that there is only a very small portion of simulated probabilities that overlap between the two scenarios. This further strengthens confidence in the result that the likelihood of institutionalized non-state actors’ access in an IEA increases with a higher populism score of governments participating in negotiations.

Linking these findings back to the theoretical argument, we find strong support for an association between populism and transnational, non-state access to IEAs. Populists oppose the old and established elites, they seek to preserve the will of the people and, hence, state sovereignty – also at the international level. In the context of environmental global governance, this implies that populists may aim at designing institutions in such a way that – or only joining them when – the influence of the ‘old elites’ is lowered. Indeed, one way of achieving this is to institutionalize the access of non-state actors in IEAs, which allows for the inclusion of populists’ constituents instead. Despite the general pattern that civil society can facilitate environmental policies and outcomes, however, we believe that populists favoring non-state actors’ access in IEAs will have the opposite effect. We return to this in the conclusion.

The appendix summarizes several robustness checks that further increase the confidence in our main result. First, we consider several alternative measures and operationalizations of non-state actors’ access to IEAs as coded in Koubi et al. (2020). Second, we model the influence of spatial diffusion effects by including region fixed effects. Third, we control for the total number of negotiating parties of a treaty. Fourth, we explore different types of standard errors. Fifth, we distinguish between the two mechanisms of our argument, i.e., anti-elitism and people-centrism, by disaggregating Populism Score. This latter analysis shows that anti-elitism seems to be the driving force behind our results. Sixth, we model a possible curvilinear impact of GDP per capita, additionally control for public-goods problem types, and consider further environmental issue dummy variables. Finally, while we focus on populism in general here, the appendix considers different government compositions and distinguishes also between leftist and rightist populists. All these additional analyses further support the core finding reported here.

Coming to the control variables, Figure 3 presents changes in the predicted probability of Non-State Access equaling 1 when altering a given explanatory variable. Several interesting results emerge from it. First, GDP is negatively signed and significant. The substance of the effect suggests that moving from the 25^th percentile to the 75^th percentile of GDP lowers the likelihood of a treaty-country combination having institutionalized non-state actor access by about 9 percentage points. This finding reflects the argument that particularly powerful states are concerned about sovereignty and will try to avoid cuts into their decision-making power (Bernauer et al. 2010, 2013b). Second, the variables for income, regime type, and trade openness are insignificant. Third, agreements dealing with air pollution are particularly more likely to see non-state actors’ access institutionalized. These estimates mirror the claim that different issue areas are characterized by different problem structures, which in turn shape IEA design in certain ways (e.g., Böhmelt and Pilster 2010, Mitchell et al. 2020). On average, air pollution agreements paired with states have a probability of non-state actor access that is 12 percentage points higher than other issue areas.

Conclusion

Populist leaders perceive domestic politics as being profoundly characterized by a struggle between the ‘pure people’ and a ‘corrupt elite’ (Mudde 2004). As a result, anti-elitism and the belief that people’s sovereignty is under threat are essential traits of this ‘thin-centered ideology’ (Hawkins and Kaltwasser 2018; see also Akkerman et al. 2014). In line with other research (e.g., Verbeek and Zaslove 2017, Jahn 2018, Copelovitch and Pevehouse 2019), we argue that these features spill over to international politics, where populist leaders criticize international institutions as being elitist and insulated from the public, technocratic, and intrusive in their internal affairs (Börzel and Zürn 2021, Lake et al. 2021). Consequently, when considering the participation in global-governance institutions at all, populists would seek to ensure that veto opportunities within these organizations exist that can lower the influence of the established elites (Pevehouse 2020, Börzel and Zürn 2021). We have claimed that populist leaders may pursue this objective by, inter alia, institutionalizing the access of non-state actors in the design of international institutions.

Using cross-sectional data with information on the design of IEAs, we find that non-state, transnational actors’ access in IEAs’ designs becomes more likely with a higher populism score of the governments involved in the negotiations. This result adds an additional – and, depending on the perspective, less positive – explanation beyond the standard functional justification found in the literature for the increased access of non-state actors in global governance (e.g., Tallberg and Jönsson 2010, Tallberg et al. 2013, 2014, 2016, 2018, Böhmelt et al. 2014). Hence, we hope to have made relevant contributions to the literatures on environmental politics and populism (e.g., Lockwood 2018, Beeson 2019, Huber 2020, Huber et al. 2020, 2021, Böhmelt 2021, Jahn 2021, Lockwood and Lockwood 2022), the design of international institutions (Koremenos et al. 2001), populism beyond domestic politics (e.g., Börzel and Zürn 2021, Lake et al. 2021), and on non-state, transnational, and civil society actors (e.g., Tallberg et al. 2013, 2014, 2016, 2018, Böhmelt 2013a, 2013b, Böhmelt et al. 2014, Green 2018).

The observation that populism can lead to more access of non-state actors in IEAs may appear counterintuitive, particularly to those regarding civil society involvement in global governance as conducive to problem solving (e.g., Böhmelt 2013a, 2013b, Böhmelt et al. 2014, Tallberg and Jönsson 2010, Tallberg et al. 2014, Bernauer et al. 2020; see also Raustiala 1997), and also in view of research showing that populism leads to lower national environmental performance and legislative action that does not favor green interests (see, e.g., Lockwood 2018, Beeson 2019, Böhmelt 2021, Jahn 2021, Lockwood and Lockwood 2022). In our case, though, institutionalized non-state actors’ access in global environmental governance is unlikely to promote pro-environmental legislation, performance, and – eventually – quality as populists merely tend to use this avenue of influence to grant their own ‘anti-elite’ constituents more power – as indicated by the examples of Brazil or Russia above.

With all that said, it is unclear to what extent non-state actors’ access contributes to the fulfillment of populist leaders’ interests. And while we have focused on IEAs, our argument should apply to all sorts of policy fields in global governance, but we currently lack the data to explore this. Similarly, the data at hand allow for conclusions about a rather narrowly defined set of actors only: states that negotiated and eventually ratified an IEA. We cannot directly assess the influence of populism at various other stages of the (selection) process leading to the design and ratification of environmental agreements. More in-depth, qualitative analysis will be required here. Finally, analyzing populists’ preferences for IEA designs in a more nuanced way, potentially with micro-level survey data, would constitute an interesting future study that could cross-check our findings at the individual level. We believe that those streams constitute promising venues for future research, which can overcome the limitations of our design or focus on the implications of our findings.

Acknowledgments

We thank the journal’s editor, Sherilyn MacGregor, and the anonymous reviewers for their helpful comments and feedback.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09644016.2022.2105055.

Notes

1. See online at: https://www.nytimes.com/2019/08/27/world/americas/bolsonaro-brazil-environment.html.

2. See online at: https://news.mongabay.com/2019/12/cop25-brazils-official-presence-diverges-widely-from-its-public-persona.

3. Available online at: https://unfccc.int/documents/184482.

4. Populism is often combined with other ideologies, such as nationalism, liberalism, or religious fundamentalism and, therefore, populists can be found at different locations along the left-right political spectrum (e.g., Akkerman et al. 2017, Rooduijn and Akkerman 2017).

5. Indeed, populist governments may prefer to withdraw from IEAs as well or abstain from them completely. This is not within the scope of our article, though. Instead, we focus on those cases where populists negotiate such institutions and prefer to stay. Remaining at the bargaining table can be more beneficial because this gives influence on the further development of an IEA, compared to being absent. Even the US under the Trump administration remained part of the UNFCCC, for instance.

6. As Huber et al. (2020: 2) conclude, ‘populist individuals are more likely to oppose the central actors involved in top-down climate policy (namely, the political elites that adopt climate policies) and are likely to reject climate-change-related mitigation policies proposed by the same actors.’

7. See online at: https://unfccc.int/documents/323052.

8. Available online at: https://unfccc.int/documents/187488. Mudde (2004) discusses how the main characteristics of populism are the same across left-wing and right-wing ideologies. We nonetheless distinguish between left-wing populists and right-wing populists in a robustness check in the appendix.

9. See online at: https://iea.uoregon.edu.

10. On one hand, the focus on states’ executive heads is likely valid for parliamentary democracies as well. While, for example, the foreign secretary or environmental minister may not be from a party other than the prime minister’s platform, the latter usually is responsible for foreign-policy decision-making and ultimate decisions in this context. However, we also added a robustness check to the appendix where we distinguish between the populism score of major governmental partners (either the party of a single-party government or the Head of Government belongs to this party) and junior partners (average of the smaller parties the Head of Government does not belong to, but one or more cabinet ministers do, or the party is not represented in government at all).

11. See online at: https://iea.uoregon.edu/international-environmental-agreements-ieas-defined.