Between Contestation and Support: Explaining Elites’ Confidence in the International Criminal Court

Abstract

The International Criminal Court (ICC) has come under challenge in recent years as some countries have decided, or considered, to withdraw from it. Against this backdrop, an emerging literature has begun to examine attitudes toward the court among the general public as a key court constituency. However, little is known about how domestic elites perceive the court. This research gap is particularly surprising given that domestic elites have a considerable impact on both public and state support of the court. This article explains why political and societal elites across world regions have confidence or lack confidence in the ICC. We present the results from a unique survey of 722 elite respondents conducted from 2017 to 2019 across six countries: Brazil, Germany, the Philippines, Russia, South Africa, and the United States. We furthermore enrich our analysis by using public opinion data to draw comparisons between elites and the general public. The analyses reveal that the views of elites are most consistently related to their perceptions of other, more well-known international organizations and their country’s relationship with the ICC. Our findings indicate both similarities and differences between how elite and public opinion about the ICC are formed, demonstrating the value of further research on elite opinion on international courts.

Introduction

During the past decade, the International Criminal Court (ICC) has been under pressure. The years running up to 2020 saw a backlash against the court (see, for example, Clarke 2019; Brett and Gissel 2020, 34–55), which not only entailed that the relationship between the African Union and the ICC had come under strain but also culminated in some states taking steps to withdraw from the court. While South Africa and The Gambia ultimately reversed their earlier withdrawal decisions, Burundi left the court in October 2017, followed by the Philippines in March 2019. In September 2020, an Independent Expert Review that sought to improve the court and the Rome Statute system delivered its final report (ICC 2020). Even more recently, the ICC Prosecutor Karim Khan’s March 2022 announcement that he had opened an investigation into the situation in Ukraine propelled the court into the headlines (ICC 2022).

Against the backdrop of these developments, questions arise about how the court has been perceived by key constituencies. An emerging literature, including in this journal, has studied public opinion on the ICC (Zvobgo 2019; Chapman and Chaudoin 2020; Dancy et al. 2020; Carrington and Sigsworth 2022). Yet little is known so far about how domestic elites perceive the court and the challenges that it has encountered. ¹ To address this research gap, our article is the first to compare, and seek to explain, domestic political and societal elite opinion on the ICC across world regions. With the term “domestic elites,” we refer to individuals who “hold leading positions in key organizations in society that strive to be politically influential” (Dellmuth et al. 2022, 13). Our approach thus goes beyond a focus on political elites who are dominant in extant scholarship on elite opinion (see, for example, Binder and Heupel 2015; Persson, Parker, and Widmalm 2019) and instead includes individuals in leading roles in government bureaucracies, civil society, political parties, media, research, and business. Assessments of the ICC from these domestic elites are key as their attitudes relate to their behavior toward the court and their views impact on both state and public support for the court (Kim 2019; Dellmuth and Tallberg 2021; Uhlin and Verhaegen 2023). For these reasons, we ask the following questions: how much confidence do elites across world regions have in the ICC; how is the court perceived by both political and societal elites; and why do these elites have or lack confidence in the ICC?

To answer these questions, this article uses unique survey data from 722 political and societal elites in Brazil, Germany, the Philippines, Russia, South Africa, and the United States (collected between October 2017 and August 2019) (Verhaegen et al. 2023). We enrich these data with public opinion data from the seventh wave of the World Values Survey (WVS-7) in five of the countries (not South Africa) (Haerpfer et al. 2022). ² We examine how much confidence individuals have in the court, how this varies between different subsections of society, and how individual-level factors can explain the variation between individuals. For the purposes of this study, we define confidence as reflecting an individual’s judgment about the performance of an authority (regarding both its intentions and outcomes) (Levi, Sacks, and Tyler 2009, 356; Turner et al. 2016). It reflects the belief, which may or may not be held by individuals, that an authority is “motivated to deliver on their promises and do what is right for the people they serve” and that this authority will also deliver on these promises (Levi, Sacks and Tyler 2009, 356). In the case of the ICC, such promises revolve around the court’s overarching goal of holding perpetrators of international crimes accountable.

With our analysis, we respond to Nathan Carrington and Claire Sigsworth’s (2022, 472) suggestion in this journal to conduct additional comparative research on perceptions of international courts, including the ICC, across countries. The related extant literature falls into four main strands. A first relevant body of research has studied the highly specialized groups of professionals that populate and surround, and move between, international criminal courts and tribunals, including international judges, prosecutors, and academics (see, for example, Christensen 2017, 2019; Batesmith 2021). However, within this literature, little is known about how other types of elites—especially, those that shape domestic politics—perceive these courts. A second related body of scholarship concentrates on public opinion and international criminal courts. While this research often has a focus on perceptions of the International Criminal Tribunal for the former Yugoslavia, it more recently expanded to public opinion on the ICC (see, for example, Klarin 2009; Meernik 2015; Zvobgo 2019; Chapman and Chaudoin 2020; Dancy et al. 2020; Carrington and Sigsworth 2022). Such initial research is situated within a wider, and partly comparative, research agenda on public opinion on international courts (see, for example, Caldeira and Gibson 1995; Gibson and Caldeira 1995; Kelemen 2012; Voeten 2013; Madsen et al. 2022).

This article then speaks to a third and growing literature on the (communicated and perceived) authority, legitimacy, and performance of international courts and international organizations more generally (see, for example, Alter, Helfer, and Madsen 2018; Grossman et al. 2018; Squatrito et al. 2018). Such a literature includes recent work on backlash against international courts (see, for example, Madsen, Cebulak, and Wiebusch 2018; Voeten 2020) as well as (often comparative) International Relations scholarship on the sociological legitimacy of international organizations (see, for example, Hurd 2003; Tallberg, Bäckstrand, and Scholte 2018; Tallberg and Zürn 2019). Within this latter work, legitimacy is typically defined as audiences’ ‘beliefs … that a political institution’s exercise of authority is appropriate’ (Tallberg and Zürn 2019, 585), which is often measured by confidence levels (see, for example, Dellmuth and Tallberg 2021). While we recognize that this research has provided relevant insights for legitimacy beliefs, we agree with Carrington and Sigsworth (2022, 454) that, while overlapping, legitimacy and confidence are analytically distinct. In this article, we thus exclusively focus on confidence in the ICC.

Finally, a growing literature on elite cuing and international law investigates how far public opinion may be swayed by elite cues or cues about the legality of a given policy (see, for example, Kim 2019; Dellmuth and Tallberg 2021). However, from this literature, it is still largely unclear why and to what extent elites themselves have confidence or lack confidence in international legal institutions such as the ICC, making an investigation into elite confidence levels all the more pertinent. Such a need is particularly pronounced given that extant research suggests that elite behavior differs from the behavior of the general public, including with regard to international law (Hafner-Burton et al. 2014; LeVeck et al. 2014; Hafner-Burton 2021).

Our descriptive analysis shows that confidence levels among the surveyed elites mostly reflect their respective country’s relationship with the court. While such a pattern is generally repeated among the general public, surprisingly high confidence levels in the Philippines and low confidence among the Brazilian general public stand out. Given the limited previous research on elite opinion and international (criminal) courts specifically, we draw on public opinion scholarship, among others, to develop three sets of hypotheses for our explanatory analysis. First, we theorize that individual elites may have more confidence in the ICC if they support addressing human rights at the international level rather than at the domestic level. Second, we propose that variation in confidence levels is associated with whether individual elites themselves value impartiality in global governance as a normative commitment that is fundamental to the legitimacy of international courts. We hypothesize that individual elites who value impartiality have more confidence in the ICC. For elites from withdrawing countries, however, we theorize that those who value impartiality more might have less confidence in the ICC since such withdrawal decisions occurred in the context of critiques regarding the ICC’s impartiality. Third, and based on public opinion research that found that individuals use heuristics when assessing international organizations (see, for example, Anderson 1998; Harteveld, van der Meer, and De Vries 2013), we hypothesize that individual elites may rely on their opinion of other institutions—namely, domestic courts or the United Nations (UN)—as shortcuts to form opinions on the ICC. To further qualify this theorization, we suggest that individuals who have less knowledge and prior experience with the ICC rely more on such associations with other institutions.

In our explanatory analyses, we find evidence for the use of heuristics by elites, albeit in a somewhat unexpected way. As our most consistent finding, our analysis most importantly reveals that elites extend their views of the UN as a particularly well-known global governance institution to the ICC. Interestingly, this association remains the same regardless of whether individual respondents have more or less knowledge of global governance or even whether they have direct prior experience with the ICC or not. Moreover, the elites’ confidence in the ICC is not as clearly related to their views of national courts in contrast to our findings based on the WVS-7 data and indications in previous public opinion research on the ICC (Voeten 2013, 429–32; Carrington and Sigsworth 2022, 465; see also relatedly Caldeira and Gibson 1995). This finding is important as it indicates that, at least in part, elite opinion about the ICC is formed differently from public opinion and may be more focused on perceiving the court as falling within a broader category of global governance institutions rather than courts. The analyses further reveal a weak indication among surveyed elites that the belief that human rights should be addressed at the international level is related to higher levels of confidence in the ICC. Finally, we observe that, in withdrawing member states, domestic elites who highly value impartiality in international organizations tend to have more confidence in the ICC than elites who value impartiality less. Based on these findings, we conclude that additional research on how domestic elites view international courts is key for a comprehensive account of international courts’ perceived performance as it is assessed across key constituencies, as well as their legitimacy and authority more generally. Taken together, our findings ultimately underline the value of, and call for, a more comprehensive research agenda on elite opinion on international courts.

The article proceeds as follows: after briefly contextualizing the relationship with the ICC of the six countries included in the study, we discuss the research data and descriptive findings. Subsequently, we outline our three lines of explanation and corresponding sets of hypotheses for why elites more specifically have confidence or lack confidence in the ICC. We then turn to our operationalization, methods of analysis, and findings as well as comparisons to perceptions among the general public. In the conclusions, we summarize the findings and reflect on their implications.

Background, Data, and Descriptive Analysis

When the Rome Statute entered into force in 2002, the ICC was seen as ushering in a new era of dealing with mass atrocity crimes. ³ During the time frame chosen for this study (2017–19), which was almost two decades later, and with ongoing investigations in about a dozen different countries, the ICC came under pressure. The most visible signs were the threats of several member states to withdraw (including Kenya and Namibia) and, in the case of The Gambia and South Africa, decisions to withdraw that were later reversed (after a change in government and a decision of the South African High Court respectively). Burundi and the Philippines, however, left the ICC. ⁴ These decisions partly occurred in the context of discussions on the immunity of sitting heads of states—including, especially, a long-standing debate on member states’ obligations to extradite Sudan’s then President Omar al-Bashir, including during visits to attend the African Union’s Assembly—and criticism of the ICC’s relationship with the UN Security Council (see, for example, Mills 2012; Han and Rosenberg 2021). Additionally, decisions to withdraw were fueled by, and interlinked with, allegations of bias against Africans within the work of the court, and corresponding criticism raised within the African Union (see, for example, Brett and Gissel 2020, 34–55; see also African Union 2017). In 2019, the Assembly of States Parties (2019b) decided to initiate a wider expert review of the functioning and effectiveness of the court following calls for such a process from a wide range of member states (see also ICC 2020).

Consequently, the 2017–19 time frame chosen for this study is of particular interest for analyzing perceptions of political and societal elites of the ICC as it captures a time at which the court was at a critical juncture. Furthermore, in the context of debates about state withdrawals, our study has been uniquely constructed to capture elite confidence toward the ICC within states that have varying relationships with the court since it includes two member states (Brazil and Germany), two non-members (Russia and the United States), and two countries that decided to withdraw (the Philippines and South Africa). Moreover, these six selected countries reflect diversity across, among others, world regions, geopolitics, geo-economics, and domestic political regimes. ⁵

How much confidence political as well as societal elites have in the ICC matters for two reasons. First, confidence (or a lack thereof) among political elites is important specifically in terms of state support and cooperation, which is likely to directly affect the court’s work and performance. Political elites (including politicians and government bureaucrats) decide whether to support the court, for example, in response to cooperation requests or with potential financial contributions. As bureaucrats, political elites may also be involved in implementing court decisions and cooperation requests as the ICC’s “compliance partners” (Alter 2014, 53). For states with a more conflictual relationship with the court, political and other elites seeking to hinder compliance might actively oppose the court (Chaudoin 2016). Additionally, resistance against international courts may also be the result of critiques from societal elites and civil society in particular (Madsen, Cebulak, and Wiebusch 2018, 204).

Second, the confidence of political and societal elites also influences how an issue is perceived among the general public by framing it one way or another (see, for example, Chong and Druckman 2007; Dellmuth and Tallberg 2021). Ultimately, elites are likely to affect how much support the court receives more broadly across society, which is not only important for perceptions of the court held among the general public, in and of itself, but also may in turn further cement elite opinion (see, for example, Kim 2019). Such a role may be taken up not only by political elites (see, for example, Wolf 2013; Kendall 2014; Chaudoin 2016; Dancy et al. 2020) but also by societal elites, for example, within media or civil society organizations (see, for example, Geis, Ristić, and Petrović 2019, 26–28; Wanigasuriya 2021, 113–20). Media representation, in turn, at times features analyses and statements by academics, emphasizing the role of both media and research elites (Madsen, Cebulak, and Wiebusch 2018, 205). Both the general public as well as legal experts and civil society organizations are seen as important audiences by the court itself, highlighting the at least perceived prominence of these groups of elites (Bexell, Jönsson, and Stappert 2021, 495). Within the remainder of this section, we briefly discuss each of these countries’ relationship with the ICC, the article’s data, and descriptive findings.

Relationships with the ICC

To begin with, the study includes two ICC member states: Brazil and Germany. Brazil has been a member since the court’s inception, forming part of a large contingent of Latin American states that signed the Rome Statute (see generally Carrasco 2010). During the 2019 Assembly of States Parties (2019a, 2), Brazil again emphasized its role as “a proud founding member” of the ICC. Germany has been a core supporter of the court and was highly active during the Rome Statute negotiations. Indeed, as Claus Kress (2006, 31) observed, “[b]y the year 2001, … Germany had become a driving force behind international criminal law.” German scholars have been described as norm entrepreneurs in their efforts to push for the introduction of German legal concepts to ICC jurisprudence (van Sliedregt 2016, 3), thereby indicating engagement with, and confidence in, the court.

Russia and the United States, in turn, are not members of the ICC. While Russia initially signed the Rome Statute, it never ratified it and is thus not a member of the ICC. Russia has acted as an observer state at the Assembly of States Parties and occasionally made statements during the General Debate. It also allowed for the ICC’s referral of the situations in Darfur and later in Libya as a permanent member of the UN Security Council. However, Russia subsequently vetoed attempts to refer the situation in Syria and was able to gain Chinese support for such a veto by citing concerns regarding the expansive interpretation of Resolution 1973 on Libya (Ralph and Gallagher 2015, 569). ⁶ In January 2016, the ICC Prosecutor gained judicial authorization to open an investigation into the situation in Georgia regarding crimes allegedly committed in 2008, ⁷ potentially by South Ossetian forces, Georgian armed forces, and Russian armed forces. ⁸ Following a (second) declaration by Ukraine to accept the ICC’s jurisdiction, the ICC’s Prosecutor in 2015 announced the decision to extend the focus of an ongoing preliminary investigation into the “Maidan events” to include crimes allegedly committed in Crimea and Eastern Ukraine after February 20, 2014 (ICC 2015; see also ICC 2016, paras. 154–83; Marchuk 2016). The preliminary investigation was ongoing at the time of data collection (2017–19), which thus preceded Russia’s February 2022 invasion of Ukraine.

The United States has historically been a supporter of international criminal justice and actively contributed to the negotiations of the Rome Statute. However, while President Bill Clinton signed the Rome Statute, it was ultimately not ratified. Under President George W. Bush, the American approach toward the ICC turned hostile, and the Bush administration pushed for the ratification of bilateral agreements designed to hinder extradition of American military personnel to the court (Schabas 2004). Consequently, previous research has shown that, by 2005, the United States was one of a handful of countries in which negative views of the court were more prevalent than positive ones among the general public (Voeten 2013, 427). ⁹ During the Obama administration, the United States again re-engaged with the court, including by participating as an observer state in the 2010 Kampala Review Conference and by actively pushing for a UN Security Council referral of the situation in Libya to the ICC (Fairlie 2011). However, after the ICC then Prosecutor Fatou Bensouda declared that she would seek authorization to open an investigation into the Afghanistan situation, and also, explicitly, to investigate US personnel, the Trump administration was openly hostile and revoked her US visa in early 2019 (BBC News 2019).

Finally, we also include two countries that just withdrew or decided to withdraw from the ICC, the Philippines and South Africa. The Philippines joined the ICC comparatively recently, in November 2011. Yet, under then President Rodrigo Duterte, the Philippines announced its withdrawal from the ICC in March 2018, a month after the court opened a preliminary investigation into potential crimes committed “in the context of the ‘war on drugs’ campaign launched by the Government of the Philippines” (ICC 2018). In the Philippines’ withdrawal notification, the decision to leave the court was justified with a critique of the court’s impartiality. ¹⁰ The withdrawal announcement came in the middle of the data collection process of the elite survey we use in this contribution. As we will elaborate below, this occurrence allows for testing the effect of the withdrawal announcement.

South Africa was initially a staunch supporter of the ICC and actively contributed to the Rome Conference negotiations with what has been described as one of the “most cosmopolitan and hopeful visions” for the newly negotiated court (Beresford and Wand 2020, 545). However, by 2016, the South African government declared that it would withdraw from the court, citing concerns about fairness, even though this decision was subsequently overturned by the South African High Court for failing to adequately involve Parliament. ¹¹ Nevertheless, it has been suggested that the African National Congress (ANC) government was particularly dissatisfied with the court’s response in the context of Article 97 consultations on South Africa’s obligations regarding al-Bashir’s arrest, which was perceived as insufficient and dismissive, including in light of concerns regarding the possibility of peace negotiations in Sudan (Beresford and Wand 2020, 550–51).

Data

Information on elite confidence in these six countries comes from the LegGov Elite Survey (Verhaegen et al. 2023). Additionally, we use public opinion data from the WVS-7 to contextualize our findings about elites (Haerpfer et al. 2022). The LegGov Elite Survey covers elites in different sectors of politics and society, working in a broad variety of issue areas in six countries across world regions. The LegGov Elite Survey goes beyond an exclusive focus on political elites (politicians and government bureaucrats) employed in most studies of elite opinion. Instead, it also includes societal elites in academia and research, business, civil society, and media since both governmental and non-governmental sectors aspire to influence issues addressed in global governance.

In the absence of an exhaustive database of elite individuals and organizations from which random samples could be drawn, the LegGov Elite Survey used quota sampling to assure that the respondents would cover a wide variety of functions, organizations, issue areas, and governance-level orientations. With a quota sample, it is not possible to extrapolate results beyond the sample. As a first step, key organizations within different sectors in the six countries were identified (Hoffmann-Lange 2009). As a second step, within those organizations, people in leading strategic positions and working on substantive issues were identified (as distinct from purely administrative responsibilities). Within each country sample, quotas required that half of the respondents would be political elites (twenty-five each for government bureaucracy and party politics) and the other half societal elites (twelve to thirteen each for business, civil society, media, and research). The goal of the sampling strategy was to acquire a diverse sample of elites in each country. Respondents were not selected based on their prior experience with the ICC. The technical report of the survey includes more information on quotas and the profiles of the respondents (Verhaegen et al. 2023).

In total, 722 elite individuals were interviewed: 124 in Brazil, 123 in Germany, 122 in the Philippines, 108 in Russia, 123 in South Africa, and 122 in the United States. Overall, the survey reached a response rate of 31.2 percent (22 percent in Russia, 25.6 percent in the United States, 35 percent in Germany, 37.8 percent in Brazil, 39.7 percent in South Africa, and 56.2 percent in the Philippines). In the context of quota sampling, response rates do not have implications for representativeness. Yet we recorded them to indicate the process of the elite survey. The response rates were relatively high for an elite survey (Walgrave and Joly 2018). The survey consisted of closed-ended questions and was conducted by telephone (79 percent of the interviews) or online when a telephone interview was not possible (21 percent of the interviews).

For five out of six countries included in the LegGov Elite Survey (the exception being South Africa), WVS-7 provides public opinion data on confidence levels in the ICC. A total of 8,896 individuals were interviewed in these countries (1,762 in Brazil, 1,528 in Germany, 1,200 in the Philippines, 1,810 in Russia, and 2,596 in the United States). ¹² This survey of representative samples of the population of these countries was fielded in roughly the same period as the LegGov Elite Survey. The periods in which respondents were surveyed is summarized in Table 1. For most countries, no exceptional events related to the ICC occurred between the fall of 2017 and the fall of 2019. The Philippines, however, decided to withdraw from the ICC in the middle of the data collection of the elite survey (the withdrawal announcement was made public on March 14, 2018, and notification was deposited on March 17, 2018). We use this turning point as a natural experiment to test the effect of the withdrawal announcement on elites’ confidence in the ICC. The Philippines’ withdrawal from the ICC became effective on March 17, 2019, and was therefore in place by the time the general population in the Philippines was surveyed in December 2019.

Table 1. Data collection period used for the survey data

Both the LegGov Elite Survey and WVS-7 asked identical questions to measure confidence in the ICC, making them comparable. Respondents in each survey were asked how much confidence they have in the ICC: a great deal of confidence (coded as three), quite a lot of confidence (coded as two), not very much confidence (coded as one), or none at all (coded as zero). ¹³ The question wording of all used variables in this article and the descriptive information of these variables is included in Appendix 1.

Descriptive Analysis: Elite Perceptions and Public Opinion

How much confidence do elites from member, non-member, and withdrawing member states have in the ICC, and how does this compare to the general public in their countries? We find that, among the surveyed elites and in line with the differing relationships of these six countries with the court, confidence in the ICC is higher in member states than in non-member states. The highest confidence levels are observed among the interviewed German and Brazilian elites (see Table 2). The active involvement of Germany in the court is reflected in the very high average confidence among German elites (2.455/3). Out of the six countries, surveyed elites in Brazil have the second-highest average confidence (1.991/3). In contrast, confidence is significantly lower among South African respondents than among surveyed elites in any other country in our sample (1.198/3). ¹⁴ In the Philippines, however, the observed confidence of surveyed elites in the ICC is neither exceptionally high nor low (1.779/3). Interestingly, the data do indicate a drop after then President Duterte’s announcement to leave the ICC. Respondents interviewed before March 14, 2018, reported an average confidence of 2.000 (fifty-eight respondents), while interviewees after that date only reported an average confidence of 1.545 (fifty-five respondents) (difference significant at p < 0.01; t = 2.926; paired t-test). ¹⁵ In non-member states, we observe medium confidence (United States, 1.614/3) to low confidence (Russia, 1.378/3) in the ICC among surveyed elites, which corresponds to the respectively distant and mixed relationship between these countries and the court.

Table 2. Mean confidence in the ICC per country

Sources: LegGov Elite Survey and WVS-7.

Note: For the WVS-7 data, post-stratification weights based on age and gender are used.

Furthermore, we inquire into variation in confidence levels across elites that are active in different elite sectors and that work with different issue areas. First, our survey includes elites in the “political” sector (that is, party politics and government bureaucracies) and “societal” sector (that is, civil society, media, business, and research). ¹⁶ One might assume that the observed pattern across member, non-member, and withdrawing states holds especially for political elites since they are closer to the decision-making process regarding ICC membership and non-membership. Yet, as can be seen in Table 2, the pattern is not clearly more outspoken among political than among societal elites.

Second, previous research has observed that the ICC receives support from human rights non-governmental organizations and academics (see, for example, Glasius 2006; Christensen 2017). Therefore, one may infer that elites working in the area of human rights are likely to have higher confidence in the ICC than elites working in other areas. The LegGov Elite Survey interviewed people working in the fields of human security (which comprises human rights), sustainable development, trade, economics, and as so-called “all-rounders.” About one-fifth of the respondents work in the field of human security. However, average confidence is nearly identical among respondents working in the field of human security (1.725/3) as among the rest of the respondents (1.749/3). Moreover, if we split the sample in member, non-member, and withdrawing member states, no significant differences are observed between respondents in the field of human security compared to other issue areas.

As a final step in the descriptive analysis, we turn to data from WVS-7. For the countries for which public opinion data is available, we detect a slightly different pattern compared to our findings from the elite survey (see the last column of Table 2). As for the elite samples, we observe high confidence in the ICC in Germany (confidence is statistically significantly higher in Germany than in Brazil, Russia, and the United States). However, confidence in the Philippines is at a similarly high level, even though the WVS-7 data were collected after the Philippines’ withdrawal from the court had become effective. For Brazil, Russia, and the United States, we note more similar, and rather low, confidence levels. Overall, we also observe that, in all countries except for the Philippines, mean confidence in the ICC is lower among the general public at large than among elites, which is a recurring finding across global governance institutions in various issue areas and with different institutional characteristics (Dellmuth et al. 2022). Following this descriptive analysis, the next section turns to explaining elite confidence in the ICC, focusing on the LegGov Elite Survey data.

Explaining Elites’ Confidence in the ICC: Theory and Hypotheses

How can we explain to what extent individual elites have confidence in the ICC? Building on extant research on the ICC and international courts more generally, we distinguish between three possible lines of explanation related to: (1) support for addressing human rights at the international level; (2) valuing impartiality and the critique of bias; and (3) using attitudes about domestic courts and the UN as heuristics.

Support for Addressing Human Rights at the International Level

As a first expectation, we suggest that the extent to which individuals have confidence in the ICC can be explained by the degree to which they support addressing human rights at the international level rather than at the domestic level. Ultimately, such a theorization suggests that confidence in the court is based on whether individuals agree with the underlying set-up and purpose of the institution as an international court addressing gross human rights violations (as opposed to, for example, specific decisions on individual cases) (see, for example, Stappert and Gregoratti 2022, 122–23). After all, there are considerable structural difficulties hindering the enforcement of human rights domestically as it is often states themselves that commit or enable mass atrocity crimes against their own citizens, potentially severely curtailing domestic courts’ ability to address such human rights violations. The question of whether human rights violations should be addressed at the international or domestic level is thus linked to the ICC’s principle of complementarity since the court’s exercise of jurisdiction is limited to instances in which there are no national prosecutions, including in cases in which a state “is unwilling or unable genuinely to carry out the investigation or prosecution.” ¹⁷

Two existing studies suggest that a belief that human rights should be addressed at the international level and, indirectly, with considerations surrounding the complementarity principle, may explain confidence in the ICC. First, the ICC itself has been found to be using references to the principle of complementarity to bolster its legitimacy. Kelebogile Zvobgo and Stephen Chaudoin (2022, 8–9) have recently observed that the ICC’s references to complementarity have increased over time and within a context of growing criticism of the court and, with it, a self-depiction of the court as fair and acting within its delegated authority vis-à-vis domestic courts. ¹⁸ Second, such an expectation is foreshadowed by another study by Zvobgo (2019), who used an online survey experiment to investigate the effects of both national interest and human rights frames on the support of the ICC among members of the public in the United States. ¹⁹ Of particular interest for this study are some of the effects that she estimates for her control variables. Specifically, she found that the conviction that international organizations (p < 0.001) and the UN more specifically (p < 0.01) are “useful for solving practical world problems” are particularly strong predictors for support (Zvobgo 2019, 1076). Indeed, she discovered that the coefficient for whether international organizations are useful is even larger than the coefficient for her initial frame treatment (Zvobgo 2019). ²⁰ Hence, we formulate the first hypothesis:

Hypothesis 1 Individuals who think that human rights issues should be addressed at the global level have more confidence in the ICC.

Valuing Impartiality and the Critique of Bias

A second set of hypotheses draws on the expectation that confidence in the ICC can be explained by the extent to which individuals value impartiality in global governance institutions. As Tom Tyler (2006) has famously shown regarding domestic legal institutions, members of the general public are more likely to follow their decisions, and behave in accordance with the law, if they believe that these institutions act fairly (procedural justice) (see also Tyler and Huo 2002). Impartiality and fairness are the backbone of international courts’ claim to legitimacy (regardless of whether such a claim is successfully implemented or not) (see, for example, Luban 2010, 579–80). Among international organizations, international courts are distinctive in that they are designed to protect judicial decision making from political influence (Alter 2008). Ultimately, international courts’ legitimacy rests on the extent to which they are able to make their decisions based on the law and the available evidence, and not on politics, so that they are—and are perceived as being—independent and impartial. Support for the ICC among core advocates—namely, civil society organizations and human rights lawyers—has been grounded in the court’s ambition to remain independent from politics (Vinjamuri 2016). Among diplomatic elites specifically, New York-based diplomats assigned to the UN headquarters in 2016 were found to support a greater focus on powerful states in the ICC’s investigations (Bocchese 2021, 286) and, thus, the impartial investigation of situations regardless of power politics.

When making decisions on whether individuals have confidence in the ICC, we thus expect an assessment of the court’s impartiality to be key. As an individual-level explanation that would allow us to observe variation and test such an expectation, we hypothesize that individuals who value impartiality are more likely to have higher confidence in the ICC. After all, in such a case, this individual’s preference for impartiality would resonate with a core value of the ICC. Thus, we expect that:

Hypothesis 2a Individuals who prioritize impartiality in global governance institutions have more confidence in the ICC.

However, this would only be the case should such an individual believe that the ICC was impartial. Individuals who value impartiality in global governance institutions might in fact be disappointed with the ICC since it is exactly the court’s impartiality that was under critique at the time of the data collection and in the context of a long focus of court investigations on situations in Africa (see also generally, for example, Han and Rosenberg 2021). Furthermore, debates about the applicability of immunity of heads of state and the ICC’s relationship with the UN Security Council have similarly revolved around questions of a feared undue influence of power politics on court investigations. ²¹ The withdrawal decision by South Africa explicitly referred to “perceptions of inequality and unfairness” and a “perceived focus of the ICC on African states, notwithstanding clear evidence of violations by others.” ²²

For the two countries included in our study that sought to withdraw and, in fact, did withdraw from the court (South Africa and the Philippines respectively), we should therefore expect to find a negative relationship between prioritizing impartiality and confidence in the ICC. Such a finding would suggest that, due to their country’s experience with, and debates on, the ICC that questions its impartiality, those elites who value impartiality in global governance institutions are disappointed with the ICC. We therefore formulate the following hypothesis:

Hypothesis 2b In withdrawing member states, individuals who prioritize impartiality in global governance institutions have less confidence in the ICC, while individuals from remaining member states who prioritize impartiality in global governance institutions have more confidence in the ICC.

Using Attitudes About Similar Institutions as Heuristics

Finally, a sizeable literature exists on the use of heuristics for assessments of international and supranational organizations, especially on public trust in the European Union (EU) (see, for example, Anderson 1998; Harteveld, van der Meer, and De Vries 2013; Dominioni, Quintavalla, and Romano 2020). For example, Tobias Lenz and Lora Anne Viola (2017) showed that people use heuristics—institutions that they are more familiar with or ideational prototypes of a perfect institution—as a point of reference when assessing the legitimacy of global governance institutions. Heuristics thus allow for more efficient (albeit potentially biased) opinion formation, especially in situations in which information that the opinion could be based on is not readily available.

The ICC is a specialized institution that most people do not regularly engage with. Consequently, both elites and members of the general public may use heuristics when forming an opinion about the court. Following the logic of Lenz and Viola’s (2017) study, the heuristics that are likely to be used relate to the ICC being both a court and a global governance institution. First, individuals may use domestic institutions as a heuristic when forming opinions on the ICC—especially domestic courts as the judicial institutions with which they are likely to be more acquainted. Such an expectation is indicated by extant research on the use of heuristics for assessments of international and supranational organizations, and limited research on the ICC specifically. Within EU research, Eelco Harteveld, Tom van der Meer, and Catherine De Vries (2013), among others, found that citizens use domestic political institutions as a heuristic in their views of the EU (see also Anderson 1998). In other words, citizens’ trust in the EU can best be understood as being rooted in their trust in domestic political institutions (Harteveld, van der Meer, and De Vries 2013; see also Anderson 1998). Similarly, using public opinion data, including from the fourth wave of the WVS, Lisa Dellmuth and Jonas Tallberg (2015) found a positive relationship between confidence in the UN and confidence in national governments (operationalized as confidence in respective national parliaments and civil services).

For the ICC, the relevant heuristic, however, should not be confidence in the respective legislative or executive branches of government but confidence in the judiciary. In this regard, the work of James Gibson and Gregory Caldeira (1995, 1998; Caldeira and Gibson 1995) has been pioneering in understanding public opinion on the European Court of Justice (as it then was). In early studies drawing on public opinion survey data from 1992, they found that rule-of-law attitudes in some EU member countries played a role in perceptions of the legitimacy of the court (Calderia and Gibson 1995; Gibson and Caldeira 1995). They concluded that, by that time, people generally had too little interest in the court to form strong opinions on it so they substituted attitudes toward it with their opinion of domestic law and institutions (or the broader global governance institution, which, in this case, was the EU) (Calderia and Gibson 1995; Gibson and Caldeira 1995).

Regarding the ICC, Erik Voeten (2013, 429–32) used the 2008 Afrobarometer responses from members of the Ugandan general public to inquire into whether trust in the ICC is based on trust in domestic courts. ²³ Voeten (2013, 432) found a “strikingly large effect” among Ugandan respondents since “[t]hose who have some trust in courts are on average twenty percentage points more likely to also trust the ICC.” Similarly, Carrington and Sigsworth (2022, 465; supplementary material, 13) found a statistically significant positive association between the UK Supreme Court’s legitimacy and ICC legitimacy among the UK public. As these findings were confined to members of the general public within individual countries, it remains unclear whether the same association applies to the elites studied here. We therefore test the third hypothesis:

Hypothesis 3a The higher an individual’s confidence in national courts, the higher their confidence in the ICC.

Alternatively, individuals may equate their view of the ICC with their perception of international organizations more generally. After all, for the Court of Justice of the European Union (CJEU, as it became), public attitudes toward the court have repeatedly been found to be at least partly derivatives of their attitudes toward the EU (Caldeira and Gibson 1995; Gibson and Caldeira 1995; Voeten 2013, 423–24; Pollack 2018, 162–72). The UN is a particularly well-known and prominent international organization so that individuals who are less acquainted with the ICC may use their pre-existing views of the UN as a heuristic to form opinions about the ICC. Considering the ICC specifically and drawing on data from Gallup’s 2005 Voice of the People survey, Voeten (2013, 427–28) found a statistically significant and positive correlation between support for the ICC and the UN. Consequently, he concluded that an individual’s support for the ICC may be “a byproduct of support for global governance more generally,” which in the survey was equally measured as positive attitudes toward the UN (Voeten 2013, 427). We thus propose an alternative version of the third hypothesis:

Hypothesis 3b The higher an individual’s confidence in the UN, the higher their confidence in the ICC.

At the same time, and conversely, existing research on trust in the EU has also found that, at times, individuals may have more trust in European institutions, especially when they are dissatisfied with their domestic institutions, thereby reversing the effect (Sánchez-Cuenca 2000; Kritzinger 2003; Muñoz, Torcal, and Bonet 2011). Individuals living in countries where domestic institutions perform comparatively less well, including due to higher levels of corruption, have exhibited more trust in European institutions and higher support of European integration, at a statistically significant level (Sánchez-Cuenca 2000; Muñoz, Torcal, and Bonet 2011, 565–67). Relatedly, Sylvia Kritzinger (2003, 237) has argued that support for European integration is driven partly by disappointment with domestic institutions that leads individuals to lend support to the EU as a form of “symbolic protest.” Recent research in this regard has revealed that both a positive and a negative spillover of trust may exist at different points in time and that such a spillover may flow both from the domestic institution to the European level and the other way around (Dominioni, Quintavalla, and Romano 2020).

A similar dynamic may be at play regarding confidence in the ICC. It might be that, when individuals have low confidence in national courts, they actually transfer their hopes for accountability to the ICC. Indirectly, higher confidence in the ICC may thus be based on a lack of expectations that the domestic courts that individuals have more experience with may successfully investigate gross human rights violations. ²⁴ Such a dynamic has been found to be at play within post-conflict societies, specifically in the former Yugoslavia, as communities that collectively experienced violence had their support for domestic institutions eroded so that these communities have increased their support for (international) legal prosecutions of gross human rights violations (Elcheroth and Spini 2009; Meernik and King 2014). Based on these observations, we propose a counter-hypothesis to Hypothesis 3a:

Hypothesis 3c The lower an individual’s confidence in national courts, the higher their confidence in the ICC.

Finally, the use of heuristics may be particularly pronounced for individuals who have limited or little knowledge of global governance more generally and who are thus likely to only have some knowledge of its most prominent institutions, such as the UN. ²⁵ After all, it has been argued that individuals will resort to using heuristics because they lack knowledge of the institution in question (see, for example, Anderson 1998). In the case of the ICC, the use of the UN as a substitute may be exacerbated for individuals who may either confuse it with the International Court of Justice, the UN’s judicial organ that is also located in The Hague, or (erroneously) assume that it is part of the (admittedly rather complex) UN system. In contrast, individuals who in fact have prior experience with the ICC, and who are thus more familiar with the institution, should be less likely to rely on heuristics.

As mentioned above, Caldeira and Gibson (1995) have argued that people have too little interest in the (now) CJEU to form a strong opinion on it and, thus, have based their attitude toward the court on either their opinion of domestic law and institutions or the EU more generally. Regarding the ICC, Patrik Vinck and Phuong Pham (2010, 439), in discussing the Central African Republic, and Carrington and Sigsworth (2022, 465–66), in discussing the United Kingdom, reveal that members of the general public who have more knowledge of the court are more likely to see the court in a positive light. Relatedly, James Meernik and Kimi King (2014, 15) have found that individuals who knew about the Geneva Conventions were substantially more inclined to support punishing individuals who broke them. However, the results of Terrence Chapman and Stephen Chaudoin’s (2020, 1317) recent survey experiment are less straightforward and suggest that, while awareness of the ICC leads to increased support when the possibility of investigations is discussed in abstract terms, this effect is not repeated for support of specified investigations (see also Carrington and Sigsworth 2022). For the purposes of this study, rather than theorizing knowledge of the court as potentially leading to greater confidence in the institution, we thus hypothesize that prior knowledge and actual experience with the court impact the extent to which individuals rely on confidence in domestic courts and the UN as heuristics:

Hypothesis 3d The positive relationship between confidence in national courts/the UN and confidence in the ICC is stronger if respondents know less about global governance.

Hypothesis 3e The positive relationship between confidence in national courts/the UN and confidence in the ICC is stronger if respondents have less prior experience with the ICC.

Operationalization and Methods of Analysis

To test the hypotheses about explanations for elite confidence in the ICC presented above, OLS regression models using the LegGov Elite Survey data were estimated. The models include country fixed effects to account for the clustering of respondents within countries. Confidence in the ICC, measured as described earlier in this article, is the dependent variable in the analyses. The dataset further allows for the operationalization of the three lines of explanation for the variation in confidence in the ICC that were laid out earlier. The full question wording and descriptive statistics of these survey questions are presented in Appendix 1. Belief in addressing human rights at the international level was measured with a survey question that asked respondents whether they thought that human rights should be addressed on the subnational, national, regional, or global level. The variable in the models shows whether a respondent opted for the global level (code 1) or another level (code 0). The global level was chosen by 69 percent of the respondents.

The role of impartiality was studied by a measure of how important respondents found impartial procedures in the workings of international organizations in contrast to transparency in decision-making procedures and taking decisions based on expertise. Respondents were asked to indicate their first and second priority from this list of three procedural qualities. Respondents who prioritized impartiality received code 2 (28 percent of respondents), respondents who selected impartiality as the second priority received code 1 (34 percent of respondents), and respondents who did not select impartiality received code 0 (38 percent of respondents).

Respondents were expected to use two types of heuristics. They were able to extrapolate their confidence in national courts or in other global governance institutions (in our case, the UN) to the ICC. Respondents were asked about their confidence in the courts in their country and in the UN in the same way as they were asked about their confidence in the ICC. Average confidence in national courts is 2.075/3, and average confidence in the UN is 1.809/3 (as a point of comparison, the mean confidence in the ICC is 1.744/3).

Finally, to test whether respondents who know less about global governance, or who have less experience with the ICC, draw more strongly on heuristics, respondents were given a short quiz of three multiple choice questions on global governance (with three response options each, the answer “don’t know” is coded as an incorrect answer). Respondents were asked about the permanent members of the UN Security Council, the city that headquarters the International Monetary Fund, and the problems that Amnesty International deals with. Not surprisingly for a sample of politically relevant elites, 62 percent of respondents answered all questions correctly, 30 percent made one mistake, 7 percent only knew one answer, and 1 percent failed on all questions. This unequal distribution of responses required a recategorization of respondents distinguishing between those who answered all questions correctly (code 1) and those who made one or more mistakes (code 0).

Respondents were also asked where they fell in the range of having “no experience at all” (code 1) to “a lot of experience” (code 4) with the ICC. We observed that 18 percent of the respondents have “little experience” and 7 percent of the respondents have “quite some” or “a lot of” experience with the ICC. A large majority (76 percent) reported no experience at all with the ICC. Because of this distribution, responses were recoded as well to a dichotomous variable where “no experience at all” was coded as 0 and all other respondents received code 1.

Findings

We now turn to testing the hypotheses on explanations for elite confidence in the ICC. The regression analysis presented in Model 1 in Table 3 only shows support for the logic of heuristics. The more confidence elites in our sample have in the UN, the more confidence they also have in the ICC (confirming Hypothesis 3b). Furthermore, respondents who reported having some experience with the ICC tended to have more confidence in the court than respondents who did not have any experience. However, in contrast to previous public opinion research, no significant relationship was observed between confidence in national courts and confidence in the ICC. Hence, the elites in our survey do not project their confidence (or lack thereof, as expected in Hypothesis 3c) in the courts in their country on the ICC. This finding differs from Voeten’s (2013, 432) conclusion that the views of individuals on the ICC are an extension of their approach toward domestic courts as well as a similarly strong positive association found by Carrington and Sigsworth (2022, 465; supplementary material, 13; see also relatedly Caldeira and Gibson 1995). However, none of these studies relied on elite surveys but, rather, on public opinion data. Since we do observe that, when respondents have more confidence in the UN, they tend to have more confidence in the ICC, we conclude that the heuristic our elite respondents use is rather that of evaluating the ICC as part of the broader category of global governance institutions than seeing the ICC as part of the broader category of courts. ²⁶

Table 3. Explaining confidence in the ICC

Source: LegGov Elite Survey.

Notes: Models are OLS regression with country fixed effects. * p < 0.05; ** p < 0.01; *** p < 0.001.

Models 2–5 further inquire into the logic of heuristics by including interaction effects to test whether respondents who have less experience with the ICC, or less knowledge about global governance, tend to draw more strongly on heuristics (Hypothesis 3d and Hypothesis 3e). The results, however, show no statistically significant interaction effects. Interestingly, the data thus do not confirm the expectation that respondents who are less acquainted with global governance or the ICC draw more on heuristics when expressing confidence in the ICC. Our findings are therefore in line with research on public opinion in the EU, which also did not find a significant moderating effect of knowledge about the EU on respondents’ attitudes toward it (Harteveld, van der Meer, and De Vries 2013, 557). Our findings add that even actual experience with the ICC does not seem to diminish the extent to which respondents draw on other global governance institutions as a heuristic. It is important to note, however, that Model 1 in Table 3 shows a direct relationship between experience with the ICC and confidence in the court. All else being equal, respondents who have experience with the ICC have significantly more confidence in the court.

Model 1 in Table 3 shows no statistically significant relationship between valuing impartiality and confidence in the ICC (Hypothesis 2a). To further inquire into the logic of valuing impartiality, we shift our gaze to Table 4. Hypothesis 2b proposed that the relationship between prioritizing impartiality and confidence in the ICC may indeed not be as straightforward as proposed in Hypothesis 2a. Hypothesis 2b suggested that the relationship may differ depending on whether respondents lived in a member state of the ICC that is withdrawing, a member state where ICC membership is less problematized in public discourse and where no withdrawal procedures have been initiated, or a country that is not a member. To empirically test this assumption, an interaction effect is estimated that evaluates whether there is a difference between withdrawing, remaining, and non-member states. Table 4 shows that the relationship between prioritizing impartiality and confidence in the ICC differs between respondents, depending on the status of their country regarding the ICC. Figure 1 visualizes the interaction effect.

Table 4. The role of impartiality perceptions in explaining confidence in the ICC in withdrawing and not withdrawing member states

Source: LegGov Elite Survey.

Notes: OLS regression model. Country fixed effects are not included since variation between countries is already accounted for by grouping respondents in member, non-member, and withdrawing member states. * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 1. Visualization of interaction terms in Model 6.

Note: Predictive margins with 95 percent confidence intervals. Source: LegGov Elite Survey.

The results show that, in non-member states (Russia and the United States), respondents who prioritize impartiality more tend to have less confidence in the ICC. For the group of withdrawing member states (the Philippines and South Africa), however, we observe that respondents who more highly prioritize impartiality tend to have more confidence in the ICC than respondents who value impartiality less. Hence, we observe the opposite of what was expected for respondents from withdrawing member states in Hypothesis 2b. Consequently, we find no support for our expectation that, due to the framing of debates on the ICC in South Africa and the Philippines, individuals who value impartiality as an institutional feature of global governance institutions are actually disappointed with the ICC. Categorizing countries according to their relationship with the ICC further shows that, overall, confidence in the ICC is higher in member states that are not planning to withdraw—that is, Brazil and Germany—than in the other two groups of countries. This finding is in line with the descriptive observations listed earlier in this article.

Finally, Model 6 in Table 4 adds another insight. Within Model 1 (shown in Table 3), we do not find a statistically significant relationship between a belief in addressing human rights at the global level and confidence in the ICC (Hypothesis 1). However, when we include these interaction effects and do not include country fixed effects but, rather, control for the clustering of respondents in countries by making three categories of countries (members, non-members, and withdrawing countries), we do observe that respondents who prefer human rights to be addressed on the global level tend to have more confidence in the ICC (Hypothesis 1), and we do observe a positive association between confidence in national courts and confidence in the ICC (Hypothesis 3a). Consequently, our findings indicate some weak support for the assumption that the extent to which an individual is convinced by the ICC’s core purpose—namely, addressing gross human rights violations through an international court—is a predictor of confidence in the ICC. We also find weak support for the logic of using heuristics from the national level. Future research is required to further specify under what conditions these logics may be at play.

Comparing Elites to the General Public

While it is clear how elite opinion toward the ICC matters in its own right, one of the main limitations of our study is that WVS-7 does not include all of the survey questions required to test the three lines of explanation for variation in confidence in the ICC. Hence, we cannot inquire whether the same associations can be observed among members of the general public as in the LegGov Elite Survey. Further research is therefore required to answer the questions of whether a preference to solve human rights problems at the international level drives confidence in the ICC among the general public. Furthermore, additional research could fruitfully inquire into whether, in the case of the general public, a preference for impartiality would be significantly related to confidence in the ICC and how this dynamic may play out across (withdrawing) member states.

However, the data of WVS-7 do not entirely leave us in the dark either. For five out of six countries in our study (South Africa is the exception), the data allow us to test the relationship between confidence in national courts and the UN as well as confidence in the ICC. The data also permit for testing whether this relationship differs depending on how much respondents know about global governance. The results of the analyses are shown in Appendix 2. For the general public, we observe a positive relationship between confidence in the United Nations and confidence in the ICC (Hypothesis 3b), which is similar to the observations in the elite survey data. However, we also observe a statistically significant positive association between confidence in national courts and confidence in the ICC (Hypothesis 3a). Such a finding is different from the findings in the elite survey data but in line with extant research on public perceptions of the ICC (Voeten 2013, 432; Carrington and Sigsworth 2022, 465). Additionally, we find that, also in the public opinion data, there is no support for the hypothesis that respondents who know less about global governance would rely more on these heuristics (Hypothesis 3d).

Finally, Appendix 3 presents analyses that use the LegGov Elite Survey data, but only include the variables that are also available in the WVS-7, to test whether the slightly different observations may be due to the exclusion of a number of covariates in these analyses. These analyses are in line with the full models presented in Table 3. The more confidence that interviewed elites have in the UN, the more confidence they tend to have in the ICC (Hypothesis 3b). Yet no significant association is observed between confidence in national courts and the ICC (Hypothesis 3a). The hypothesis about respondents with lower knowledge about global governance drawing more on heuristics is not supported by these data either (Hypothesis 3d). These findings suggest that confidence in the ICC is partly driven by other explanations among the public at large than among the specific subset of the population that can be categorized as “elites.” Similar observations have been made for other international organizations (Dellmuth et al. 2022; see also relatedly Hafner-Burton et al. 2014; LeVeck et al. 2014).

Conclusions

This study has aimed to explain why individual political and societal elites have confidence or lack confidence in the ICC. The study’s data were set within a crucial time frame for the court (2017–19) as it found itself faced with the prospect of country withdrawals. Our explanatory analysis of variation in elite confidence in the ICC provides two key insights into what role individual-level factors play. First, and most importantly, the data clearly indicate that the surveyed elites tend to draw on their opinions about the UN as a key international organization. We found a similarly consistent relationship between confidence in the UN and confidence in the ICC in the WVS-7 public opinion data, which is in line with extant research on public opinion on international courts (Caldeira and Gibson 1995; Voeten 2013, 427–28). As members of the broader public of which they were a part, the elites relied on their views of the UN as a particularly well-known global governance institution when forming opinions about the ICC.

We expected to find equally strong support for the use of domestic courts as a heuristic. As our second key insight, however, we did not observe a clear association between confidence in domestic courts and confidence in the ICC among the surveyed elites. In the main analyses, no significant association between confidence in domestic courts and confidence in the ICC was observed. Yet, when we controlled for the clustering of respondents in member, non-member, and withdrawing countries, instead of using country fixed effects, we did observe a positive association (p < 0.05). ²⁷ This less robust association between confidence in domestic courts and confidence in the ICC among elites is striking as—unlike the use of the UN as a heuristic—it is in contrast to the consistent observations of such a relationship among the general public in the WVS-7 data as well as in previous public opinion research (Voeten 2013, 432; Carrington and Sigsworth 2022, 465; see also relatedly Caldeira and Gibson 1995). Our findings from our elite survey thus indicate that, compared to the general public, elites perceive the court differently, at least in part, and more as falling within a broader category of global governance institutions than courts. Ultimately, based on our results, elites may thus harbor different expectations of the court’s work, with potentially far-reaching consequences for how the court is assessed across different constituencies.

The descriptive analysis revealed that elite confidence levels in the ICC largely reflect their country’s relationship with the court. Confidence is highest among the surveyed elites in Germany and Brazil as member states. Confidence among the South African elites is lower than in the other countries included in this study. In the Philippines, we observed that confidence among the elites surveyed after Duterte’s decision to withdraw from the court (fifty-five respondents) is lower than among the elites surveyed before this announcement (fifty-eight respondents). For the general public, however, the results were more mixed. Here, confidence is also particularly high among the German general public but surprisingly low among Brazilian respondents and high among respondents in the Philippines (surveyed after the withdrawal announcement was made). ²⁸

Together, our findings thus have important implications for understanding differences in public and elite perceptions of the ICC and of international courts more generally. Crucially, they indicate that elite confidence in the ICC is determined in part by different factors than those that determine public confidence in the court. For elites, confidence in other international organizations rather than confidence in domestic courts, as well as the relationship of the elite’s country with the ICC, relate to their levels of confidence in the ICC. Such a finding underlines the significance of further research to explain elite confidence in international courts, including—as we did—elites who may not themselves decide on their respective country’s support for the court but who may influence public opinion via elite cueing. Within the context of a broader literature on international court authority, our study thus emphasizes the value of further interrogating how the construction of international courts’ “extensive authority” among such a broader set of elites and its authority among the general public are related (Alter, Helfer, and Madsen 2018, 32–33). Indeed, our study suggests that, to fully understand international courts’ ability to effectively perform their duties in the eyes of different constituencies, as well as their legitimacy and authority, further research on the perceptions of domestic elites is key.

In addition, some of the less consistent findings of our analysis might be of interest for future research as well. With a statistically significant positive association within Model 6 (p < 0.05), but not within Models 1–5, we observed that elites who find that human rights should be dealt with on the global level tend to have more confidence in the ICC. In Model 6, we further found that, in withdrawing member states (the Philippines and South Africa), elites who highly value impartiality in international organizations tend to have more confidence in the ICC than respondents who value impartiality less. This finding is in contrast to non-members (Russia and the United States) where elites that value impartiality more tend to have less confidence in the court. This is an unexpected, albeit potentially interesting, finding. The extent to which individuals value impartiality over other institutional objectives seems to matter but in ways that are contrary to what we expected. In particular, it may be the case that it is not the overall distinction between non-member states and withdrawing states that is important in this regard but, rather, divergences across how impartiality is perceived across elites from different states. However, additional research would be required to further investigate this potentially important, yet inconsistent, dynamic, including on how exactly such respondents understand impartiality regarding the ICC. For example, Sophie Rosenberg (2017) traced critiques of the ICC Prosecutor’s strategies to tensions between “legalist” (highlighting procedural impartiality internally) and “political” (emphasizing perceived neutrality externally) understandings of prosecutorial impartiality, foreshadowing the promise of further empirical research on (potentially divergent) impartiality perceptions among elites.

Finally, elites who have more direct experience with the ICC tend to have higher confidence in the court. This could imply that these experiences were inducing confidence, but it might also be that it is elites who have more confidence in the court who choose to interact with it. Future research could fruitfully explore this further with longitudinal research designs, including for, and across, other international courts.