Elite Communication Matters for Legitimacy Beliefs

Our general expectation for why citizens’ legitimacy beliefs toward IOs would be receptive to elite communication was grounded in three assumptions: Citizens’ political awareness tends to be low, citizens therefore rely on heuristics to form political opinions, and reliance on heuristics makes citizens susceptible to elite influence. These assumptions are anchored in cognitive psychology and inform research on opinion formation in domestic politics. We saw no reason why they would not be equally applicable in the context of global governance. In fact, we suspected they might be even more relevant in this setting, as citizens tend to be even less aware of global politics than domestic politics.

The findings lend extensive support to our general expectation: Citizens rely on elite communication as a cognitive shortcut to opinions about IOs, and elites are thereby given an opportunity to shape how citizens evaluate IOs. Our analyses show this effect for a variety of elites and for messages invoking a broad range of institutional qualities of IOs. These findings tie in well with earlier observational research on public opinion toward the EU (Hooghe and Marks Reference Hooghe and Marks2005; Gabel and Scheve Reference Gabel and Scheve2007; De Vries and Edwards Reference Edwards2009), but establish these effects with greater causal certainty and for a broader sample of IOs.

First, a multitude of elites affects how citizens conceive of IOs, from the domestic party elites that people usually turn to for guidance to the global elites that always compete for influence in world politics. We arrived at this result by first examining member governments, NGOs, and IOs as globally active elites, exploring their influence comparatively across five prominent global or regional IOs (EU, IMF, NAFTA, UN, and WTO) based on a vignette experiment with citizens in three countries (Germany, the UK, and the US) (Chapter 4). All three types of elites are frequent communicators about the merits and demerits of IOs, and therefore the key focus of research on legitimation and delegitimation in global governance. Our analyses establish that these communicative acts by member governments, NGOs, and IOs, indeed, have effects on citizens’ legitimacy beliefs, albeit not to the same extent for all three actor categories, as we discuss further below.

We then turned to domestically oriented elites, examining the effects of cues from political parties based on two vignette experiments involving party communication regarding two IOs (NATO and UN) in two countries (Germany and the US) (Chapter 5). Political parties and candidates frequently communicate about IOs when debating political issues with an international dimension, but have traditionally received limited attention in global governance research. Our findings show that communication by political parties can effectively shape how citizens perceive the legitimacy of IOs, even if these effects are subject to variation.

Second, elites may choose from a large menu of salient institutional features in order to influence the legitimacy perceptions of citizens toward IO. We first established this finding for the procedures and performances of IOs, whose influence on legitimacy beliefs is debated in a large literature. We explored this issue comparatively across seven IOs in different issue areas (ASEAN, AU, EU, IMF, NAFTA, UNSC, and UNFCCC) based on a vignette experiment with citizens in four countries in diverse world regions (Germany, the Philippines, South Africa, and the US) (Chapter 6). Our analyses confirm that information about both types of qualities impact people’s legitimacy beliefs, contrary to a common expectation that public opinion toward IOs would be driven primarily by procedural (input) or performance (output) considerations.

We then extended this analysis by also considering the potential importance of an IO’s authority and purpose for its perceived legitimacy. Recent studies suggest that these two qualities may shape people’s opinions about IOs, but do not properly test these expectations. We examined this issue across hypothetical IOs and based on a conjoint experiment with citizens in two countries (Germany and the US) (Chapter 7). Our results show that both an IO’s authority and its social purpose matter for citizens as they form opinions about its legitimacy, even when considering its procedures and performance. Yet, in terms of the strength of the effects, procedure and performance remain the more important qualities.

Effects Are Conditioned by the Communicative Setting

Next to its general expectation, our theory produced a set of conditional expectations about the circumstances under which elite communication should have particularly strong or weak effects on legitimacy beliefs. We argued that elite influence is likely to vary with conditions associated with three features of the communicative setting: the elite, the message, and the citizen. Specifically, we identified six moderating factors: elite credibility, elite polarization, tone of the message, object of the message, citizens’ political awareness, and citizens’ political beliefs. Our empirical analyses broadly support these expectations but also point to interesting exceptions. In the following, we discuss the findings on each factor individually, integrating evidence from across the four empirical chapters. Overall, our findings resonate well with observational research highlighting some of the same factors we emphasize (Hooghe and Marks Reference Hooghe and Marks2005; Gabel and Scheve Reference Gabel and Scheve2007; Bechtel et al. Reference Bechtel, Jens, Hangartner and Marc2015), which further strengthens our confidence in these results.

We expected the credibility of elites to condition the impact of their communication on citizens’ legitimacy beliefs toward IOs. When elites are perceived to be more credible, they should be more likely to sway the opinions of citizens. In this respect, citizens are not mindless followers of elites, but selective in listening to those elites who enjoy the most credibility in their eyes. In the domestic context, in which issues usually are more ideologically polarized, we expected partisanship to be the central source of credibility for elites. In the global context, in which issues often are less polarized and elites less strongly linked to specific partisan positions, we expected credibility instead to be tied to perceptions of impartiality.

Our analyses offered evidence on elite credibility in the domestic context, when exploring the influence of party cues on partisans and independents in the US and Germany (Chapter 5). Specifically, we examined whether citizens are more receptive to party elites they sympathize with than to party elites they disagree with or do not identify with. Our findings offer support for this expectation, but not unreservedly. Party cue effects are primarily identified in the US setting and mainly concentrated on the Democrats. Yet, if we take citizens’ pretreatment opinions into consideration, then party cue effects are significantly stronger across the board for those citizens with more positive attitudes toward the IOs and issues in question. We will shortly return to the question of why party cue effects were stronger in the US than in Germany, when we consider elite polarization as a conditioning factor.

In addition, we evaluated the importance of elite credibility in the global context, when examining whether elites perceived as more impartial and unbiased are better positioned to shape citizens’ legitimacy beliefs toward IOs (Chapter 4). Specifically, we expected NGOs to be perceived as most credible, member governments as slightly less credible, and IOs themselves as least credible, resulting in corresponding patterns in influence. The findings are largely consistent with our expectations. Communication effects tend to be stronger for more credible elites – NGOs and member governments – than for less credible elites – IOs themselves. IOs are not found to be effective when communicating positive messages about themselves; however, when admitting to mistakes, their communication has a strong negative effect on citizens’ legitimacy beliefs, likely reflecting the greater credibility of this admission. Contrary to the pattern for domestic party elites, these globally oriented elites are more influential in Germany and the UK than in the US, indicating that country context shapes citizens’ relative susceptibility to communication by varying types of elites.

The second elite factor that we expected would have a conditioning effect was the polarization of elites. This expectation is rooted in research on party cueing, where polarization is seen to have two parts: the ideological distance between parties and the ideological homogeneity within each party. The expectation is that greater polarization in both respects leads to stronger party cue effects, since it makes it easier for partisans to identify and follow the party line.

We found significant support for this expectation when examining the influence of party elites in the US and Germany (Chapter 5). Overall, party elites are considerably more effective communicators in the more polarized US context compared to the less polarized German setting, confirming the importance of ideological distance between the parties. The polarization between Democrats and Republicans in the US two-party system greatly facilitates elite influence compared to the German multiparty system, where the two main contending parties – CDU/CSU and SPD – even have ruled together in a grand coalition since 2013. There is more mixed evidence that ideological homogeneity within parties matters for their effects on citizens’ legitimacy perceptions. When manipulating the information respondents received about the homogeneity within parties on the issues in question, we could establish a conditioning effect among Republican and CDU/CSU partisans, but not among Democratic and SPD partisans. Possibly, this mixed record reflects a stronger pretreatment expectation of intraparty homogeneity for the first two parties, especially the US Republicans, compared to the latter two parties, especially the US Democrats.

Turning to the communication as such, we expected the tone of the message to condition its impact on citizens’ legitimacy beliefs. Building on research in comparative politics, economics, and psychology, we theorized that negative elite messages would have stronger effects on legitimacy perceptions than positive messages. This expectation recognizes that communication about IOs seldom is neutral, but charged with an evaluative tone, making it an act of legitimation or delegitimation. It is by praising or criticizing IOs that elites seek to influence the public perception of these organizations.

We found strong support for this expectation when evaluating tone as a conditioning factor (Chapter 4). The tone of a message matters for the likelihood that it catches people’s attention and influences their opinions of IOs. Negative messages are more effective than positive. This is true regardless of whether elites invoke the procedures or the performances of IOs in their communication. We also found this result to hold at the level of individual countries and IOs. Negative messages have a stronger effect on legitimacy beliefs in all three countries (Germany, UK, and US). In addition, negative messages have a stronger impact in the case of all IOs for which significant effects were identified (EU, IMF, UN, and WTO). When we disaggregated the analysis to specific IOs in specific countries, the results were more variegated, but broadly consistent with the overall pattern.

A second feature of the message that we expected would have a conditioning effect was the object of communication. Does the communication refer to IOs that have been subject to high levels of prior contestation and on which citizens already have hardened opinions, or to IOs that have been less publicly debated and on which people’s opinions are less crystallized? Drawing on research on attitude strength and motivated reasoning, we expected the level of prior contestation of an IO to condition the effectiveness of elite communication.

This expectation offered a valuable way of interpreting variation across IOs observed in several analyses. Patters in the effectiveness of elite communication closely fit the notion that citizens respond less strongly to messages about IOs that have been subject to more intense public debate in the past, leading people to develop stronger priors. We found that elite communication more often was effective in the context of the IMF, UN, and WTO, than in the context of the EU and NAFTA (Chapter 4). We observed this pattern regardless of whether we focused on the communicating elites, the features of IOs invoked in these messages, or the tone of the communication. While, for sure, the IMF, UN, and WTO do not fly beneath the radar of public attention, they have been far less intensely debated in recent years than the EU and NAFTA. European integration has been highly politicized in the member states for a long period of time – a pattern lately reinforced by populist mobilization against the EU and Britain’s choice to leave the organization. NAFTA, too, has been subject to extensive domestic contestation, especially in the US, which was a contributing factor behind its recent renegotiation and conversion into the USMCA. This interpretation also helped us to understand the variation in communication effects we observed in a comparison between the IMF, UNFCCC, and UNSC (Chapter 6). The analysis showed that citizens’ legitimacy beliefs were more easily swayed in the case of the IMF than in the case of the UNSC and especially the UNFCCC. This pattern well reflects people’s level of familiarity with these IOs and the salience of climate change in public debate in recent years.

As a third category, we theorized the moderating impact of factors associated with citizens themselves. To begin with, we expected citizens’ political awareness to shape their receptivity to elite communication. Specifically, the more politically aware citizens are, the more responsive they should be to elite cues about IOs. This expectation builds on an influential literature in political communication, theorizing that politically more sophisticated citizens are more likely to make use of heuristics as a shortcut to opinions. Integrating new information is aided by a preexisting level of political awareness that allows individuals to interpret and systematize this information.

We found only weak support for this expectation. Our analyses of communication by global elites (Chapter 4) and communication on procedure and performance (Chapter 6) offered no support for a conditioning effect of citizens’ political awareness, using two alternative indicators – discussion of politics and education. Similarly, our analysis of communication about authority and purpose (Chapter 7) found that neither political knowledge nor education affected citizens’ receptivity to information about these qualities of an IO. However, our analysis of communication by domestic elites (Chapter 5) established that knowledge about global governance strengthens the effects of party cues in the expected direction among Democratic partisans in the US, but not among Republican partisans in the US or among CDU/CSU or SPD partisans in Germany. Taken together, these results indicate that political awareness is not a decisive conditioning factor for citizens’ receptivity to elite communication. This may reflect a simultaneous occurrence of two forces pulling in different directions: on the one hand, less politically aware citizens demanding guidance from elites to a greater extent; on the other hand, more politically aware citizens being able to integrate information from elites to a greater extent.

Finally, we expected citizens’ political beliefs to matter in multiple ways for the impact of elite communication. This expectation goes hand in hand with the idea that political parties are important communicators on issues of global governance, shaping the opinions of partisans, as discussed earlier. Citizens should then be most receptive to elite communication when they are ideologically proximate to the elites issuing these messages. In addition, this expectation builds on the idea that citizens may interpret the same information about IOs in different ways, and with different consequences for legitimacy beliefs, depending on their preexisting political beliefs. In all, this expectation underlines the inherently political nature of elite communication and its effects on citizens’ legitimacy beliefs toward IOs.

We found moderate support for this expectation. Our analysis of communication by domestic party elites (Chapter 5) showed that citizens’ partisan identity affects the impact of party cues on legitimacy beliefs. When supporters of the main political parties in the US and Germany were confronted with basic information on the parties’ positions on issues of global governance, their legitimacy beliefs toward IOs shifted in the expected direction. Yet statistically significant effects were mainly concentrated on Democratic partisans in the US. We also examined whether the strength of a citizen’s partisan identification mattered for the impact of party cues and found mixed evidence. For Democratic partisans, the strength of the identification had a positive impact on their receptivity to party cues in the context of NATO. However, similar effects could not be established for the UN, for Republicans, or for partisans in Germany. This mixed record should be understood against the backdrop of varying degrees of political polarization in the US and Germany, not only among elites, as discussed earlier, but also among citizens. Supporters of the Democrats and Republicans were much more likely to hold different pretreatment opinions on global governance issues than supporters of the CDU/CSU and SPD. Moreover, US respondents were considerably more likely to recall the cued information on party positions than German respondents, pointing to the larger role of partisan identification in the US context.

Our analysis of communication on the social purposes of IOs (Chapter 7) presented clear evidence that citizens’ political beliefs matter in how they evaluate such information. In both the US and Germany, information on the social purposes of IOs yielded effects on citizens’ legitimacy beliefs, consistent with the idea that such purposes are politically interpreted by citizens. Compared to a baseline of ensuring peace and security, the social purposes of climate change and free trade had clear negative effects on IO legitimacy in the US. In Germany, similar negative effects were recorded for free trade and public health. To further evaluate this link, we also examined if these effects were conditioned by citizens’ level of political ideology and found evidence for this in the US but not in Germany. For Democrats, a purpose of poverty alleviation contributes positively to legitimacy beliefs, while a purpose of free trade promotion has a negative effect, compared to a baseline purpose of ensuring peace and security. For Republicans, both poverty and climate change mandates lead to less confidence in IOs. In Germany, the results were not contingent on partisan identification, likely reflecting a lower level of partisan polarization over international issues.

Taken together, our findings show that elite communication shapes citizens’ legitimacy beliefs toward IOs. Citizens are sensitive to communication by global as well as domestic elites, and to messages that invoke a variety of IO features, from the purpose and authority of these organizations to their procedures and performance. These effects tend to be particularly strong under certain conditions – when elites enjoy greater credibility and are more polarized, when messages are negative in tone and focus on IOs less contested in the past, and when citizens hold firmer political beliefs. Conversely, the impact of elite communication tends to be weaker, or even nonexistent, when elites are less credible and polarized, messages are positive and focused on highly contested IOs, and citizens hold softer political beliefs. Citizens’ political awareness does not appear to matter systematically for the effectiveness of elite communication, contrary to well-established expectations in earlier research.

Generalizability

Our confidence in these findings is boosted by the reliance on experimental methods for causal inference, which have allowed us to by-pass problems of complex causality and omitted variables that typically afflict observational studies of elite influence. In addition, we have established these results through a comparative analysis that encompasses multiple and diverse IOs and countries. These strengths notwithstanding, we should recognize the limitations of our design and how they may be addressed in future research.

First, future research may benefit from evaluating our theory in other IO and country contexts as well. While broad in scope compared to most experimental designs, our study is still limited to a handful of IOs and countries, with an emphasis on relatively well-known organizations and advanced democracies. Our findings suggest that some conditions at the level of IOs (prior contestation) and countries (elite polarization) systematically shape how elite communication influences the legitimacy beliefs of citizens, but a broader sample may yield additional insights (Guisinger and Saunders Reference Guisinger and Saunders2017).

Second, future research may usefully extend the analysis to more complex communicative settings. While we examined a variety of elites and messages, our design focused exclusively on one-off effects of elite communication. This was an essential first step for this research agenda. Next steps could involve examining elite communication in contexts characterized by competing elite messages (Chong and Druckman Reference Chong and Druckman2007b; Ghassim Reference Ghassim, Bexell, Jönsson and Uhlin2022), simultaneous cues from social peers (Kertzer and Zeitzoff Reference Kertzer and Zeitzoff2017; Isani and Schlipphak Reference Isani and Schlipphak2020), longer time horizons (Druckman et al. Reference Druckman, Hennessy, Charles and Webber2010), and reciprocal effects between elites and publics (Steenbergen et al. Reference Steenbergen, Edwards and de Vries2007).

Third, we arrived at our results using survey-experimental methods. While these methods have advantages in terms of identifying causality, they raise the question of whether we should expect to find the same effects outside the survey experimental setting. In this book, we have worked with a three-pronged strategy to address this concern: designing experiments in ways that mimick real-world conditions, relying on population-based samples, and discussing our findings in view of available observational evidence. In addition, we note that systematic comparisons between experimental results and real-world outcomes reveal a high level of external validity (Hainmueller et al. Reference Hainmueller, Hangartner and Yamamoto2015). Going forward, we see extensive potential in combining observational and experimental methods in ways that are complementary and help to ensure both internal and external validity. On the observational side, big data on social media offer new opportunities to study communication effects. On the experimental side, survey experiments may be complemented by field experiments, which can be designed in ways that are very similar to real-world elite communication (cf. Broockman and Butler Reference Broockman and Butler2017; Nielson et al. Reference Nielson, Hyde and Kelley2019).

Legitimacy and Legitimation in Global Governance

Most obviously, our findings have implications for the understanding of legitimacy and legitimation in global governance. Recent years have seen an upsurge of interest in legitimacy, understood as audience beliefs about an IO’s exercise of authority, and in legitimation, understood as the process by which actors seek to influence such beliefs (for overviews, see Tallberg et al. Reference Tallberg, Bäckstrand and Scholte2018; Tallberg and Zürn Reference Tallberg and Zürn2019). Our results support a perspective on legitimacy beliefs as formed in a social process of legitimation and delegitimation, where evaluative claims in the public realm shape individual attitudes. As such, our findings have consequences both for research on the sources of legitimacy for IOs and for scholarship on legitimation and delegitimation processes in global governance.

The literature on sources of legitimacy for IOs seeks to identify the principal drivers of such beliefs (e.g., Hurd Reference Hurd2007; Norris Reference Norris, Esmer and Pettersson2009; Johnson Reference Johnson2011; Voeten Reference Voeten2013; Dellmuth and Tallberg Reference Dellmuth and Tallberg2015, Reference Dellmuth and Tallberg2020; Schlipphak Reference Schlipphak2015; Scholte and Tallberg Reference Scholte, Tallberg, Tallberg, Bäckstrand and Scholte2018; Anderson et al. Reference Anderson, Bernauer and Kachi2019; Nielson et al. Reference Nielson, Hyde and Kelley2019; Tallberg and Zürn Reference Tallberg and Zürn2019; Bernauer et al. Reference Bernauer, Mohrenberg and Koubi2020; Dellmuth et al. Reference Dellmuth, Scholte, Tallberg and Verhaegen2022b). Earlier research has suggested that legitimacy beliefs are shaped by individual-level factors (such as cosmopolitan identity), organization-level factors (such as institutional procedures), and societal-level factors (such as a country’s prior experiences with an IO). Our findings show that elite communication constitutes a fourth, independent type of explanation for citizens’ legitimacy beliefs toward IOs. Indeed, our design allowed us to establish that elite communication matters irrespective of other potential drivers of legitimacy beliefs. Legitimacy perceptions are not set, but continuously evolving, as citizens integrate arguments and information from trusted elites.

It may even be that elite communication as a process shapes other factors contributing to citizens’ legitimacy beliefs. Also accounts that emphasize institutional or societal factors are essentially information based (Mansfield and Mutz Reference Mansfield and Mutz2009, 432; Curtis et al. Reference Curtis, Jupille and Leblang2014, 721). Since citizens have few personal experiences of IOs, elite communication in news media and social media becomes the primary way by which they acquire information about an IO’s institutional qualities and a country’s experiences with an IO. In practice, then, institutional and societal factors may not affect legitimacy beliefs independently of elite communication, but as mediated through elite communication.

Our findings are of particular importance for the debate on institutional sources of legitimacy beliefs (Scharpf Reference Scharpf1999; Bernauer and Gampfer Reference Bernauer and Gampfer2013; Dellmuth and Tallberg Reference Dellmuth and Tallberg2015; Scholte and Tallberg Reference Scholte, Tallberg, Tallberg, Bäckstrand and Scholte2018; Anderson et al. Reference Anderson, Bernauer and Kachi2019; Dellmuth et al. Reference Dellmuth, Scholte and Tallberg2019; Bernauer et al. Reference Bernauer, Mohrenberg and Koubi2020). Our results suggest that citizens care equally about IOs’ procedures and performance when developing legitimacy perceptions. In addition, they demonstrate, for the first time, that citizens care about the authority and purpose of IOs. Theoretical accounts that privilege the one or the other institutional quality, therefore, appear misguided.

Establishing elite communication as a driver of legitimacy beliefs also has implications for the large literature on legitimation and delegitimation in global governance (e.g., Zaum Reference Zaum2013; Binder and Heupel Reference Binder and Heupel2015; Gronau and Schmidtke Reference Gronau and Schmidtke2016; Bäckstrand and Söderbaum Reference Bäckstrand, Söderbaum, Tallberg, Bäckstrand and Scholte2018; Zürn Reference Zürn2018; Dingwerth et al. Reference Dingwerth, Witt, Lehmann, Reichel and Weise2019; von Billerbeck Reference Von Billerbeck2020; Bexell et al. Reference Bexell, Jönsson and Uhlin2022). While earlier research has made important advances in mapping and explaining the communicative practices that defenders and opponents of IOs use, studies of the effects of legitimation and delegitimation on citizens’ legitimacy beliefs are scant. This book offers such an analysis. Our findings confirm that discursive legitimation and delegitimation matter and identify the conditions under which such practices are particularly impactful. They suggest that the success of legitimation and delegitimation in changing opinions toward IOs varies systematically with features of the communicative situation – the elites, the messages, and the citizens. While a large variety of elites aim to influence public perceptions of IOs using a broad range of messages, some communication gets through to citizens more effectively than other communication.

Specifically, our findings suggest that legitimation and delegitimation by member governments and NGOs are more effective in shaping popular legitimacy beliefs than self-legitimation by IOs. A large part of the literature thus appears to focus on practices – IO self-legitimation – that ultimately carry less significance for the perceived legitimacy of these organizations. While these studies are still revealing in terms of how IOs choose to present themselves to the outside world, future research could devote more sustained attention to legitimation and delegitimation by more impactful actors. Notably, our research underlines the importance of member governments as key actors shaping the perceived legitimacy of IOs. Citizens tend to listen to their national government in debates over international cooperation, especially when they identify with the party in office. Similarly, NGOs wield influence over how citizens regard the legitimacy of IOs, irrespective of whether they criticize or praise these organizations. This result ties in well with earlier research on the framing power of NGOs in global governance (Keck and Sikkink Reference Keck and Sikkink1998; Joachim Reference Joachim2003). Finally, our findings underline the importance of extending the study of legitimation and delegitimation of IOs to domestic partisan elites. Political parties communicate about IOs in ways that matter for popular legitimacy, especially in more polarized environments, but have so far been absent from the study of legitimation and delegitimation in international relations.

Elite Communication and Public Opinion

Our findings also have consequences for the understanding of elite communication in politics generally (e.g., Zaller Reference Zaller1992; Chong and Druckman Reference Chong and Druckman2007a; Gabel and Scheve Reference Gabel and Scheve2007; Levendusky Reference Levendusky2010; Bullock Reference Bullock2011; Druckman et al. Reference Druckman, Peterson and Slothuus2013; Guisinger and Saunders Reference Guisinger and Saunders2017; Torcal et al. Reference Torcal, Martini and Orriols2018). While research in American and comparative politics has spearheaded the analysis of elite influence, it remains focused on political parties and candidates as communicators in domestic political settings. Yet elites are not only national in orientation, and even domestic elites communicate about international politics as well. This book joins the few contributions that examine how elite communication operates under conditions of global governance. With the growing internationalization of politics, it becomes increasingly important for public opinion research to explore how political communication and opinion formation work in the global realm. By expanding the scope, our analysis yields novel insights for the general study of elite communication.

To start with, it shows that citizens form opinions on global politics in ways very similar to how they form opinions on domestic politics or other matters in their everyday lives. When confronting informational constraints, people rely on heuristics to make decisions. Just as previous analyses have confirmed the importance of elite communication in the domestic context, we have demonstrated how elite communication shapes people’s attitudes toward global politics. In addition, our analyses show that factors that are known to moderate the impact of elite communication in the domestic context also explain variation in these effects in the global setting – the role of partisanship as a source of elite credibility, the degree of polarization among elites, the tone of messages, and the strength of citizens’ partisan identification. The exception is citizens’ political awareness, which is a well-known moderating factor in the domestic context (Sniderman et al. Reference Sniderman, Brody and Tetlock1991; Zaller Reference Zaller1992), but which did not systematically shape the impact of elite communication in our analyses.

Yet our findings also show that the specific political context of global governance matters for how elite communication influences public opinion. Two broad contextual differences merit particular attention: The elites and the issues debated.

The elites that engage in communication about global politics are broader than the partisan elites who dominate domestic politics. Understanding elite influence in the context of global governance, therefore, requires an appreciation of the breadth and diversity of these elites, as well as the specific conditions under which they operate. As we have shown, elite credibility in the global context is not always down to partisanship. While partisanship is a central source of credibility for domestic political elites, other sources of credibility are at play for the range of elites active in the global setting. For member governments, NGOs, and IOs themselves, being perceived as an impartial source of accurate information is particularly important when seeking to influence people’s legitimacy beliefs. Tellingly, IOs appear incapable of shifting people’s opinions when communicating positively about their own operations and activities, reflecting their low credibility as sources of unbiased information about themselves. In contrast, IOs are a credible and influential source of information when communicating about other actors, such as states (Chapman Reference Chapman2009). Our findings tie in well with other recent research, which shows that epistemic communities may influence public support for international cooperation (Maliniak et al. Reference Maliniak, Parajon and Powers2021), underscoring how elite credibility in the global domain often is linked to other sources than partisanship. In all, examining elite communication across a broader range of elites than is conventional in this field allows us to identify novel conditions and dynamics shaping their influence.

In addition, the issues that elites communicate about in global politics have features that matter for the impact on public opinion. At a general level, the public tends to be less informed about international issues (Delli Karpini and Keeter Reference Delli, Carpini and Keeter1996), tends to find international issues less politically salient (Aldrich et al. Reference Aldrich, Gelpi, Feaver, Reifler and Thompson Sharp2006), and tends to have more ambivalent attitudes on international issues (De Vries Reference De Vries2018), compared to domestic political issues. For all these reasons, public opinion on international issues should be particularly malleable by elites. Our findings support this expectation, showing how multiple types of elites can influence people’s legitimacy beliefs toward a broad range of IOs in a diverse set of countries. Global politics appears to present an unusually hospitable environment for elite influence. It occupies that soft spot where the public is aware of an issue and finds it reasonably important, but does not have sufficiently hardened attitudes to resist elite communication.

At the same, the issues that elites communicate about in global politics vary in ways that matter for the impact on public opinion (Guisinger and Saunders Reference Guisinger and Saunders2017). Our findings show that the level of prior contestation of an issue shapes the ease with which elites can sway citizen opinions. When issues have been more extensively contested in the past, citizens are more likely to have developed hardened attitudes on these issues, making elite influence less likely. These results underline the importance of studying elite communication comparatively across issues. For instance, had we only focused on the EU, as in most existing studies, we would have underestimated the effects of elite communication about IOs. As Guisinger and Saunders (Reference Guisinger and Saunders2017, 425) put it: “[T]he single-issue nature of most survey experiments masks systematic variation in how elite cues affect attitudes across international issues.” In all, global governance presents a context that is particularly conducive to exploring how the nature of issues affects the scope for elite influence.

Elite Influence in a Democratic Perspective

Our findings also raise an important normative issue: Is it good or bad from a democratic perspective that citizens’ evaluations of IOs are susceptible to elite communication? Should we celebrate or lament the fact that elites appear quite capable of shaping citizens’ perceptions of IO legitimacy?

As political authority increasingly has moved from the national to the international level, the conditions for democratic engagement by citizens have changed as well. Whereas, before, citizens needed to understand the basics of national politics in order to exercise their role as democratic subjects, they must nowadays also comprehend and develop opinions toward global politics. With the shift of power to political institutions beyond the nation-state comes new expectations on citizens. What is the role of elite communication in helping people navigate the new landscape of global politics?

It is well known that citizens experience competence constraints already in the domestic context (e.g., Converse Reference Converse and Apter1964; Sniderman et al. Reference Sniderman, Brody and Tetlock1991; Delli Carpini and Keeter Reference Delli, Carpini and Keeter1996; Kuklinski Reference Kuklinski2001; Achen and Bartels Reference Achen and Bartels2016; Lupia Reference Lupia2016). People tend to be relatively unaware and unknowledgeable about politics. And while there is variation in this respect, such differences only underscore the inequalities involved in citizens’ potential for democratic engagement. These challenges are likely compounded in the international context, since people tend to be even less aware and knowledgeable about global issues (Delli Carpini and Keeter Reference Delli, Carpini and Keeter1996). Having the requisite competence and resources as a citizen to develop opinions and make political choices in global politics is demanding (Dingwerth Reference Dingwerth2014).

Elite communication has the potential of both frustrating and helping citizens as democratic subjects in global politics. According to the pessimistic interpretation, elite communication exploits citizens’ informational deficits and manipulates their opinions. If citizens possess such limited information about politics that their opinions on policies, politicians, and institutions can be easily manipulated by elites, this is bad news for democratic politics. Then citizens have no genuine interests and beliefs, and public preferences no stability. In this view, reliance on elite communication may lead citizens to disregard the policy information they possess, to shy away from independently assessing the merits of different positions, and to arrive at positions that do not fully reflect their preferences. As Bullock (Reference Bullock2011, 496) puts it: “[T]he greatest concern about elite influence on public opinion has been that it causes people to hold positions that they would not hold if they knew more facts.” In a larger perspective, substituting one’s own judgment with a mechanical adoption of elite positions, even those of political parties, may be to the detriment of democracy (Achen and Bartels Reference Achen and Bartels2016). This concern is rooted in theories of democracy that stress the importance of bottom-up processes of preference formation and representation (Pateman Reference Pateman1970; Pitkin Reference Pitkin1972).

The optimistic interpretation instead emphasizes communication from elites as part of a democratic learning process. Elite communication and contestation are natural and necessary components of the process through which individuals develop political attitudes. Most information about politics that citizens acquire originates with elites, and forming an opinion involves assessing, accepting, or rejecting competing messages communicated in the public realm. When elites systematically influence the choices that citizens make, this contributes to greater democratic effectiveness by aiding citizens in the “correct” political choices in view of their preferences (Kuklinski and Quirk Reference Kuklinski and Quirk2001). As Sniderman and Bullock (Reference Sniderman, Bullock, Saris and Sniderman2004, 346) explain in the context of party influence: “Citizens in representative democracies can coordinate their choices insofar as the choices themselves are coordinated by political parties.” This optimistic view is rooted in theories privileging elite competition as a normal component of democratic rule (Schumpeter Reference Schumpeter1947).

These drawbacks and benefits of elite communication are particularly salient in the context of global governance. Since citizens typically are less informed about global politics, they are not only more vulnerable to elite influence but also more likely to be aided by it. Moreover, since the public sphere is less developed in global politics, it is more easily captured by determined elites but also in greater need of enrichment from political and societal elites.

Adjudicating whether elite influence on balance is positive or negative from a democratic perspective is difficult on both empirical and normative grounds. Yet only recognizing this issue presents us with a set of important questions for future research to explore. How is citizen competence on global issues distributed across groups in society? Do structural conditions allow for a multitude of elite voices to be heard on global issues? Are citizens exposed to competing elite messages or only messages from likeminded elites? Are there systematic biases in terms of how elite communication affects citizens? Some of these questions have recently become topical as we seek to understand the recent backlash against IOs.

The Backlash against International Cooperation

Finally, our findings have implications for how to understand the backlash against IOs. The past decade has witnessed growing resistance to IOs, expressed in a variety of ways. States have chosen to withdraw from IOs such as the EU, ICC, UNESCO, and WHO (von Borzyskowski and Vabulas Reference von Borzyskowski and Vabulas2019; Voeten Reference Voeten2020). Voters in several countries have used popular referenda to challenge the terms of international agreements (Walter Reference Walter2020). Populists with antiglobalization messages have been elected to government in, for instance, Brazil, India, the Philippines, and the US, and gained strong electoral support in a broad range of countries (Copelovitch and Pevehouse Reference Copelovitch and Pevehouse2019; Hooghe et al. Reference Hooghe, Lenz and Marks2019).

A common explanation for this backlash refers to globalization’s effects on societies, which are seen to have caused public grievances with international cooperation. As globalization has deepened, spurred by the actions and policies of IOs, it has affected societies in ways that have led people to turn against international cooperation. Some studies emphasize the economic consequences of globalization, pointing to rising inequality and an ensuing division of society into winners and losers (e.g., Rodrik Reference Rodrik2018). Other studies stress noneconomic consequences of globalization, such as cultural concerns, value shifts, and a rise in antiimmigration attitudes, as sources of growing discontent (e.g., Norris and Inglehart Reference Norris and Inglehart2019). The emerging consensus is that both economic and noneconomic concerns contribute to the backlash against globalization (Walter Reference Walter2021).

At the same time, there is little empirical support for a large-scale shift toward more negative public attitudes vis-à-vis IOs. As discussed in Chapter 2, public support for IOs appears to hold quite steady over time, even if there is variation across IOs and countries. The UN experienced a decline in average confidence until 2014, but has seen growing support since. The EU witnessed a decline in average trust during the Eurozone crisis, but has experienced a significant rise since 2013. Average support for the AU has declined slightly since 2009, but overall remained at a high level. Public support for Mercosur has been on a solid rise since 2002. Moreover, IOs still tend to enjoy greater support than national governments in most countries. These patterns echo findings in other studies which show that public opinion toward IOs is moderately supportive in general (Dellmuth et al. Reference Dellmuth, Scholte, Tallberg and Verhaegen2022b) and hardly on a clear downward trajectory over the past decade (Nguyen and Spilker Reference Nguyen, Spilker, Elsig, Hahn and Spilker2019; Tallberg Reference Tallberg, Barnett, Pevehouse and Raustiala2021; Walter Reference Walter2021; however, see Bearce and Jolliff Scott Reference Bearce and Jolliff Scott2019). A recent comprehensive study of trends in more than thirty IOs concludes that there is no such thing as a general legitimacy crisis in global governance (Sommerer et al. Reference Sommerer, Agné, Zelli and Bart2022). Many IOs have not experienced any legitimacy crises over the past 35 years, and among those who have, the timing varies and appears largely dictated by contextual circumstances.

How can this paradox be explained? While we have witnessed a political backlash against IOs, involving actions by voters in elections and referenda, public opinion toward IOs has held quite steady. Our findings can help to shed light on this puzzle. They support the interpretation that the recent backlash against IOs reflects the success of political entrepreneurs in exploiting specific public grievances for political gain (De Vries and Hobolt Reference De Vries and Hobolt2020; Walter Reference Walter2021). While IOs continue to enjoy reasonably broad public support across the membership, antiglobalization elites have been skilled in identifying and nurturing grievances in particular groups and countries. As De Vries et al. (Reference De Vries, Hobolt and Walter2021, 314) put it: “Political entrepreneurs try to successfully ignite opinions that lay dormant or mobilize aspects of preexisting discontent that are most advantageous to politicize. In doing so, political entrepreneurs seek to gain electoral advantage from driving a wedge between mainstream elites and their supporters by mobilizing opposition to international cooperation.”

Communication is the method of persuasion for both antiglobalization elites and the elites who rush to the defense of international cooperation. Our findings suggest why the advocates of global governance so far have come up short against its critics. To begin with, negative communication about IOs is considerably more common than positive communication (Chapter 2). While likely more numerous, the elites supportive of global governance are also more silent. In addition, negative communication about IOs is more effective than positive communication (Chapter 4). In particular, citizens are often not convinced by IOs’ own attempts to talk up their legitimacy. The elites of discontent thus appear to hold the upper hand, and there is a risk of a downward spiral in IO legitimacy as a result.

For advocates of global governance, eager to push back against antiglobalization elites, the lessons are threefold. First, they need to step up their efforts. The supporters of international cooperation have to be as energetic in their defense as the opponents are in their offensive. And there is no reason to wait until IOs are under attack to explain the benefits of international cooperation. Recent research shows that positive communication about IOs can help to neutralize the effects of negative communication (Ghassim Reference Ghassim, Bexell, Jönsson and Uhlin2022). Second, advocates of global governance need to work through the channels that are most credible in the eyes of citizens. Mobilizing civil society and national political elites to speak up on behalf of IOs is more effective than having these organizations trumpet their own credentials. Leaving the legitimacy of global governance to the public communication departments of IOs is a losing strategy. Instead, NGOs, national governments, and political parties need to shoulder their responsibility in countering antiglobalist narratives. Third, advocates of global governance should work to reform IOs in ways that yield greater legitimacy in the eyes of citizens. When IOs are perceived as more democratic, effective, and fair, they are also rewarded with greater legitimacy. While institutional reforms involve significant hurdles, steps to enhance transparency, participation, efficiency, problem-solving, impartiality, and fair distribution will not only improve the inherent qualities of global governance but also have positive knock-on effects on its popular legitimacy.