Discussion
The present study investigated factors associated with European public trust in scientists at a crucial moment in the COVID-19 pandemic, exploring text responses concerning COVID-19 origins and intentions to accept COVID-19 treatments and vaccines. Our findings indicate higher levels of trust in scientists within higher-income countries of Belgium, Italy and Sweden, and less trust in Ukraine, a lower-income country. Consistent with previous research,22 23 we observed higher trust among older individuals, those in higher-income countries and those with higher educational levels. In contrast to a Wellcome Trust study,22 we found greater trust among respondents over 45 years old.
Political affiliation appears to play a role in trust levels. Our study suggests that those aligned with the political left and centre exhibited greater trust in scientists compared with their right-leaning counterparts, aligning with trends observed in the USA and Europe, where public perceptions of science have been influenced by political discourse.24 25 In contrast, one German study examining changing levels of trust in science over the pandemic reported that trust increased at its outset but declined over time, more so among right-wing voters.
Health information sources also influenced participants’ trust of scientists. Participants who used print and online newspapers, magazines, television, radio, news websites or apps, websites of mainstream organisations, as well as those who obtained information through personal conversations with friends and family or from healthcare environment trust scientists in their countries more than those who do not. Our findings differ from a broader literature on information sources during the COVID-19 pandemic, which have not addressed correlations with trust in scientists.26
Participants’ direct experience with COVID-19 (close contact with COVID-19-like symptoms or knowing someone who had been in an ICU), as well as not knowing someone admitted to an ICU for COVID-19, were positively associated with trust in scientists. These apparently contradictory findings, and particularly not knowing someone admitted to an ICU, could have resulted from the large sample size leading to more variables being statistically significant.
This study also found that beliefs in conspiracy-related narratives—that the pandemic resulted from a deliberate release of SARS-CoV-2 and that 5G technology exacerbated COVID-19 symptoms—were associated with lower levels of trust in scientists. Similarly, previous studies reported that belief in conspiracy theories was negatively associated with public trust in science27 and adoption of protective behaviours.28 The emergence and circulation of conspiracy narratives (beliefs that ‘major public events are secretly orchestrated by powerful and malevolent entities acting in concert’26) and an infodemic (a plethora of correct and incorrect health information) have been crucial features of this pandemic.20 29 30
Our analyses of open-text responses effectively recast our questions about public trust in scientists. First, we found that both respondents trusting and distrusting scientists supported their claims by citing existing scientific research and expressed a need for additional scientific data and research. These results suggest that respondents—even those not trusting national scientists—sought to employ evidence-based thinking. This finding appears to counter claims that those not trusting mainstream scientific discourse act out of emotion or irrationality.15 Although this question requires further investigation, our results suggest that scepticism of existing scientific knowledge on the eve of COVID-19 vaccine rollout in Europe fuelled demands for more scientific evidence. Whereas physicist Edwin Hubble31 argued that ‘a healthy dose of scepticism’ is a prerequisite to scientific thinking, European respondents in this study were highly sceptical of expert judgement. Although Atul Gawande indicates that a scientific mindset ‘observe[s] the world with an open mind, gathering facts and testing […] predictions and expectations against them’,32 distrustful participants in the present study signalled that they undertook similar approaches: they cited published works, they gathered observations, although anecdotal, but ones they considered to be facts. Appropriating this sceptical stance and scientific processes, respondents not trusting mainstream scientists across all surveyed countries suggests a desire to ‘outscience’ the scientists.
For Feinstein,33 laypeople are outsiders to science; they rely on scientific knowledge communicated to them. This outsider status uncovers a deep paradox of trust in science: the promise of modern science is to ‘know the truth instead of just trusting what you are told’, and yet trust in science is equally essential, when laypeople cannot surmount barriers of highly specialised, complex scientific knowledge.34 The information revolution once raised hopes for greater public participation in science,19 but new concerns about a post-truth era35 have displaced these earlier aspirations. Still, some respondent efforts to ‘outscience’ the scientists may reflect a continued desire to engage more fully in scientific investigation and findings, to discover truth for themselves. Respondents’ clamours for more data may also result from a conscious strategy to set impossible standards of certainty, to generate doubts and to postpone decision-making about viral origins, vaccines or treatments. Proctor and Schiebinger have described a similar strategy among interest groups engaged in scientific controversies.36
Second, our findings raised the question about who can be considered a scientist. Respondents who did not trust scientists in their own countries nevertheless appeared to mention and trust individuals whom they considered to be scientists, but who were controversial or whose status was disputed (eg, Raoult, Trinca). Some commentators have attributed lack of trust in science as the result of ‘fake experts […] who do not actually have a credible scientific track record’.32 Yet, individuals mentioned by respondents—even those roundly castigated by mainstream science for their records during the pandemic—remain difficult to dismiss as ‘fake experts’. Our results highlight the major challenges that the European public face in distinguishing ‘real’ from ‘fake’ scientists. A recent report, for instance, found that 65% of online anti-vaccine content originated with some 12 individuals; our further investigation into these individuals found that half declared that they possessed a medical or biomedical degree.37 The lay public may label active, influential critics as scientists, even when the latter disseminate inaccurate information in their online profiles, are banned from medical boards, or dismissed as pseudo-scientists in news outlets or peer-reviewed journals.
Third, we found that the participants’ rationale concerning pandemic origins or plans for COVID-19 treatment or vaccination did not attribute central roles to scientists. Certain respondents appeared to assign more pivotal roles to states, politicians and pharmaceutical companies, suggesting that an intentional viral release would enable powerful actors to reduce certain populations and expenditures on healthcare or social support, or to benefit economically from vaccines. Although these narratives were marked by the absence of scientists, they implied that scientists were nonetheless carrying out agendas of more powerful actors. These results align with Harambam’s findings38 that online Dutch conspiracy narratives often challenge the image of science as a collective, impartial search for knowledge, and that ‘science’, particularly biomedical research, is corrupted by the corporate world.
The four principal narratives identified in participants’ justifications for believing in the deliberate release of SARS-CoV-2—demographic control, geopolitical advantage, financial profit and social control—were observed across all countries, languages and cultural areas, although with varying intensities. That these narratives appeared across all country populations included in this survey suggest a shared cultural and linguistic ‘informational ecosystem’ across Europe.
Conspiracy narratives, despite their fallacies, offer a window into the underlying anxieties of those believing and spreading them.39 Demographic control narratives were especially prevalent in countries with ageing populations highly affected by COVID-19 in December 2020, notably France, Belgium and Italy. These countries are currently grappling with debates over funding their social retirement systems and sustaining their models of social support. The geopolitical advantage narrative, attributing the pandemic to a Chinese attempt to undermine the West, appeared to resonate with populations anxious about the emergence of a multipolar world, in which Europe and the USA no longer dominate as global economic, cultural and military powers. Respondents from all countries evoked this narrative, but more frequently in Italy and Sweden. Finally, the financial profit and social control narratives, possibly alimented by fears of concentrated power in the hands of private and/or state actors, were somewhat more prominent in France and Italy, which in recent years have been preoccupied by debates over accumulation of wealth and power by these actors.
All these narratives often align with and sometimes explicitly reference the Great Reset, a multifaceted conspiracy theory suggesting collusion between governments and large corporations to orchestrate the pandemic.40 Significantly, science and scientists do not feature prominently in such rationale. Scepticism or mistrust of science and scientists does not develop in a vacuum, but is produced and sustained by historical events that shape contemporary attitudes. Past abuses, such as the infamous Tuskegee Syphilis Study in the USA and the Mediator obesity drug scandal, demonstrate that unethical research practices and misapplications of scientific knowledge leave indelible traces on public memory, eroding long-term trust in science and scientists.41–43
One singular feature of our study is that it employed an approach that was initially qualitative, then quantitative, then mixed. We first conducted a thematic analysis of online discourse (social listening), using this analysis of the infodemic to inform our survey questions, which integrated key quantitative measures and open-text answers. Quantitatively, we evaluated trust indicators, gauged the prevalence of prominent pandemic conspiracy beliefs, and assessed anticipated vaccine acceptance and treatment preferences. Subsequently, our qualitative analysis highlighted fluid definitions of what constitutes a scientist; the common practice of citing sources to support claims about pandemic origins among both trusting and distrusting participants; four principal themes (demographic control, geopolitical positioning, financial benefit and political control over citizens) that underlay distrustful attitudes towards scientists consistent across nations; and an in-depth mapping of trust dynamics among actors mentioned in participants’ open rationales.
Mobilising and combining quantitative and qualitative methodologies leveraged the strengths of each approach. Upstream of the study, qualitative methods crucially highlighted previously unidentified variables through inductive characterisation of online discourse (social listening). These methods also contributed significantly to elucidating and analysing public explanations of their claims around pandemic origins and attitudes towards scientists. In turn, quantitative methods produced additional precision about key indicators of trust in scientists and associated factors and cross-country comparisons of predominant online discourses. Shuttling between and combining mixed-methods appear especially apt for analyses of rapidly changing, polarising and complex subjects like trust in scientists and science during an epidemic or pandemic.
Limitations of the study
This study has multiple limitations. First, because the survey was administered online, respondents with better computer and internet access and higher levels of education were more likely to be recruited and to participate. Moreover, because Ipsos survey panels do not include participants over 65 years old due to uneven internet knowledge and use, we were unable to collect and analyse responses from older populations, which would have been illuminating.
Second, the survey was conducted over a 12-day period in December 2020, and in seven European countries. Our results are neither representative of all European countries, nor of high-income countries outside of Europe and middle-income and low-income countries (where populations may be less trustful of scientists). Trust in scientists may have changed significantly since December 2020. For this reason, our findings and conclusions are applicable to the seven European countries where the study was conducted. Although they reveal trust at a specific moment in the past, we employ these results to raise questions implicit in studies of trust in scientists and to encourage further investigation.
In investigating factors associated with trust in scientists, the present research did not address trust in science more generally. The survey contained specific questions about scientists’ honestly, integrity and intention to act in the interest of the public. ‘Science’ is a broad term, encompassing multiple actors, processes and practices. We compare our results with studies addressing both trust in ‘scientists’ and ‘science’, although we recognise that these terms do not have the same meaning for respondents.
The study’s large (N=7000) sample size in countries with diverse health, political, social, economic and cultural indicators, and conditions led to more statistically significant variables. A study of individual countries could shed additional light on more significant associations.
Finally, not all respondents explained their responses in text boxes. It is possible that respondents more adamant about their claims were more likely to respond to these specific questions. That said, in a previous publication, we found that responses to questions about intention to accept COVID-19 vaccination expressed conditionality.44 We would suggest, then, that these responses reflect a broad range of opinions, and not just those more convinced of their claims.
