The Public Legitimacy of Multistakeholder Partnerships in Global Environmental Governance: Evidence from Survey Experiments in Brazil, the United Kingdom, and the United States

Partnerships between public, civil society and private actors have become one of the central features of global environmental governance and the sustainable development agenda (Andonova 2010; Andonova et al. 2009; Bernstein 2013, 20; Visseren-Hamakers et al. 2011). Examples of multistakeholder partnerships (MSPs) addressing environmental and sustainable development issues include the World Commission on Dams, the Renewable Energy and Energy Efficiency Partnership, and the Climate, Community and Biodiversity Alliance. The United Nations (UN) and other international organizations (IOs) have turned toward MSPs as important mechanisms, or even panaceas, for collective problem solving across various policy fields. This is illustrated by the prevalence of MSPs as implementation mechanisms with regard to the millennium development goals, the UN Global Compact, and the 2030 Agenda and sustainable development goals (SDGs). The 2030 Agenda explicitly designates the seventeenth SDG to the revitalization and mobilization of partnerships as key governance mechanisms to accelerate implementation of sustainable development (Beisheim and Simon 2018). In the UN secretary-general report Our Common Agenda, launched in 2021, a key recommendation is to boost partnerships between the private sector, subnational actors, and civil society. The call for “networked multilateralism,” where MSPs will play a central role, was reaffirmed by the UN High-Level Advisory Board for Effective Multilateralism (2023). Currently virtually all IOs have partnership programs. Because of MSPs’ hybrid, cross-sectoral, flexible, and horizontal governance structures—as well as the potential to pool resources and competencies from multiple actors—they are expected to be effective and inclusive (Andonova 2017; Gleckman 2018; Gray and Purdy 2018; Neuner 2020; Schäferhoff et al. 2009; Sun 2017; Westerwinter 2021).

A recurrent claim is that MSPs increase the legitimacy of global sustainable governance as they fill implementation gaps left by governments and reduce the democratic deficit by being more transparent, accountable, and inclusive than intergovernmental arrangements (Bäckstrand et al. 2022). Critics, however, question the legitimacy of MSPs due to the alleged lack of transparency, effectiveness, dominance of corporations (Andonova et al. 2022; Pattberg and Widerberg 2016), and hierarchical decision-making dynamics and norms (Chowdhury 2023; Gray et al. 2022; Saffer et al. 2018). The promotion of MSPs by governments and IOs also raises the question of how MSPs—with their diverse sets of goals and features—are perceived by the public. For example, do citizens have higher confidence in MSPs that are led by corporations, governments, or civil society organizations (CSOs)? Are citizens more supportive of MSPs that achieve their goals through advocacy or through service provision? Put more broadly, what features of MSPs matter to the public? While previous studies have focused on whether MSPs fulfill procedural standards and are effective in solving sustainability problems (Andonova et al. 2022; Bäckstrand 2008; Kramarz 2020; Pattberg et al. 2012; Take 2012), our study is the first to explore citizens’ attitudes toward MSPs as tools to tackle global environmental problems. As such, we contribute to the scholarship on MSP for sustainability and public opinion of global governance institutions (Edwards 2009; Tallberg and Zürn 2019).

Unpacking the sources of public legitimacy of MSPs is an important task. Public legitimacy for national and global institutions is considered a precondition for exercising authority (Buchanan and Keohane 2006). For example, a lack of public support may create difficulties for MSPs in attracting the necessary resources for successful implementation (Sommerer et al. 2022). It may also discourage the participation of citizens and local CSOs in MSPs. Moreover, understanding the sources of public support is important because citizens’ attitudes toward different types of MSPs may influence decision makers’ preferences in global governance (Copelovitch et al. 2019).

Our study goes beyond the normative legitimacy, which evaluates whether MSPs fulfill democratic legitimacy in terms of inclusion, accountability, and transparency (Bäckstrand 2008; Bäckstrand and Kylsäter 2014). Instead, we focus on the sociological or public legitimacy of MSPs, assessing the features that garner public support (Dellmuth and Tallberg 2023). We focus particularly on MSPs in global sustainability governance, defined as “voluntary agreements between public actors (IOs, states or sub-state public authorities) and non-state actors (non-governmental organizations, companies, foundations) on a set of governance objectives and norms, rules, practices, or implementation procedures and their attainment across multiple jurisdictions and levels of governance” (Andonova 2017, 2).¹ To identify which institutional features of MSPs are likely to affect citizens’ perceptions of them, we rely on existing MSP scholarship, the literature on IO legitimacy in global governance, and discussions in UN policy circles (United Nations 2021; United Nations Department of Economic and Social Affairs 2015). Building on global environmental governance and IO literature, we identify six features that may shape citizens’ support for MSPs: performance (Andonova et al. 2022; Marx 2019; Pattberg and Widerberg 2016), transparency (Grigorescu 2020; Kramarz 2020), inclusion (Ellersiek and Beisheim 2017), leadership (Kolcava et al. 2021; Pattberg et al. 2012), social purpose (Andonova 2017), and governance function (Bull and McNeill 2007). Arguably, we cover the most salient features identified in the literature; however, additional important features of MSPs may influence public opinion, to be explored in future research.

To test whether and how these MSP features affect public support, we employ an original population-based conjoint survey experiment involving 6,056 respondents in three countries, including Brazil (2,015 respondents), the United Kingdom (2,029), and the United States (2,012), conducted between December 2020 and January 2021. The conjoint design allows us to examine how the various governance features of MSPs affect popular legitimacy. In our conjoint framework, respondents were asked to rank hypothetical MSPs with the six specified features, enabling us to estimate the influence of each feature on respondents’ support for MSPs. In selecting countries for the conjoint experiments, our ambition was to include countries from across the Global North and South that are significant players in multilateral diplomacy and global environmental governance.

Our study is the first to assess the sociological legitimacy of global MSPs by employing experimental survey methods and thereby offers novel insights and implications for understanding MSPs in global environmental politics. In terms of policy implications, the study reveals which institutional features of MSPs the public favors. Enhanced knowledge in this area can lead to more informed policy proposals concerning the development and implementation of MSPs for environmental and sustainable development. In the following sections, we briefly outline the expanding scholarship on MSPs and present our theoretical expectations. We proceed by outlining the study’s research design, followed by a presentation of the findings from our conjoint survey experiments. Finally, in the concluding section, we discuss the implications of our findings for the public legitimacy of MSPs and their role in global environmental governance.

Partnerships are increasingly gaining prominence in global environmental governance and sustainable development (Filho et al. 2022). Two diverging views on the legitimacy of MSPs emerge in the literature. In the first account, MSPs are viewed as problem-solving institutions that can increase effectiveness in global environmental governance by including public and private actors in joint agenda setting, decision-making, and implementation (Börzel and Risse 2005; Cashore et al. 2021; Kramarz 2020; United Nations 2021). As such, the public legitimacy of MSPs stems from both procedural values (e.g., transparency and inclusion) and effective outcomes (e.g., performance and efficiency) (Tallberg and Zürn 2019).

In contrast, the second account questions the legitimacy of MSPs based on the same procedural and outcome qualities. Critics here raise concerns regarding MSPs’ limited accountability, power imbalances (Saffer et al. 2018), and lack of inclusion (Andonova 2017, 6; Gideon and Unterhalter 2017; McKeon 2017; Zammit 2003). Moreover, empirical studies of the performance and effectiveness of MSPs find insufficient evidence to support the proponents’ claims of their efficiency and effectiveness (Andonova and Levy 2003; Beisheim and Liese 2014; Bull and McNeill 2007; Glasbergen et al. 2007; Kramarz 2020; Pattberg and Widerberg 2016). Empirical studies of more than 300 MSPs for sustainable development, adopted at the 2002 World Summit on Sustainable Development, have shown that the partnerships have had no or little impact on global governance (Andonova and Levy 2003; Pattberg et al. 2012). Subsequent studies of MSPs, adopted in conjunction with the UN Rio+20 summit in 2012, have also demonstrated limited effectiveness (Bäckstrand and Kylsäter 2014). Recent studies discuss the methodological limitations of previous studies with regard to effectiveness that allegedly have been overlooked in earlier research (Andonova et al. 2022). Critics of MSPs in global governance and the 2030 Agenda question their legitimacy with reference to the corporate influence that allows a business to steer the UN (Bruno and Karliner 2002, 34; Bull and McNeill 2019; Taggart and Abraham 2023). While there is a proliferating scholarship on the normative legitimacy of MSPs, there is a lack of studies exploring sociological legitimacy, that is, citizen perceptions, although citizens are the ultimate beneficiaries of MSPs. It is therefore important to investigate this issue, as it can provide valuable information about how MSPs are perceived by the communities they serve.

Scholars have paid considerable attention to the public support, legitimacy, and legitimation of IOs, multistakeholder institutions, and MSPs (Bäckstrand and Kylsäter 2014; Bexell et al. 2022; Gleckman 2018). Studies have explored how various attributes of international climate agreements (Bechtel and Scheve 2013) and intergovernmental organizations (Dellmuth and Tallberg 2023) may influence citizens’ perceptions. This literature emphasizes the importance of legitimacy and public support for IOs and governance innovations like MSPs to exercise authority, garner resources, generate compliance, and solve transnational global environmental problems. However, the current literature on public opinion in global governance has focused predominantly on traditional IOs, thus overlooking new modes of public–private governance, such as MSPs. But MSPs—whether they are international institutions in themselves, such as COVAX, or hosted by the UN—are also dependent on public support to secure resources, exercise authority, and achieve their goals. In the next section, we draw on existing scholarship on public support of global governance institutions to identify which features or attributes of MSPs affect citizens’ support for them.

When evaluating the public legitimacy of MSPs, we consider institutional features advanced in scholarly research on IOs and MSPs. These features are also mirrored in policy debates on how to improve or strengthen MSPs. The identification of institutional features is informed mainly by the MSP literature but also by recent studies on the public support of global governance institutions. We focus on six institutional features or attributes, which we label performance, transparency, inclusion, leadership, purpose, and governance function. We define performance of MSPs in terms of their problem-solving effectiveness (Dellmuth and Tallberg 2023), while transparency pertains to whether information about MSPs’ decisions and processes is publicly available (Bäckstrand et al. 2022; Koliev and Bäckstrand 2022). Inclusion relates to MSPs’ decision-making procedures and, more specifically, whether these involve MSP stakeholders or if more resourceful actors are allowed to dominate in decision-making contexts (Pattberg et al. 2012).

Our expectation is that citizens will show greater support for MSPs perceived as effective, transparent, and inclusive. Similar to studies of legitimacy of IOs (Dellmuth et al. 2019), performance, transparency, and inclusion are among the most highlighted features affecting the public legitimacy of MSPs (Bäckstrand 2008; Beisheim and Liese 2014; Mena and Palazzo 2012), private environmental governance (Neuner 2020), and UN multistakeholder arrangements in general (Beisheim and Simon 2018; Gleckman 2018). In the scholarly literature, performance is referred to as the outcome-based (e.g., problem-solving features) dimension of public legitimacy, while transparency and inclusion pertain to the procedural or input-based (e.g., rules and processes of organizations) dimension of legitimacy (Bernauer and Gampfer 2013). Previous experimental studies have shown that both dimensions matter for the public when evaluating organizations (Bernauer et al. 2020; Dellmuth et al. 2019; Tallberg and Zürn 2019). These three features are also part of a broader narrative around MSPs as new models of governance with promising enhanced effectiveness, transparency, and inclusion, in contrast to intergovernmental agreements. These narratives articulated by global political elites in UN policy fora can be conceived of as attempts to legitimize MSPs among citizens to increase acceptance and support.

The expectation that citizens would prefer effective global governance arrangements over noneffective ones is self-evident. However, in our conjoint experiment, citizens evaluate and compare features of MSPs in relation to each other. Several studies (Bernauer and Gampfer 2013; Edwards 2009; Esaiasson et al. 2012) suggest that the importance of features like effectiveness, transparency, and inclusion may vary among citizens. While we anticipate that these three features will be crucial in evaluating MSPs, we remain agnostic as to which will carry the most weight.

While MSPs typically involve various stakeholders, they are often managed by one specific private or public actor that leads and coordinates the work. For instance, the Partnership for Action on a Green Economy (PAGE) is a UN-led MSP, whereas the Global Partnership for Education is a nongovernmental organization (NGO)-led MSP. And although the Better Work Partnership includes various stakeholders, it is financed and managed by two IOs: the International Labour Organization and the International Finance Corporation. We assume that the leadership of MSPs will matter for citizens’ perceptions because they may prefer some actors over others based on their prior beliefs. Indeed, existing studies provide us with reason to assume that the type of actor leading an MSP can be a source of disapproval of or support for the MSP. For example, citizens may favor international institutions in which civil society actors are included and have influence over decision-making (Bernauer and Gampfer 2013). In the context of this study, citizens may prefer MSPs led by civil society, as they are commonly perceived as representatives of citizens’ interests and ideas. Moreover, citizens may evaluate them positively if they are seen as representing impartiality and as possessing information and skills of value to MSPs (Zammit 2003). MSPs that are led by national governments can also be positively evaluated by citizens. As Dellmuth and Tallberg (2015, 456) argued, citizens may view government involvement as a way to strengthen citizen interests.

On the other hand, MSP leadership may also be a source of contestation, in general or depending on citizens’ political orientations. The ongoing debate on whether MSPs consolidate corporate power, privatization, and marketization of global environmental governance (Carmin et al. 2003, 537; Pattberg and Widerberg 2016, 44) may also be reflected in public opinion. For example, citizens may disapprove of business-led MSPs as businesses may be perceived to use MSPs as an opportunity to symbolically respond to citizens’ preferences for sustainable products and to protect their own profit-based interests (Darnall et al. 2017; see also Darnall et al. 2010; van der Ven 2023). Citizens’ political beliefs, however, may be especially important here to understanding the public support for MSPs. The reasoning is that citizens are likely to view MSP leadership as political, particularly when it is led by actors like businesses, governments, or faith-based organizations, which often polarize citizens along left–right political affiliations. This assumption aligns with previous research on the impact of political values on citizens’ attitudes toward IOs (Edwards 2009; Hooghe et al. 2019). Thus we anticipate that citizens’ support for MSP leadership will be influenced by their political views.

Previous research suggests that support for global governance institutions also may be influenced by the overarching goal or purpose of the organizations (Scholte and Tallberg 2018, 64). The assumption here is that citizens evaluate organizations not only based on their effectiveness and how they work to achieve their goals but also based on which goals they pursue. Dellmuth and Tallberg (2023) provide the first experimental evidence of the relationship between an IO’s social purpose and public support. They argue that the social purposes of IOs have an independent effect on citizens’ perceptions of IOs. The survey experiments conducted in the United States and Germany suggest that citizens favor IOs that promote peace and security and dislike IOs that promote free trade and mitigate climate change (Dellmuth and Tallberg 2023).

MSPs in global sustainability governance are often criticized for their focus on the promotion of trade and business over issues like human rights or the just distribution of resources (Bull and McNeill 2019; McKeon 2017). This is in line with the results of general surveys measuring the preferences of citizens regarding global policy areas. Citizens generally care more about broader social issues, such as poverty, health, and security. For instance, in the MyWorld UN Global Survey, with more than 10 million respondents, citizens overall identify health, poverty reduction, and economic prosperity as the most important issues. A recent survey by Ipsos (2021) on the SDGs found that 20,000 citizens from twenty-eight countries ranked SDGs, where 1 was no poverty and 2 was zero hunger, as their top priorities. Furthermore, and in line with previous studies, we expect that the political ideology of citizens will matter here as well. Citizens’ evaluations of the social purposes of MSPs—whether to promote gender equality or to combat climate change—are likely to be politically contested. And, as Dellmuth and Tallberg (2023, 183) state, “when citizens hold a particular ideological orientation, this offers them a shortcut to political positions on a whole range of issues interpretable in ideological terms.” Depending on whether someone holds left- or right-leaning political views, that person may have different opinions on the salience and legitimacy of social purpose, such as promoting trade or reducing poverty. Following this logic, we expect that the information provided about the social purposes of MSPs will have an impact on citizens’ perceptions of them.

Finally, because MSPs have various governance functions—advocacy, rule making, service provision, implementation—we have reasons to assume that the governance functions of MSPs will matter for their public support because these determine how the problems are addressed and the goals are achieved (see, e.g., Andonova 2017; Bulkeley et al. 2014). Bäckstrand et al. (2022) find that most MSPs addressing the SDGs have governance functions described as knowledge provision and awareness raising, followed by capacity building and advocacy. Citizens may, for different reasons, view certain governance functions as more suitable or legitimate than others. For example, they may consider capacity building to be a more appropriate approach than information campaigns for MSPs working to eliminatechild labor. Moreover, the governance functions can indicate to the public whether MSPs adopt a bottom-up or top-down approach. Bottom-up governance functions, such as raising funds for project investments, increasing knowledge, and providing on-the-ground services, may be perceived as more empowering and sustainable. In contrast, top-down governance functions like advocacy, lobbying, and setting new standards may be seen as more prescriptive. The bottom-up approach is often viewed as more democratic and participatory (DeLeon and DeLeon 2002). Alternatively, the governance function may not be of importance for citizens’ perceptions, placing more weight on other features, such as perceived effectiveness, inclusion, or transparency. We note, however, that while we expect the governance functions of MSPs to influence citizens’ perceptions, we remain agnostic as to which type of governance function will matter most or least.

To test our expectations, we conducted a population-based conjoint survey experiment. Conjoint survey experiments are particularly suitable for our purpose as they allow us to establish whether various attributes matter for citizens’ support for MSPs (Bansak et al. 2016; Hainmueller et al. 2014; Leeper et al. 2020). We fielded our experiments in Brazil (n = 2,015), the United Kingdom (n = 2,029), and the United States (n = 2,012). We based the selection of countries for our survey experiments on three criteria. First, we focused on countries that are significant actors in global sustainability governance. Second, we aimed to choose countries from different regions across the Global North and South to address the bias toward the United States observed in many previous survey experiments (Bechtel and Scheve 2013; Neuner 2020). Finally, we selected countries that, within their respective regions, have relatively high levels of internet access, an important prerequisite for creating representative online panels.

The online panel survey was conducted by the well-reputed global survey company YouGov between December 2020 and January 2021. YouGov uses targeted quota sampling to approximate a representative sample. In our sample, the quotas were set based on age, education, gender, and region. For detailed information about the survey and the robustness checks, see Figures A1 and A6 in the online appendix. Moreover, we opted to provide hypothetical MSPs rather than actual ones because abstraction serves the purpose of this study, that is, to examine if various features of MSPs matter to citizens’ perceptions (Brutger et al. 2023).

We asked respondents in each country to choose between two hypothetical profiles of MSPs with randomly assigned attribute values. The attributes are the features of MSPs that are expected to affect citizens’ support, while the attribute values are discrete categories defining relevant values that can be attached to a certain attribute (see Table 1). The attributes associated with MSPs are also randomly ordered for each respondent but kept constant per respondent to avoid confusion. Each respondent was asked to evaluate a total of four comparisons with two hypothetical MSPs. Understanding sociological legitimacy as an attitudinal phenomenon, and in line with recent studies assessing public legitimacy of international treaties and global governance institutions (Bechtel and Scheve 2013; Dellmuth and Tallberg 2015), respondents were asked to indicate one MSP in which they had the “most confidence” (see Figure A1).²

Before the conjoint experiment, we presented the respondents with an introductory text to MSPs (Figure 1). We recognize that respondents may be less familiar with the concept of global partnerships than with, say, intergovernmental organizations, though they may still be able to refer to the idea of partnerships as a collaboration between private and public actors.

To ensure respondents’ comprehension of the subject of the study, we provided some brief and neutral information about MSPs. This was followed by details on the conjoint setup (see Figure 1). To ensure respondents’ understanding of the basic features and structures of MSPs, we included a comprehension test and an attention check (see Figure A1). The comprehension test asked respondents to choose between four different definitions of global partnerships, one of which was correct and had been included in the introductory text. We applied the “late screen out” rule to those who failed the tests. In total, 12,288 respondents took part in the survey experiment. Our final sample of 6,056 respondents for all three countries excludes respondents who failed our relatively elementary tests.³ In total, we obtained 48,448 observations.

In Table 1, we summarize all the conjoint attributes. The first three attributes relate to hypotheses 1–3 and specify the degree of effectiveness, transparency, and inclusivity of MSPs. For the performance attribute, we provide three values, capturing the extent to which an MSP achieves its intended goals. Similarly, for transparency, we identify three values, where MSPs are transparent, somewhat transparent, and not transparent to the public. In the inclusion attribute, we indicate two values where decision-making in MSPs involves either all (“all members have an equal say”) or only resourceful members.

The attribute leadership relates to hypothesis 4 and denotes which type of actor is in charge. Although there might be more than one leading actor, we opt to keep it simple and indicate only one actor to avoid confusion. In identifying the attribute values, we draw on the UN’s Partnership Platform database and specify six frequently leading actors, as shown in Table 1.

The social purpose attribute (hypothesis 5) identifies the general purpose of MSPs. Though MSPs may have multiple goals, we focus on single-issue goals that may affect respondents’ evaluation of an MSP. In identifying the goals, we relied on the seventeen SDGs, resulting in eight broad categories. The attribute governance function (hypothesis 6) indicates how MSPs achieve their goals and draws on existing research on the functions of MSPs (Bäckstrand and Kylsäter 2014; Pattberg et al. 2012). While the classification of governance functions of MSPs is not always clear and may combine multiple approaches, we provide five general but distinct functions, from agenda setting to rule making, that capture the broad spectrum of functions (Bäckstrand et al. 2022).

Our analysis focuses on marginal means (MMs), although we report average marginal component effects (AMCEs) in Figures A2.1 and A2.2. Leeper et al. (2020) show that AMCEs can produce misleading results, especially in subgroup analyses, as they are sensitive to the selection of the reference categories. MM analysis allows us to estimate citizens’ level of support for MSP attribute values, all else equal (i.e., average choice probabilities given specific attribute values). MMs are also relatively easy to interpret: a 0.5 (50%) MM estimate indicates that respondents are indifferent concerning particular attribute values of MSPs compared to other attribute values. In cases in which an MM coefficient is lower than the 0.5 threshold (< 50%), the respondents have less confidence in a specific MSP attribute value; when the coefficient is higher than 0.5 (> 50%), the respondents have more support for an MSP attribute compared to other attribute values.

The results of the conjoint survey experiments are presented in Figures 2a and 2b. They indicate that citizens are highly supportive of MSPs that are effective, transparent, and inclusive. These results align with hypotheses 1–3. In particular, respondents demonstrated a greater level of confidence in MSPs that are either effective (67%) or somewhat effective (55%), compared to those that are not effective (28%).

Respondents disapprove of MSPs that are not transparent (37%), whereas those that are partially transparent (52%) or fully transparent (61%) receive stronger relative support. Whether MSPs are inclusive (58%) or have a structure in which the most resourceful actors have a say (42%) also matters for respondents. In other words, respondents affirm that MSPs should be inclusive in their decision-making. Having MSPs in which members have an equal say increases support by 16 percentage points compared to MSPs in which only the resourceful members have a say. Although these results are not surprising, our analysis provides evidence that citizens view MSPs’ effectiveness as more important than their transparency and inclusion (Bernauer and Gampfer 2013; Edwards 2009; Esaiasson et al. 2012).

Our results largely support hypothesis 4a regarding leadership of MSPs. In the pooled analysis, we find that respondents prefer MSPs led by civil society actors (55%), intergovernmental organizations (53%), and universities (52%). These results are similar to those of earlier studies in which support for civil society actors in IO decision-making has been established (Dellmuth and Tallberg 2015). For other actors—for example, governments and corporations—we find that respondents are mostly indifferent. There is a slight variation across countries, the most significant being that business-led MSPs are disliked in the UK sample. However, the preference for civil society–led MSPs and the disapproval of MSPs led by faith-based organizations are robust and significant across all three countries. The lowest support for MSPs led by faith-based organizations is found in the United Kingdom (35%), followed by Brazil (39%) and the United States (42%).

With regard to the social purposes of MSPs, our evidence provides support for hypothesis 5a, that is, that social purposes matter for citizens’ evaluations of MSPs. MSPs that aim to reduce poverty (55%–58%) and promote public health (53%–55%) receive considerably more support compared to MSPs that promote other social purposes. This is in line with observational surveys showing that citizens in general view these issues as important (Ipsos 2021). MSPs that focus on gender equality (42%–43%) and trade and business (42%–48%) are in general negatively evaluated by citizens. Also, mitigation of climate change (46%–50%) is disapproved compared to other social purposes. In the Brazilian and US samples, gender equality was the least favored purpose, with a disapproval rate of 42 percent each. On the other hand, the UK sample showed the highest disapproval of trade and business as a social purpose of MSPs (42%). MSPs with the goal to protect the environment receive some support (51%) in the UK sample, but in general, citizens neither approve or disapprove them.

Indeed, promoting gender equality decreases the support by 11 percentage points compared to MSPs that work to reduce poverty, which is a substantial difference. One explanation can be that gender equality is a more politically polarizing issue compared to human rights promotion or reducing hunger. This has been observed in intergovernmental negotiations on sustainable development, where some governments try to exclude references to gender equality and women’s rights in international agreements (Kamau et al. 2018). As for other social purposes, such as human rights promotion, protection of the environment, and ensuring peace and security, citizens were in general indifferent.

Finally, our findings partially support hypothesis 5: the type of governance function adopted by MSPs does not have a significant impact on citizens’ opinions but can be an influencing factor in some cases. Figures 2a and 2b indicate that citizens generally do not prefer any specific type of governance function, with two exceptions. Citizens slightly prefer MSPs that engage in capacity-building activities (51%–52%), such as providing technical capacity, while disapproving of MSPs engaged in standard-setting activities (47%–49%).

To address hypotheses 4b and 5b, we perform subgroup analyses pertaining to political beliefs along the left–right dimension. The results pertaining to the leadership and social purpose attributes are displayed in Figure 3. Although there is some significant variation across countries, our expectations are confirmed. In the Brazilian sample, we find little support that citizens’ evaluations of MSPs are conditioned by their political values. The only substantial effect concerns the social purposes of MSPs: right-leaning respondents are more negative to the promotion of gender equality than left-leaning respondents. Nevertheless, we find substantial differences between left- and right-leaning respondents in the UK and US samples. In the UK sample, most notably, right-leaning respondents favor MSPs led by corporations, whereas left-leaning respondents disapprove of the same. Left-leaning respondents also disapprove of MSPs that promote trade and business but are positive toward MSPs that work to promote human rights. In contrast, right-leaning respondents approve of MSPs that promote trade but dislike MSPs that promote human rights. The US sample, which provides further support for our expectation, displays a similar pattern. However, in addition, we find that left-leaning respondents prefer MSPs led by governments and universities, whereas right-leaning respondents dislike both of these actors. Unique to the US sample is that right-leaning respondents show support for MSPs led by faith-based organizations, compared to other results whereby opposition to MSPs led by faith-based organizations is not dependent on citizens’ political views.

Taken together, the results in Figure 3 support our expectation that in some cases, public support for MSPs is shaped by citizens’ political values. Moreover, our results imply that the contested role of business salient in the scholarship on MSPs (Bull and McNeill 2019; Pattberg and Widerberg 2016) is not generally shared by the public but rather is dependent on citizens’ political beliefs. Political beliefs matter also for citizens’ perceptions of the social purposes of MSPs. Indeed, opposition to MSPs that focus on gender equality and climate change mitigation implies that generally politicized topics or policy areas also constitute a source of contestation among MSPs. As such, the public does not view MSPs as neutral actors or governance innovations to close implementation gaps; rather, they perceive them as political initiatives to some extent.

We conducted an additional subgroup analysis. Recent studies on the legitimacy of IOs argue that citizens’ broader political values, which fit poorly within the left–right dimension, may also form their legitimacy perceptions of global governance institutions (Ghassim et al. 2022; Hooghe et al. 2019). Specifically, Hooghe and Marks (2009) propose a framework that incorporates contrasting values: green–alternative–libertarian (GAL) and traditionalist–authoritarian–nationalist (TAN) values. Building on these studies, we investigate whether the GAL–TAN values may also influence citizens’ perceptions of MSPs. For this purpose, we include five questions that capture the GAL–TAN values (see Figure A1). Our subgroup analysis is displayed in Figures A4.1–A4.15. Overall, we find no substantial relationships between citizens’ GAL–TAN values and their evaluations of MSPs’ features. Other experimental studies obtain similar results (Ghassim et al. 2022, 11–15), providing reasons to investigate further how and when citizens’ broader values and identities relate to opinions on world politics. Finally, we provide various robustness checks in Figure A5, demonstrating that our results are robust.

As the first experimental assessment of the public legitimacy of MSPs, our findings have important implications for our understanding of MSPs. The main findings are threefold. First, we find that MSPs receive the strongest support from citizens if they can solve problems, are transparent to the public, and are inclusive in their decision-making. These results are largely in line with research on the sociological legitimacy of IOs (Tallberg and Zürn 2019). Second, the results from the conjoint experiments suggest that other features of MSPs also matter for the degree of public support. Specifically, we find evidence that the leadership feature of MSPs (e.g., if MSPs are led by governments or civil society), their social purpose (e.g., to protect the environment or promote trade and business), and to some extent also their governance functions (e.g., advocacy or capacity-building approach) matter for public legitimacy. Citizens in general prefer MSPs led by civil society actors and, to some degree, IOs but strongly disapprove of MSPs led by faith-based organizations. The public’s support for civil society actors in global institutions is well noted in existing research and attributed to their perceived closeness to citizen interests (Bernauer and Gampfer 2013). Though the disapproval of faith-based organizations in previous studies is less explored, our results indicate that citizens do not view religious actors as legitimate leaders of MSPs. With regard to social purpose, citizens show clear disapproval of MSPs that promote gender equality and climate change mitigation, while demonstrating support for those that aim to reduce poverty and promote public health. In addition, we find that the governance function exerts some, but not a significant, influence on citizens’ perceptions of MSPs. MSPs that set standards are slightly disapproved by citizens, whereas those focusing on capacity building receive some support. Third, we find that public support for MSPs in some cases depends on citizens’ political orientation. Our analysis shows that MSPs that are led by corporations and that promote trade and business are negatively viewed by left-leaning citizens while affirmed by right-leaning citizens. This highlights the significant role of political values in citizens’ perceptions of MSPs.

These results provide several important implications for the public legitimacy of MSPs. The insights provided by the survey experiments might explain why some MSPs are more successful in implementation and resource mobilization. These insights can also inform policy makers and MSPs in their policy formulation given that citizens, for instance, have a strong preference for which actors are leading partnerships and what their social purposes are. Our findings also imply that citizens value the same institutional qualities in MSPs and traditional IOs, which underscores the importance of the effectiveness of MSPs highlighted in UN policy circles. Ultimately, public support of MSPs depends on their ability to problem solve effectively and on their degree of transparency and inclusion of stakeholders. Considering previous studies of the limited effectiveness and transparency of global MSPs for environmental and sustainable development, the prospects for a high degree of public support for this mode of global governance are dim. A key observation is that MSPs with the purpose of protecting the environment or mitigating climate change do not receive as strong public support compared to MSPs with a focus on other issue areas related to sustainable development. This raises additional questions regarding the conditions under which environmental MSPs can achieve their goals (Dentoni et al. 2018). At the same time, a more nuanced picture of what we can expect from MSPs and the pathways through which they can achieve their goals calls for more systematic studies (Andonova et al. 2022).

The findings of this study also imply that the public does not view MSPs simply as neutral problem-solving institutional innovations (Börzel and Risse 2005; Gray and Purdy 2018). In contrast, they are conceived as political or politicized organizations with normative agendas, based on their leadership and social purpose. This calls for a revision of some basic assumptions in the literature on MSPs, particularly pertaining to features such as MSPs’ inclusion and governance functions. Moreover, our findings show that an MSP’s governance function may also matter for its public support. This implies that citizens value collaboration and the implementation of projects based on more bottom-up approaches, such as capacity building, rather than on top-down approaches, such as standard setting.

We also acknowledge the limitations of our study, which point to possible pathways for future research. First, our findings may be limited by the fact that we assessed only hypothetical MSPs; citizens’ perceptions may differ when citizens are exposed to real-world MSPs. Although we conducted our experiments in three countries in different regions of the world, our findings can be country specific and related to different domestic events—further research in other regions, such as Africa and Asia, is needed to provide a more complete picture. Including more open-ended questions may also provide greater insight into why citizens prefer some features over others. For example, researchers could focus on understanding when and why citizens would favor effective MSPs, while taking into account specific issue areas. Second, further research is required to understand citizens’ perceptions of MSPs in global environmental governance in comparison to intergovernmental organizations or civil society. For instance, researchers could study public support for MSPs compared to NGOs. Third, an understudied yet important research area is assessing how power imbalances among stakeholders and the underrepresentation of marginalized groups—as emphasized in the literature (Bäckstrand et al. 2022; Chowdhury 2023; Gray et al. 2022; Saffer et al. 2018)—influence the public legitimacy of MSPs. Fourth, although our research covered many important features of MSPs, a few significant features remain to be studied, such as adaptability, impartiality, and specificity. Finally, future research should systematically examine legitimacy perceptions among stakeholders, who, compared to the general public, are often more informed about MSPs. Exploring these avenues with various methodological approaches could enhance our understanding of MSPs’ public legitimacy and their effectiveness in global environmental governance.

This research was funded by FORMAS, the Swedish Research Council for Sustainable Development, under the project “Transformative Partnerships for Sustainable Development: Assessing Synergies, Effectiveness, and Legitimacy of the UN’s Multi-stakeholder Partnerships Across SDGs to Achieve the 2030 Agenda” (grant 2020-00418).