Leaders as diverse as United States President Joe Biden and Iran’s Supreme Leader Ayatollah Ali Khamenei have come forward on television to receive a COVID-19 vaccine. The idea is that, by getting vaccinated publicly, leaders signal that they are confident in the vaccine’s effectiveness and safety, thereby promoting vaccine acceptance among the broader population. Other heads of state, including Germany’s Angela Merkel and France’s Emmanuel Macron have revealed that they got vaccinated, but did not publicize the moment on television or with a picture. A small minority of heads of state publicly refused to get vaccinated. What is the relative frequency of these choices? And, to what extent do these leaders’ choices influence vaccine acceptance? These are the questions we address.

First, we create a public database of heads of state with systematic information on their support for COVID-19 vaccination, whether they are vaccinated themselves, and whether they distributed images of the inoculation. We find that 168 out of 173 global leaders (97%) explicitly supported the vaccination campaign. Most of them (80%) also made public that they received a COVID-19 vaccine, and 78% of those vaccinated publicized the news with a picture or video. We can therefore conclude that most global leaders thought it was important to communicate their vaccination status to the public using more than words.

Existing studies suggest that role modelling by political leaders helps to promote vaccine acceptance among the population. However, few studies support this with experimental evidence, and most focus on a small number of high-income countries. This study aims to fill this gap by providing experimental evidence on vaccine role modelling for a country in Africa – the region that is most lagging behind in achieving universal immunization coverage (1). We turn to the Democratic Republic of Congo, a country that has been particularly affected by declining vaccine confidence during the COVID-19 pandemic (2).

We conducted a survey with 600 Congolese citizens. Through a randomized survey experiment, 1/3 of respondents was prompted to consider the hypothetical vaccination of their president, while another 1/3 of respondents was prompted to consider the hypothetical vaccination of the Congolese Cardinal (of the Catholic Church). We compare their stated willingness to accept a COVID-19 vaccine to that of a control group who did not receive such prompt. While the survey was ongoing, president Tshisekedi publicly received a COVID-19 vaccine. We compare stated vaccine acceptance of respondents interviewed before and after Tshisekedi’s vaccination. Our analyses rely on multivariable logistic regressions controlling for respondent and household characteristics, and we formally assess the influence of potentially confounding unobserved characteristics.

While the hypothetical vaccination of the Cardinal had no significant impact on vaccine acceptance, the results for the president were moderated by public trust. For Congolese who report trusting the president, the survey experiment boosted acceptance from 27 to 52%. For those who mistrust the president, it decreased acceptance from 17 to 11%. When the president got vaccinated during the survey period, vaccine acceptance increased from 15 to 35%, but only for respondents who were aware of the president’s vaccination. These findings demonstrate that the impact of political leader’s role modelling is moderated by trust and depends on media outreach and access. When trust in leaders is lacking, or news on their actions is inaccessible, alternative ambassadors and effective communication methods become crucial in motivating and informing the public. This may be especially relevant in fragile states and remote regions.

In what follows, we first situate our contribution in the literature on vaccine hesitancy. We then present our database on vaccine role modelling of global leaders. Section 4 describes the context in which our case study took place. Section 5 presents our data and methods, while results are presented in Section 6. We conclude with a discussion of our findings in Section 7.

The World Health Organization (WHO) defines vaccine hesitancy as a “delay in acceptance or refusal of safe vaccines despite availability of vaccine services” (3). A growing body of evidence links vaccine hesitancy to demographic factors (such as gender and age), socioeconomic factors (including educational attainment and ethnic origin), as well as citizen’s perceived efficacy and safety of vaccines, which in their turn depend on previous vaccination history, (mis)information, and levels of trust in public authorities [e.g., (4–7)]. Given these determinants, it is unsurprising that vaccine hesitancy varies substantially, not only across countries, but also within countries, across different subsets of the population (8). Regarding COVID-19 vaccination, for instance, Solís Arce et al. (9) documented a wide cross-country variation in vaccine acceptance ranging from 30 percent in Russia to 97 percent in Nepal, but also large disparities within countries, such as an 18 percentage point difference between United States respondents who continued studies after secondary school and those who did not.

Importantly, vaccine-hesitant individuals may refuse some vaccines, but agree to others. The above-mentioned determinants may thus relate to the characteristics of a specific vaccine or vaccination process (3). In the case of the COVID-19 pandemic, factors that played a role in vaccine hesitancy included the many asymptomatic cases of COVID-19 which fed the idea of a rather harmless disease, the urgency surrounding the vaccine development which led some to doubt the reliability of clinical trials, and the social and economic disruption associated with the pandemic which turned out to be fertile ground for conspiracy theories (10–12).

In (2019), the WHO listed vaccine hesitancy among the main threats to global health (2023). Since then, COVID-19 caused a severe regress in global vaccination coverage and a sharp decrease in vaccine confidence (2, 15). To turn the tide and restore immunization progress, WHO, UNICEF, and other health partners announced “The Big Catch-Up” during the World Immunization Week (16). Through targeted efforts, the organizations aim to strengthen health care workforces, improve health service delivery and “build trust and demand for vaccines within communities” (16).

Scholars have argued that vaccine demand needs to be actively promoted by comprehensive communication campaigns to improve the perceived efficacy and safety of vaccines (17, 18). While there is a large body of evidence on what and how to communicate [see, e.g., (19–25)], less is known about who should communicate to reach maximum impact, and existing studies mainly focus on the United States [e.g., (26–28)].

Pioneering work, carried out across six countries, distinguished between the impact of COVID-19-related social distancing messages delivered by a well-known medical expert, a government official, a Hollywood actor, or a social media celebrity (29). The message had the largest impact on respondents’ stated intentions when delivered by the health expert, followed by the government official, who outperformed celebrities. The authors argue that, while celebrities have been shown to influence opinions about health and well-being at large, during times of crisis, health experts and government officials – who manage the crisis and are held accountable for it – may exert greater influence on public opinion.

The most prominent government official is arguably the head of state. Heads of state can influence citizen’s life and attitudes, not only by implementing policies, but also by communicating with the public, both with words and symbolic actions (30). This is in line with the social identity model of leadership, according to which leaders do not simply represent citizens’ attitudes and opinions, but can also change those, be it only for the subset of citizens that perceives the leader as part of the ‘ingroup’ (31, 32). That both words and actions by heads of state can have tremendous impact on crisis management, both positive and negative, has been amply demonstrated in the COVID-19 crisis. Both Brazil’s Jair Bolsonaro and USA’s Donald Trump, for instance, have aggravated the health crisis by downplaying the health risk of COVID-19, opposing measures to prevent its spread, and instead promoting remedies known to be ineffective (32–34). Conversely, several heads of state, including prime ministers Jacinda Ardern of New Zealand and Sanna Marin of Finland, have been credited with a better-than-average management of the health crisis (35).

One low-cost action that a head of state can take is to get vaccinated publicly. With this action, political leaders can arguably signal to the public that vaccines are safe and effective, thereby building public confidence in vaccines (18). Recent (quasi-)experimental research from a small number of high-income countries shows that citizens often follow cues from their political party’s elites, including when it comes to COVID-19 vaccination intentions (26–28, 36). For instance, while President Trump was mostly known for having anti-vaccination attitudes, he did receive a COVID-19 vaccine and, in one interview on Fox News, recommended citizens to get vaccinated as well. Levering this video in an online randomized experiment with millions of YouTube users, Larsen et al. (27) found that it significantly increased vaccination rates in treated United States counties.

We aim to contribute to this literature by systematically documenting the vaccine role modelling behavior of global leaders during the COVID-19 pandemic. In addition, we add experimental evidence on the impact of such role modelling from a low-income country context.

We systematically collected information on the vaccine role modelling behavior of global leaders. We started from the October 2021 version of the Political Leaders’ Affiliation Database (PLAD), which contained information on the leaders of 173 countries around the world on December, 31, 2020 (37). If, by the time of the roll-out of the COVID vaccines in a country, the head of state had changed, we updated the PLAD dataset. Supplementary Appendix 1 describes our coding procedure in detail. Figure one graphically presents the data; Panel A indicates whether heads of state supported the vaccination campaign, Panel B indicates whether they got vaccinated themselves, Panel C indicates whether an image of the vaccination was made available to the public.

We find that 168 out of 173 leaders explicitly supported the vaccination campaign (Figure 1, Panel A). Among those who did not, are two very explicit anti-COVID-vaccine presidents – Madagascar’s Andry Rajoelina and Brazil’s Bolsonaro – and three presidents who ‘tolerated’ the vaccination campaign, despite a lack of personal acts or words to support it (President Isaias Afwerki from Eritrea, President George Weah of Liberia, and Supreme Leader of Afghanistan Hibatullah Akhundzada). We found that 139 leaders (80%) received a COVID-19 vaccine but could not confirm the vaccination status of 32 leaders (Figure 1, Panel B). Moreover, 108 leaders (78% of those vaccinated) distributed a picture or a video of their vaccination (Figure 1, Panel C). We can therefore conclude that most heads of state thought it was important to communicate their vaccination status to the public using more than words.

The extent to which heads of state can act as credible vaccine ambassadors likely depends on the public trust they enjoy. Systematic reviews documented that a lack of trust in governments is associated with vaccine hesitancy and refusal across a wide range of countries (38–40). While quantitative evidence for low-income countries is scant, existing studies point in the same direction. Blair et al. (41) and Vinck et al. (42), studying Ebola outbreaks in Liberia and DR Congo, find for instance that respondents with low trust in government institutions exhibit less compliance with recommended behavior changes and a lower willingness to take up an Ebola vaccine. Moreover, Stoop et al. (6), leveraging data from 22 African countries, highlight that institutional mistrust – including mistrust in the head of state – is an important barrier to reaching universal child immunization. Such trust is often context-dependent and varies across subsets of the population (43). In some groups, anti-establishment sentiment can be so high that vaccination support by certain public figures can backfire (28).

In what follows, we turn to the DR Congo to analyze the impact of President Tshesekedi’s COVID-19 vaccination on citizen’s vaccine acceptance.

President Tshisekedi got vaccinated with Moderna on September 13^th, 2021. This is late compared to other African presidents (44). He initially went against the vaccine promotion strategy of his own government by refusing to get vaccinated for 6 months, casting doubt on the AstraZeneca vaccine,¹ and even promising to launch a Congolese ‘anti-COVID’ product at a meeting in Berlin in August 2021 (46). Less than 2 weeks after that statement, he made a U-turn and received his COVID-19 vaccine live on Congolese television. The news was distributed by diverse national media channels, Facebook, and Twitter.

It is not clear to what extent the news of Tshisekedi’s vaccination (and his initial reluctance) reached Congolese citizens. Since only 19.4% percent of households owns a television, only a minority would have been directly exposed to images of the president’s live vaccination. A larger proportion of households owns a radio (37.6%) or phone (51.8%). But, according to the latest data, internet penetration at home stands at only 1.3%, and a mere 1.5% of women and 5.5% of men aged 15–49 are estimated to access news on either radio, newspaper, or television on a weekly basis (13).

By the time President Tshisekedi got vaccinated, DR Congo had officially reported 56,000 COVID-19 cases and 1,066 deaths, or a death toll of just 0.0012% for a population of 90 million (44). But, with low testing and tracing capacity, these are likely underestimates. Looking at a highly visible (and exposed) subpopulation, namely members of parliament, the death toll reaches 5 % (47). Out of 640 Congolese parliamentarians, 32 died from COVID-19 (48). These high-profile cases fed the popular belief that COVID-19 is a disease of the urban elite, and therefore not a concern for ‘ordinary’ Congolese (49). Combined with conspiracy theories as well as the need of a largely poor population to provide in one’s livelihood, this led to overall low compliance with containment measures, such as lockdowns and restrictions on travel and public gatherings (50).

DRC received its first vaccines in March 2021, in the form of 1.7 million AstraZeneca doses from the COVAX vaccine sharing scheme, but rollout was delayed due to safety concerns, and eventually around 75 percent of these doses were re-exported to make sure they were used before they expired (48). September 2021 marked a new phase in the vaccination campaign, with the arrival of vaccines from Sinovac, Johnson & Johnson, Moderna and Pfizer (51–53). The president’s public vaccination with Moderna thus coincided with the arrival of mRNA vaccines in the country, and these were in first instance intended for 15 priority provinces, among which North-Kivu, where our research takes place (54). While vaccination rates increased with the arrival of these new vaccines (51), many Congolese remained reluctant. WHO statistics indicate that by December 2022 less than 7 doses per 100 population were administered in DRC (55).² Only Yemen, Eritrea and Papua New Guinea ranked lower.

Aside from rumors and conspiracy theories, the low vaccination rate was compounded by the country’s limited healthcare infrastructure, low numbers of health workers and broader governance issues, including rampant corruption and political instability (56). These governance issues not only affect the country’s ability to provide basic services to its citizens but also erode general trust of Congolese citizens in public institutions and President Félix Tshisekedi. A December 2021 opinion poll by the Congo Research Group (57) revealed that only 29% of Congolese surveyed had a positive opinion of Tshisekedi. In contrast, the Congolese Cardinal, Fridolin Ambongo Besungu, was trusted by 47%.

Within this context, we analyze the potential of president Tshisekedi to act as a vaccine ambassador and influence Congolese citizens’ COVID-19 vaccine acceptance by getting publicly vaccinated himself.

When systematically documenting the attitudes and behavior of heads of state regarding COVID-19 vaccination, the picture that emerges is overwhelmingly pro-vaccine: almost all global leaders endorsed the vaccination rollout, 80% publicly announced their vaccination and 62% did so with a picture or a video. We can thus conclude that most heads of state thought it was important to communicate their vaccination status to the public using more than words.

The cost of a leader getting vaccinated publicly is very low, but its symbolic value may be high, as demonstrated in recent studies from the United States. It is however unclear to what extent these results travel to different settings. In our DR Congo case study, we relied on a survey experiment to empirically verify the impact of such vaccine role modelling. The results indicate that the president’s hypothetical inoculation only increases vaccine acceptance among those who trust the president, while it depresses vaccine acceptance among those who do not. When the president got publicly vaccinated during the survey period, we find that it only increased vaccine acceptance among respondents who were aware of this fact.

These results have important policy relevance. They show that public vaccination of heads of state can only effectively serve as a vaccination advocacy tool if two conditions are satisfied. First, the said leader should be perceived as trustworthy by citizens. Second, the live inoculation should be widely communicated, preferably through diverse channels that also reach areas with low media access. These conditions were largely absent in our study area. Only 17% of respondents expressed trust in the president amidst the COVID-19 crisis, and only 18% of those interviewed after the president got vaccinated were aware of his vaccination. In such a context, vaccination of local public figures, for instance village leaders or respected older adult community members, might be more effective to improve vaccine acceptance. Indeed, our data shows that trust in local leaders is almost twice as high as trust in the president. Additionally, despite the remoteness of these territories, local news travels through word of mouth, as communities are tightly knit and easily exchange information. Mind however that our null result regarding the cardinal’s hypothetical vaccination suggests that it is not sufficient to pick any well-known and well-trusted person. Future research should delve deeper into identifying suitable ‘vaccine ambassadors’ across varied contexts.

The data and replication files for the analyses presented in this article are publicly available at https://doi.org/10.7910/DVN/OAYJCV.

The studies involving humans were approved by the University of Antwerp Ethical Advice Committee (file SHW_19_03). The studies were conducted in accordance with the local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin because we anticipated that many of our participants would be illiterate. In addition, given the local context, asking them to sign a document may even create suspicion. Hence, we provided participants with a verbal description of the study and its purpose. Verbal consent was sought from the participants and they were informed of their right not to participate or not to answer certain questions. Participants further had the option to discontinue the survey at any given point without consequences.

LC: Formal analysis, Writing – original draft, Writing – review & editing. EL: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. NS: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. MV: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing.