Absolute Global Democratic Deficit

The global democratic deficit can be defined in absolute and relative terms. My concept of an absolute global democratic deficit focuses on identifying the extent to which real-world global governance falls short of the aspirational ideal of democracy. In line with common arguments in the literature (Archibugi and Held Reference Archibugi and Held1995), I expected that citizens worldwide generally perceive global governance to be undemocratic in absolute terms (Hypothesis 1a ). Despite the wealth of public opinion research on the EU (Hobolt and de Vries Reference Hobolt and de Vries2016) and other IOs (Dellmuth et al. Reference Dellmuth, Scholte, Tallberg and Verhaegen2022a), studies on perceived democratic characteristics of international institutions are rare (Hobolt Reference Hobolt2012; Lee and Lim Reference Lee and Lim2022). While people may support IOs overall (Council on Foreign Relations 2012), I hypothesized that they are not content with IOs’ democratic qualities. My hypothesis thus contrasts with prior findings of a perceived democratic “surplus” in the European context (Hobolt Reference Hobolt2012, 91), which were somewhat surprising given debates about democratic deficiencies of the EU. Moreover, I expected that citizens find world politics—including the power politics outside of formal IOs, as well as the participation of other actors like multinational corporations and non-governmental organizations—even more undemocratic (Hypothesis 1b). This seems plausible, given that ordinary citizens presumably have even less (indirect) impact on the activities of such actors than on IOs, and since world politics may appear to follow an anarchic logic more so than the rule of law (Dahl Reference Dahl, Shapiro and Hacker-Cordón1999; Monteiro Reference Monteiro2014; Waltz Reference Waltz1979).

Relative Global Democratic Deficits

An alternative approach is to benchmark global governance against existing democratic regimes rather than an ideal of democracy. This is what I call the concept of a relative global democratic deficit (see also Moravcsik Reference Moravcsik2004, 337). While Buchanan and Keohane (Reference Buchanan and Keohane2006, 406) counter such a conceptualization based on real-world benchmarks, I contend that it adds valuable nuances to the extent of perceived democratic deficiencies, allowing us to put people’s absolute evaluations into perspective. Indeed, in the European context, scholars have established national institutions as important benchmarks against which citizens judge the EU (Hobolt and de Vries Reference Hobolt and de Vries2016, 423). I break the concept of relative global democratic deficiencies down into three sub-types.

Minor. First, we may employ advanced democratic regimes as the benchmark for global governance’s democratic qualities, similar to previous research in the European context (Hobolt Reference Hobolt2012, 92). I define advanced democracies as political systems in which institutions such as elections and the rule of law are relatively far developed and well entrenched (Held Reference Held2006). While advanced democracies are a reasonable benchmark, they set a high bar such that using this yardstick only allows for establishing minor democratic deficiencies.

Many scholars of global governance argue that world politics is at least somewhat democratically deficient (Dingwerth, Blauberger, and Schneider Reference Dingwerth, Blauberger and Schneider2011). Considering that a parliament and other meaningful ways of citizen participation, as well as further democratic elements like the rule of law, are arguably lacking or deficient at the global level, I conjectured that most people perceive global governance to fall short of the standards set by advanced real-world democracies (Hypothesis 2a). This would reflect prior findings of perceived democratic deficiencies of the EU when compared with the more advanced national democratic systems in Europe (Hobolt Reference Hobolt2012, 92, 99). Besides my substantive focus on global governance more broadly, the difference between the present study and research in the European context (Rohrschneider Reference Rohrschneider2002; Sánchez-Cuenca Reference Sánchez-Cuenca2000) is that respondents compared international governance to their home countries, which I conceptualize and operationalize separately (refer to the subsection after the next one).

Major. One may also use less advanced, developing democracies as the point of reference. I define developing democracies as political systems in which institutions such as elections and the rule of law have emerged but are not yet fully entrenched (Held Reference Held2006). The comparison is a useful exercise because it allows for identifying shortcomings with respect to certain minimal standards set by existing democracies. Hence, if one finds that a regime falls short of the democratic standards set by less advanced democracies, one might speak of major democratic deficiencies.

Many global democracy theorists deem present-day world politics to be highly undemocratic (Archibugi Reference Archibugi2008; Cabrera Reference Cabrera2018; Held Reference Held1995). Given that developing democracies are generally characterized by regular elections, the rule of law, relative freedom of the press, and other democratic features, it seems that global governance falls significantly short of such standards. I expected that the lack of basic democratic institutions at the global level would also be evident to non-expert observers of the international system, so that I would find widespread perceptions of major democratic deficiencies among the international public (Hypothesis 2b). Moreover, I suspected that citizens would perceive clear differences when comparing global governance’s democratic quality to advanced versus developing democracies (Hypothesis 2c). Once again, such findings would be reflective of prior results at the European level, where it has been found that citizens of less advanced national democracies tended to find the EU more democratic than their home countries, whereas the opposite tended to be true for citizens of more advanced democracies (Hobolt Reference Hobolt2012, 92, 99).

Domestic. My next concept relates the regime quality of people’s home country to the level of democracy at the global level. Similar benchmarking approaches have been developed in the context of public opinion about the EU (De Vries Reference De Vries2018). While possibly incoherent from a third-person perspective—in cases when people consider their home country democratic or undemocratic, contrary to standard measures (Kruse, Ravlik, and Welzel Reference Kruse, Ravlik and Welzel2019)—domestic global democratic deficits may constitute the most accurate reflection of people’s subjective experience of the global system’s democratic deficiency. Other studies have argued that citizens tend to extrapolate their (lack of) trust in domestic institutions onto the EU (Anderson Reference Anderson1998; Janssen Reference Janssen1991) and other IOs (Dellmuth et al. Reference Dellmuth, Scholte, Tallberg and Verhaegen2022b; Dellmuth and Tallberg Reference Dellmuth and Tallberg2018). While they assess domestic institutional trust as an explanatory factor for confidence in IOs, the present study assumes diverging individual assessments based on domestic regimes’ democratic qualities and theorizes differences in citizens’ relative evaluations of global governance’s democratic characteristics compared to domestic regimes.

Irrespective of potential incongruences between external and subjective assessments, and in line with my earlier hypotheses about a minor and major global democratic deficit, I expected perceptions of a domestic global democratic deficit to align with expert evaluations of the democratic qualities of respondents’ home countries. That is, I hypothesized that citizens of relatively democratic countries find global governance less democratic than their country (Hypothesis 3a), but that citizens of relatively undemocratic states are more positive about global governance’s democratic credentials (Hypothesis 3b). Such a result would reflect prior findings in the European context where this association has been established (Hobolt Reference Hobolt2012, 92, 99; Rohrschneider Reference Rohrschneider2002; Sánchez-Cuenca Reference Sánchez-Cuenca2000). Table 1 summarizes my different concepts of global democratic deficiencies.

Table 1

Definitions and conceptions of the global democratic deficit

Input and Output Aspects

While the concepts presented earlier relate to the extent of a global democratic deficit, now I delve deeper into the types of deficiencies. A classic idea in the literature on the democratic qualities of political regimes is to distinguish between input and output aspects. Since Scharpf (Reference Scharpf1999) employed the input-output distinction to discuss the democratic legitimacy of EU governance in the wake of debates about the European Union’s alleged democratic deficit (Føllesdal and Hix Reference Føllesdal and Hix2006), it has become the standard conceptualization in this literature (Loveless and Rohrschneider Reference Loveless and Rohrschneider2011, 14). Other influential accounts argue for “throughput” (Schmidt Reference Schmidt2013) as a third aspect of the legitimacy of governance institutions besides input and output, or “fairness” and “purpose” as additional elements besides procedures and performance (Dellmuth, Scholte, and Tallberg Reference Dellmuth, Scholte and Tallberg2019, 634; Gregoratti and Stappert Reference Gregoratti, Stappert, Bexell, Jönsson and Uhlin2022, 122). Like related research on the EU (Hobolt Reference Hobolt2012), I stick to the arguably more established distinction between input and output, partly to avoid conceptual overlaps between aspects such as “throughput” and “input.” For the present paper, I apply these concepts to global governance’s democratic qualities.

Input. Drawing on input considerations, a democratic system is one in which governance is ultimately based on the public will, channeled through institutions that allow for citizen participation and representation (Przeworski Reference Przeworski, Shapiro and Hacker-Cordón1999; Schumpeter Reference Schumpeter1942). In other words, democracy requires that citizens have real choices and a say in how they are governed. According to proponents of such a conception, democracy is realized procedurally through free and fair elections between competing political candidates or parties (Dahl Reference Dahl1989; Norris Reference Norris1999a, ch. 11).

In world politics, there is no global parliament composed of directly elected representatives. Moreover, existing IOs have only opened up to non-governmental stakeholders relatively recently and mostly through limited participation channels for civil society organizations (Tallberg et al. Reference Tallberg, Sommerer, Squatrito and Jönsson2013). The form, extent, and content of such restricted participation by civil society actors is arguably not very visible to the global public, mostly known to expert audiences, and relatively limited in impact. Notwithstanding the complexities of internal processes and certain levels of autonomy that IO bureaucrats enjoy, by and large IOs arguably continue to be in the hands of their member states (Koremenos, Lipson, and Snidal Reference Koremenos, Lipson and Snidal2001). Hence, their relation and accountability to citizens is—at best—indirect via national governments, presumably leading people to feel rather distanced, disconnected, and disempowered regarding IOs and world politics more broadly. I therefore expected that citizens perceive the relative lack of public participation in global governance quite clearly, such that I would find a notable global democratic deficit from an input perspective (Hypothesis 4).

Output. Evaluating democracy based on output considerations shifts the focus. In this conception, democratic systems must ensure that governance is for the people as well (Hobolt Reference Hobolt2012; Munck and Verkuilen Reference Munck and Verkuilen2002; Schmidt Reference Schmidt2013). In other words, for a political regime to be considered democratic with a view to its output aspects, it must produce instrumental benefits for its subjects, for example, in terms of ensuring minimal living standards or providing social welfare that goes beyond such standards. The extent to which output aspects should be viewed as constitutive of the democratic legitimacy of governance regimes has been much debated (Steffek Reference Steffek2015), especially in the EU context where some scholars argue that the EU’s democratic legitimacy depends almost entirely on its output (Majone Reference Majone1998), whereas others consider a focus on output misleading (Bellamy Reference Bellamy, Mair and Thomassen2013).

While I am also skeptical of arguments limiting democratic legitimacy to output aspects, it is certainly conceivable that public perceptions of global governance’s democratic legitimacy are influenced by output considerations as well. Research has shown that (expected) performance and instrumental motives are important factors in explaining perceptions of the EU’s democratic qualities (Hobolt Reference Hobolt2012), public confidence in IOs such as the United Nations (UN) (Dellmuth and Tallberg Reference Dellmuth and Tallberg2015), and attitudes toward proposed global democratic institutions (Ghassim Reference Ghassim2020). Indeed, as opposed to (mostly indirect and limited) avenues for public participation described earlier, IOs are frequently in the media with their statements and actions on highly visible issues in world politics, for example, the World Health Organization (WHO) during the COVID-19 pandemic or the UN during humanitarian crises. For citizens of countries that are directly implicated by IOs’ actions, their statements and actions may be perceived as divisive and controversial—for example, the ICC in Kenya (Mueller Reference Mueller2014). Generally, however, IOs arguably tend to take relatively consensual positions on global issues on behalf of their broader membership and presumably largely endorsed by the international public. We may thus expect that citizens evaluate the actions that IOs take and the benefits they provide rather positively (Hypothesis 5a).

At the same time, we may expect that citizens are less positive about the output of world politics as a whole. In global politics, there are many more actors than just IOs. Perhaps the most salient systemic characteristic of world politics are the power differentials between countries in the international system, which constitute an essential basis or assumption of all major theories in the field of International Relations (Mingst and Arreguín-Toft Reference Mingst and Arreguín-Toft2016), and are presumably evident to any observer, including average citizens. As a result of these obvious power imbalances, benefits in international relations are arguably distributed rather unevenly. While these power dynamics certainly play out within IOs as well, they are perhaps more visible outside of them. Considering this, it thus seems likely that citizens would deem world politics in general rather undemocratic compared to IOs (Hypothesis 5b).

Explanatory Factors for Perceptions of Global Democratic Deficiencies

Numerous drivers of public opinion on the democratic quality of global governance are conceivable (Hatemi et al. Reference Hatemi, Funk, Medland, Maes, Silberg, Martin and Eaves2009). In the context of public opinion on the EU, scholars have concentrated on approaches relating to “utilitarian” considerations, “identity” aspects, as well as “cue-taking and benchmarking” processes (Hobolt and de Vries Reference Hobolt and de Vries2016, 419–23; Loveless and Rohrschneider Reference Loveless and Rohrschneider2011). Drawing on such research without any claim to comprehensiveness, my study focuses on five potential explanatory factors at the levels of countries and individuals.Footnote ¹ Let me now address them in turn.

Country-level factors. First, in the European context, economic aspects have been shown to determine public support for the EU (Anderson and Kaltenthaler Reference Anderson and Kaltenthaler1996; Carrubba Reference Carrubba1997; Eichenberg and Dalton Reference Eichenberg and Dalton1993, Reference Eichenberg2007) and perceptions of the EU’s democratic quality (Hobolt Reference Hobolt2012). Similarly, I suspected that economic aspects such as national living standards may explain public perceptions of global democratic deficiencies. Such associations could be in different directions. On the one hand, citizens of countries with higher living standards may consider present-day global governance beneficial to them and thus bearing output legitimacy. As a result, they may deem global governance more democratic than citizens of poorer countries (Hypothesis 6a). On the other hand, citizens of wealthier countries may find global governance less democratic due to cost-benefit calculations (Hypothesis 6b), assuming that richer countries tend to be the biggest contributors to the budgets of IOs and the largest providers of governmental foreign aid.

Second, in addition to economic aspects, political factors may explain public attitudes toward supranational governance, as has been shown in the European context (Anderson Reference Anderson1998; Rohrschneider and Loveless Reference Rohrschneider and Loveless2010). In particular, I expected that a country’s regime type may explain perceptions of a global democratic deficit in different ways. For one, citizens of freer countries may find global governance more democratic (Hypothesis 7a) because they have at least some indirect participation channels, for instance, by electing national representatives who then act on their behalf in international fora. Conversely, citizens of freer countries may deem global governance less democratic (Hypothesis 7b) because their participatory rights are most curtailed, relatively speaking. That is, while they do have free elections and other ways of democratic participation at the national level, such institutions are missing at the global level, so that the participative discrepancy between the domestic and international sphere is clearest for them; whereas for citizens of non-democratic countries participative avenues are limited at both the national and global levels such that they are not used to anything else. Studies of the European context have shown that public perceptions of the EU’s democratic quality are indeed inversely related with the democratic qualities of national regimes (Hobolt Reference Hobolt2012, 92, 99; Rohrschneider Reference Rohrschneider2002; Sánchez-Cuenca Reference Sánchez-Cuenca2000).

Third, a country’s population size may explain public perceptions of global democratic deficiencies in different ways. On the one hand, a larger national population may lead citizens of such countries to view global governance as more democratic (Hypothesis 8a). One reason could be that their country tends to have more say in IOs and world politics more generally, given that the most populous countries tend to be regional hegemons and often have formal or informal privileges in IOs. On the other hand, citizens of more populous countries may find global governance less democratic (Hypothesis 8b). One reason may be that rights and privileges in IOs and world politics are not distributed strictly proportionally to population size, but usually follow other principles like economic power as in the case of the Bretton Woods institutions.

Knowledge. Scholars who argue that global governance is (un)democratic base their assessment on the ontological and epistemological assumptions that there are facts about the democratic qualities of the international system, and that these facts are accessible by observing world politics (Archibugi and Held Reference Archibugi and Held1995). If they are right, then this should be possible for any observer, including average citizens. Thus we may expect that people who know more about the facts of world politics come to similar assessments as many scholars of global governance who conclude that world politics is highly undemocratic (Held Reference Held1995). Conversely, research in the European context shows that individuals with more knowledge about the EU tend to support it more and find it more democratic (Armingeon and Ceka Reference Armingeon and Ceka2014, 95; Hakhverdian et al. Reference Hakhverdian, Van Elsas, Van Der Brug and Kuhn2013; Hobolt Reference Hobolt2012, 97–99; Inglehart Reference Inglehart1970; Karp, Banducci, and Bowler Reference Karp, Banducci and Bowler2003). This may be since the EU is relatively democratic (Dingwerth, Blauberger, and Schneider Reference Dingwerth, Blauberger and Schneider2011) and, in this sense, an arguably somewhat atypical IO. Yet if the link between more general knowledge and stronger support holds for other IOs, we may in fact expect greater knowledge about world politics to lead people to find global governance more democratic.

To further tailor my argument, I thus limit it to knowledge about arguably undemocratic characteristics of global governance such as the permanent member states’ veto rights in the UN Security Council (Bosco Reference Bosco2009) or the weighted voting system and leadership conventions in the Bretton Woods institutions (Lesage et al. Reference Lesage, Debaere, Dierckx and Vermeiren2013). Indeed, while there are long-standing academic debates about alleged democratic deficiencies of the EU (Bartl Reference Bartl2015; Goodhart Reference Goodhart2007; Majone Reference Majone1998), deficits in other IOs are essentially undisputed, even by those who argue that the EU is democratic (Moravcsik Reference Moravcsik2004, 363). I hypothesized that if and when ordinary people come to know about such features of the international system, they should be especially likely to deem global governance undemocratic (Hypothesis 9). Finding such an effect would mean that more information about relevant aspects of global governance would heighten public perceptions of global democratic deficiencies.

Framing. Another potential driver of public perceptions relating to global democratic deficiencies is framing. Global governance has become increasingly contested in recent years, as populist leaders and authoritarian states have challenged existing international institutions (Posner Reference Posner2017; Voeten Reference Voeten2020). Especially under the assumption that ordinary citizens are often not acquainted in detail with the institutional setup of global governance and other aspects that are considered democratic or undemocratic (Dellmuth Reference Dellmuth2016), framing may be an important driver of individual perceptions. Since there seems to be widespread agreement among scholars that there are immense democratic deficiencies at the global level (cf. Nye Reference Nye2001), we may suspect that standard arguments on supposed global democratic deficits are more convincing than claims that global governance is in fact democratic.

One of the principal arguments for the democratic deficiency of global governance is the lack of avenues for public participation in global decision-making, since there are no institutions such as a world parliament, for instance (Falk and Strauss Reference Falk and Strauss2001). Conversely, a key assumption at the heart of our present-day international system and a core argument for its democratic sufficiency is that citizens are adequately represented in global governance by their national governments. We may expect the former argument to be more convincing, given the apparent scholarly consensus on the existence of a global democratic deficit. Yet studies on counter-framing effects suggest that two frames pointing in opposite directions tend to cancel each other out (cf. Chong and Druckman Reference Chong and Druckman2012), which may especially apply in the present context where strong public views are presumably rare. Thus, my hypothesis that common framings negatively affect perceptions of the democratic qualities of global governance, even when a negative frame is countered by a positive one (Hypothesis 10), remains to be tested. All my hypotheses are summarized in table 2.

Table 2

List of hypotheses

* Refer to note 2 for links to my pre-registrations.

Absolute and Relative Global Democratic Deficits

In order to investigate the existence and extent of a perceived absolute global democratic deficit, I asked respondents how democratic or undemocratic they consider IOs or world politics respectively to be. By not specifying a real-world benchmark for people to base their assessment on, respondents were implicitly asked to evaluate the democratic qualities of global governance based on their perception of ideal-type democracy. As answer choices, I offered unidirectional response scales from “not democratic at all” to “completely democratic.”

I inquired about people’s perceptions of a minor democratic deficiency in global governance by asking respondents how democratic they deem IOs or world politics respectively to be compared to “advanced democracies like Germany.” Similarly, regarding major global democratic deficits, I asked respondents to evaluate IOs or world politics respectively against the benchmark set by “developing democracies like South Africa.” Evaluations could range from “much less democratic” to “much more democratic.” While one may argue that these operationalizations assume some familiarity by respondents with the political systems of the countries used as benchmarks, I believe that Germany’s and South Africa’s characterizations as examples of the underlying concepts of developed and developing democracies respectively provided survey participants with the necessary information for the present purpose. In the European context, scholars have relied on two separate questions about satisfaction with democracy at the national and EU levels respectively, in order to derive relative deficits indirectly (Hobolt Reference Hobolt2012, 96). Here, I employed one question for each type of relative global democratic deficiency to obtain direct measures. Online appendix 1 further discusses my operationalizations. I randomly allocated respondents to subgroups where they were asked either about a minor global democratic deficit or a major one, but not about both. This was intended to ensure that differences between responses to these two questions do not just arise because respondents aimed to be logically consistent in their answers, but because the two concepts are indeed capable of capturing different extents of perceived global democratic deficiencies.

My method of testing for the existence and extent of perceived domestic global democratic deficits resembled my approach to the other two kinds of relative democratic deficiencies that I conceptualized. I asked respondents to indicate if they deem IOs or world politics to be more or less democratic than their country of residence, that is, the survey country. I focused on the country of residence because it is likely that this is the country whose political system people generally tend to be most familiar with as it usually has the greatest impact on their daily lives. I refrained from ambiguous terms like “home country” because in cases of immigrants, in particular, it would not have been clear what the question was about—their country of origin or their country of residence.

Input and Output Aspects

Regarding input deficiencies, I asked the following question: “From your perspective, how much or little opportunity do citizens like yourself have to influence “the policies of international organizations” or, in the broad definition group, “world politics?” Based on this input perspective, global governance may be considered democratic if people generally feel that they have a say.

Of course, in any political system—including democratic ones—individual citizens only have a limited impact on public policies. The more important consideration is therefore whether citizens believe that people like them (for example, those of a similar political conviction) are adequately represented and sufficiently influential. Hence, similar to other surveys in different world regions (Dulani Reference Dulani2009, 35; Gallup Reference Gallup1996, 252; National Survey of Libya 2011, 27), I operationalized this conception as “citizens like yourself” rather than just “you.” While respondents’ interpretations of “citizens like yourself” may differ, I do not consider this an issue for the present purpose, given that respondents may legitimately conceptualize like-minded citizens along different dimensions (for example, nationality, social status, age, ethnicity, or political convictions), which are all valid interpretations for the present purpose of identifying an input-based global democratic deficit.

On the output side, I asked the following question: “In your opinion, to what extent are the policies of international organizations”—or “is world politics”—“to the advantage or disadvantage of citizens like yourself?” The purpose of this wording was to aim at the potential benefits or costs that global governance may cause for citizens, going to the heart of output-focused evaluations of political regimes. In order to prompt respondents to think broadly about the advantages and disadvantages that global governance provides, I intentionally kept the operationalization general, not specifying the (dis)advantages as material or otherwise.

Knowledge and Framing Experiments

To find out if more relevant knowledge causes people to deem global governance less democratic, I gave respondents the answers to three questions relating to characteristics of the international system, which are often considered undemocratic in different ways: first, the vote share system in the World Bank, which is based on economic contributions and power, rather than more conventional democratic principles such as proportional representation; second, the identity and nature of privileges of the five veto powers in the UN Security Council, which runs counter to principles such as sovereign equality that may be considered more democratic; and third, the WHO’s main funding sources, which may be considered undemocratic due to a potential reliance on private donors and thus special interests. Exposing respondents in the knowledge treatment group to these three pieces of information allowed me to estimate the effect that more relevant knowledge has on public attitudes regarding global democratic deficiencies.

While my knowledge treatment arguably constitutes a kind of framing in itself, my second experiment (which another group was randomly assigned to) exposed respondents more classically and directly to common normative framings about global democratic deficiencies. The treatment vignettes distilled the gist of standard arguments: “Some argue that there is no public involvement in international organizations, which makes them undemocratic. Others argue that citizens are adequately represented by their national governments, which makes international organizations democratic.” In the respondent group where global governance was defined more broadly, the term “international organizations” was simply replaced by “world politics.” I exposed respondents to only one treatment with opposing frames (rather than separate treatments on global democratic deficiency versus sufficiency), in order to prevent “experimenter expectancy effects” (Sanderson Reference Sanderson2010, 47–49), since opposed frames capture mixed messages in real-world debates more accurately, and due to practical considerations of statistical power given the available sample sizes. Online appendices 1.3 and 2 present my knowledge and framing treatments in the second main round, while online appendix 3 summarizes additional survey elements in the two main rounds.

Country Selection

I conducted the survey in 17 diverse countries across the global South, North, East, and West. Several criteria were used in the selection of survey countries. First, I wanted to include a variety of countries in terms of population size, average national income, and regime quality, due to the theoretical considerations outlined earlier. Second, I wanted my sample of countries to be as globally representative as possible, that is, geographically and culturally diverse. After conducting a pilot survey in the UK, I ran the two main rounds of survey experiments in 16 countries, for which the aforementioned characteristics are summarized in table 3.

Table 3

Survey country characteristics

Note: Population figures are based on data from the survey years 2019 and 2021 respectively and stated in millions (United Nations 2022).

Freedom ratings are the summary assessments of Freedom House for the survey years 2019 and 2021 (Freedom House 2021). In my study, I concentrate on the Freedom House data (rather than alternative measures) due to its availability for the survey years and its unambiguity offering one summary measure.

GDP per capita is stated in 2015 US Dollars for the survey years 2019 and 2021 (World Bank Group 2023a).

Survey Implementation

After translating the questionnaire into the target languages (online appendix 4), I programmed the survey experiments using Qualtrics. I collaborated with YouGov in 2018 for the pilot survey, Dynata in 2019 for the first main survey round, as well as Qualtrics and its partners (Cint, Dynata, Lucid, and Toluna) in 2021 for the second main survey round. Thus, my first main round took place before COVID-19 was known, while the second main survey was fielded in the middle of the pandemic, allowing me to examine to some extent the potential stability of or variations in views on global democratic deficiencies in times of a global crisis compared to more ordinary times. In table 4, I summarize the main characteristics of my different survey rounds.

Table 4

Characteristics of the different survey rounds

For each country, I used sampling quotas reflecting population averages for various regions, as well as age-gender groups, and—in the second main survey round—education. Online appendices 5 and 6 provide further details on my samples and weights. The aggregate sample amounted to 41,966 respondents across the 17 survey countries (Ghassim Reference Ghassim2024b). Due to deviations from my target quotas at the sampling stage, I used entropy balancing (Hainmueller Reference Hainmueller2012) to reweight the survey data, so that it matched the targeted quotas as closely as possible. In the pooled analyses that follow, each country sample is weighted equally rather than proportionally to population size, so that the results are not dominated by the biggest countries in the sample (especially China and India).

Input and Output Perceptions

Let me now delve deeper into input and output aspects of global democratic deficiencies. Average perceptions of input-related democratic qualities are remarkably robust across both the narrow and broad definitions of global governance (figure 5). When conceptualized in terms of IOs and world politics, 73% and 72% respectively across countries in the second survey round stated that citizens like themselves have little influence on global governance, as expected in Hypothesis 4. As online appendix 8.1 shows, the difference in means between the narrow and broad global governance conceptions is not statistically significant. Moreover, note here that the “very little” input category constitutes the plurality of responses across countries (figure 5).

Figure 5

Input- and output-related global democratic deficit

Survey questions: “From your perspective, how much or little opportunity do citizens like yourself have to influence [the policies of international organizations / world politics]?” and “In your opinion, to what extent [are the policies of international organizations / is world politics] to the advantage or disadvantage of citizens like yourself?” (refer to online appendix 1.3)

Note: Refer to the notes below figure 1. These data are from the second main survey round in which the input and output questions were included (table 4).

Further data: Tests for statistically significant differences are provided in online appendix 8.1.

At the country level, negative attitudes on input to IOs range from 42% in Egypt to 95% in Hungary. Regarding world politics, negative attitudes range from 46% in Egypt to 96% in Hungary thinking that citizens have little opportunity for influence (online appendix 10.9). The negative public sentiment regarding the input qualities of IOs is expressed well by a respondent in my Kenya survey: “I have never witnessed a situation in Kenya where citizens have been asked to give their views, improvements, etc., on international organizations. It would be a good thing for citizens to feel actively involved in how these organizations are run/managed because whatever the organizations do directly affects citizens who have no say in them.” Regarding world politics, one respondent in Turkey simply summarized his apparently widely shared sentiment as follows: “I don’t think we can contribute to world politics as individuals.”

My findings suggest—as expected—that people generally perceive a global democratic deficit based on input considerations and that a strong perception of a deficit in this respect is indeed widespread. The divergence compared to my main results on the global democratic deficit may be explained by people’s implicit associations with the term “democratic” that I employed in my questions on absolute and relative global democratic deficiencies. Even though I explicitly defined “democratic” in terms of citizen participation, it may be that respondents evaluated democracy in IOs and world politics from other perspectives as well. For instance, as noted, some people may hold a state-centered perspective on democracy in IOs, rather than citizen-centered one (Lee and Lim Reference Lee and Lim2022).

Another explanation for the results here is that people generally feel like they cannot influence politics much—neither at the global level nor at the national level. Thus, the high percentages of input-related deficiencies would not be specific to global governance, but observable in any governance system. This line of thought is evident in the following comment by a respondent in my Australia survey: “Only people in politics and the rich have a say, us citizens can vote but each person is 1/1000000 with a small voice.”

Lastly, one may note that my question asked about the extent of opportunities for people to influence IOs or world politics; and if people deem it democratically justified that citizens should not have much opportunity to influence global governance, then my finding of widespread input deficiencies do not per se indicate democratic deficiencies. Conversely, the results here may be taken as most indicative of widespread public perceptions concerning global democratic deficiencies, given that respondents may interpret the absolute and relative questions with different reference points in mind, for example, a world without IOs (see ms. p. 11).

The picture changes significantly with respect to output considerations (figure 5): While an average of 58% of citizens across the ten countries in the second survey round consider the policies of IOs to be to their advantage, perceptions are not exclusively positive—partly due to contemporary events, as the following comment from my Colombia survey in 2021 illustrates: “Unfortunately, the WHO has NOT been of any use in the pandemic.” At the country level, positive attitudes about IOs’ outputs range from 34% in Colombia to 78% in Egypt (online appendix 10.10). Meanwhile, an international average of 53% deem world politics as a whole to be to their disadvantage (figure 5). At the country level, negative attitudes about world politics’ output range from 35% in Egypt to 69% in Colombia. As online appendix 8.1 shows, the difference in means between the narrow and broad global governance conceptions is highly statistically significant. We may therefore conclude that the output-based democratic quality of global governance depends on how broadly it is conceived.

Output perceptions of IOs tend to be relatively positive (in line with Hypothesis 5a), whereas output perceptions of world politics as a whole tend to be rather negative (in line with Hypothesis 5b), thus helping to explain greater perceptions of a global democratic deficit when defined more broadly. In general, output perceptions are significantly more positive than input perceptions, which helps explain the relatively positive public views of global governance’s democratic qualities overall. Indeed, online appendices 8.2 and 8.3 show that citizens’ perceptions of a global democratic deficit (based on my narrow and broad conceptions) are more strongly associated with perceived output qualities of global governance than with perceptions of its input qualities—a finding that contrasts with research on the EU’s perceived democratic qualities where procedural (that is, input) aspects have been found to be more important (Hobolt Reference Hobolt2012, 89).

One interpretation of these findings is that input aspects are not so important for citizens’ overall evaluations of global governance’s democratic qualities as long as basic standards are fulfilled (for example, national governments having a say on citizens’ behalf), and so long as people deem global governance’s outputs in terms of public goods acceptable. At the same time, this output-dependence—without clear links to input channels and public appreciation thereof—may place IOs in a delicate position, given that national governments have incentives to claim successes for themselves while blaming IOs for failures. Lastly, one may argue that the global democratic deficit’s essence from an output perspective is that IOs’ competence domains are too restricted. Thus, while the international public on average perceives that IOs bring advantages, they could benefit even more if IOs’ functional remit was broadened (Fabre, Douenne, and Mattauch Reference Fabre, Douenne and Mattauch2023; Ghassim, Koenig-Archibugi, and Cabrera Reference Ghassim, Koenig-Archibugi and Cabrera2022), which would thus imply global democratic output deficiencies despite the apparent public content regarding existing IOs.

Country Characteristics as Explanatory Factors

Moving on to my exploration of country characteristics as possible explanatory variables for public perceptions of global democratic deficiencies (or a lack thereof), I concentrate on three factors: economic power, regime type, and population size. Figure 6 shows the results for perceptions of an absolute global democratic deficit based on my narrow definition in terms of IOs, while figure 7 presents the findings for my broader definition in terms of world politics.

Figure 6

Absolute global democratic deficit (IO definition) by country type

Note: The categories for national income levels are based on the World Bank Group’s (2023b) official categorizations.

Regime types are based on Freedom House’s (2021) categories. Population data comes from the United Nations (2022), and is split into the three categories at the cutoffs of 65 million and 1 billion inhabitants.

Figure 7

Absolute global democratic deficit (world politics definition) by country type

Note: Refer to the notes below figure 6.

First, I observe that perceptions of global democratic deficiencies seem to be positively associated with countries’ income categories—a finding that appears to hold for both the narrow and broad definitions. While my multivariate ordered logistic regressions with country fixed effects do not confirm these associations (online appendix 11.1), the finding is corroborated by my analysis of global democratic input deficiencies, perceptions of which also increase along with countries’ average income levels (online appendix 11.2). However, the picture is somewhat complicated by my analysis of output deficiencies (online appendix 11.3): Citizens from high income countries are more likely to find global governance (defined both narrowly and broadly) democratically deficient than people from lower-middle income countries, but less likely than citizens of upper-middle income countries. In my univariate regression analyses, GDP per capita is positively and significantly associated with output quality, while the association is reversed when controlling for freedom scores and population size. In general, the results from these different analyses arguably tend toward the finding of more widespread public perceptions of global democratic deficiencies as national income levels increase (in line with Hypothesis 6b). This may be due to cost-benefit calculations, as citizens of richer countries believe that they contribute more than what they get in return from IOs and world politics.

Second, my analyses appear to show that the correlation between regime type and perceptions of global democratic deficiencies is not linear, but that citizens in the freest countries find global governance least democratic—a finding that again seems to hold for both the narrow and broad definitions of global governance (figures 6 and 7). Once again, however, my regression analyses do not confirm the statistical significance of this finding (online appendix 11.1). The association between freedom in a country and perceptions of global democratic deficiencies seems to hold in my input and output analyses as well (online appendices 11.2 and 11.3), although the “partly free” countries tend to cluster with “free” countries rather than “not free” countries as in the analyses of absolute global democratic deficits. Yet again the regression analyses with country fixed effects paint a more diverse picture with significant associations in different directions (online appendices 11.2 and 11.3, respectively). Thus, the evidence is once more mixed—with a tendency toward a (bivariate) negative association between domestic freedoms and perceptions of global democratic deficiencies (in line with Hypothesis 7b). Such inverse perceptions of national and supranational institutions have been perceived in the EU context as well (Sánchez-Cuenca Reference Sánchez-Cuenca2000). One interpretation is that citizens in freer countries are most discontent with the curtailment of their participative rights in global governance compared to domestic politics.

Third, my analyses appear to show that the more populous a country, the more democratic citizens find global governance. This association appears in my analyses of absolute global democratic deficiencies, although here it is only citizens of “very large” countries (that is, China and India) who clearly perceive global governance to be more democratic—both narrowly and broadly defined (figures 6 and 7). While my regressions do not corroborate the significance of this finding (online appendix 11.1), the input analyses in online appendix 11.2 provide further evidence for this result by showing that citizens in more populous countries perceive more input opportunities in global governance. Lastly, with regard to output deficiencies, this positive association is only evident in the broader definition of a global democratic deficit with respect to world politics, not when more narrowly defined in terms of IOs, and my regression analyses once more complicate the picture with significant associations in different directions (online appendix 11.3). Overall, there is mixed evidence for a positive association between national population size and perceptions of global democratic qualities (in line with Hypothesis 8a). One interpretation is that citizens in larger countries may find global governance more democratic because they feel that their country has more sway in international institutions as a result of its size.

Knowledge and Framing as Drivers of Perceptions

The principal result of my knowledge and framing experiments is that, in my pooled analysis across all survey countries, there are no main treatment effects—contrary to Hypotheses 9 and 10 (figures 8 and 9). Disaggregating the analyses by countries, there are instances in which significant differences between control and treatment groups are observable—both in expected and unexpected directions (online appendices 10.11 and 10.12). Yet, as in the cross-country analyses, most treatments do not yield any significant results at the country level.

Figure 8

Knowledge and framing treatment effects, IO definition

Notes: The plot shows the control and treatment groups for different conceptions of the global democratic deficit in the second main survey round. The y-axis represents parts of the six-point scales for the different questions, with 3.5 constituting the middle. The plot shows the point estimate and 95% confidence interval of each experimental condition. None of the differences-in-means between the treatment and control groups is statistically significant at conventional levels. Online appendix 9 contains the detailed data and results of each experiment.

Figure 9

Knowledge and framing treatment effects, world politics definition

Note: Refer to the notes below figure 8.

Regarding the first experiment, the aggregate null findings are especially surprising in light of the fact that knowledge has been shown to be a significant factor in the context of public attitudes toward the EU (Karp, Banducci, and Bowler Reference Karp, Banducci and Bowler2003). One interpretation is that the treatment was not strong enough to capture the supposed effects of greater knowledge about undemocratic characteristics of global governance on perceptions of global democratic deficiencies. Indeed, my experiment does not test the effect of greater relevant knowledge more broadly, but rather the specific effect of three pieces of information about global governance (some of which respondents may already be aware of). Another possible reason for the null result is that citizens’ assessments of global governance as not highly undemocratic reflects their general support for international organizations and cooperation, as well as the output benefits they perceive (Ecker-Ehrhardt Reference Ecker-Ehrhardt2014). Democratic deficit perceptions may thus be somewhat decoupled from specific institutional characteristics that scholars point to when they highlight democratic deficiencies in global governance.

Lastly, it may be that the treatments do not provide knowledge that is relevant to people’s views on global democratic deficiencies. For example, while people may already know and have formed their judgment based on the fact that delegates at the UN are not directly elected, membership and veto powers in the Security Council may not be relevant to their views on a global democratic deficit. However, respondent comments such as this one from my Egypt survey illustrate that my chosen knowledge treatments are relevant for at least some people’s perceptions of global democratic deficiencies: “The right of veto makes decisions apply only to small countries. Thus, there is no democracy in the United Nations system, especially the Security Council.”

Similarly, framing the debate around the supposed global democratic deficit, as it often is, does not affect citizens’ perceptions of a global democratic deficit along any of my measures in the cross-country analysis. Yet the arguments provided do resonate with some respondents, as evidenced by the following comment in my survey in France reflecting on IOs: “These organizations operate with member countries and their official representatives. These are not all democratically elected as we understand in France and Europe.”

One potential explanation for the observed null effect at the aggregate level is that the arguments for the democratic deficiency and sufficiency of global governance cancel each other out. Indeed, this neutralization effect is what scholars have observed in other experiments with opposing frames (cf. Chong and Druckman Reference Chong and Druckman2012). Another possible explanation is that the treatments were—yet again—not strong enough. It seems conceivable that a broad debate on global democratic deficiencies would bear greater potential to affect public attitudes in this respect, since the arguments on both sides would be elaborated more clearly than possible within my limited survey experiment. Nonetheless, given the findings, I conclude that public views of global governance’s democratic qualities are more stable than anticipated and—by and large—not affected by two factors that are thought to be influential.