Trust in scientists amidst the pandemic: Panel evidence from 12 countries

Yann Algan, Daniel Cohen, Eva Davoine, Martial Foucault, Stefanie Stantcheva 15 December 2021



The global pandemic – nearly two years along now – has imposed numerous challenges and hardships upon society. Amidst the terrible death toll, widespread illness, and economic turmoil, divisions have emerged within and across nations about governments’ imposition of non-pharmaceutical interventions (NPIs), such as school closures, non-essential business shutdowns, curfews, lockdowns, and mask requirements.  

Varying social compliance or noncompliance with such NPIs has variously facilitated or constrained government action, and compliance has generally decreased over time (Daniele et al. 2020). Perhaps unsurprisingly, attitudes that existed among people prior to the pandemic have proven indicative of their support for, and compliance with, NPIs. Numerous studies have linked, for example, support for NPIs to pre-crisis trust in government (Bargain and Aminjonov 2020), to local civic-mindedness (Barrios et al. 2020), or to trust in others (Durante et al. 2021). 

Trust in scientists as a key driver

It turns out, however, that trust in scientists plays a key role in public perceptions and attitudes, including compliance with and support for NPIs. In fact, experimental evidence from nine countries suggests that a high level of trust in science generates a much larger increase in compliance than does trust in government (Bicchieri et al. 2021). Our new panel data focused on 12 countries between March and December 2020 enables analysis of how varying levels of trust in scientists, government, and others in general affect compliance with and/or support for NPIs and vaccination (Algan et al. 2021). 

We distinguish between three types of trust: trust in government, trust in others (generalised trust), and trust in scientists. By comparing the marginal effects of each type of trust on outcomes that are vital to the fight against the pandemic, we can gain a clearer understanding of how social attitudes and behaviours interact with policy. Trust in scientists emerges as the main driver of attitudes and behaviours related to compliance and support for intervention measures. The degree to which trust in others and trust in government affect attitudes and behaviours is less clear cut. 

Trust levels and support for NPIs: Mixed patterns

Trust in scientists appears to be a key driver of such support. The NPI support index is 0.07 higher for high-trust respondents, relative to a mean level of the NPI support index of .75 among respondents who expressed low trust in scientists (see Figure 1). The starkest difference in NPI support between respondents who trust scientists more and those that trust them less is in the US. 

Figure 1 Difference in support for NPIs among respondents with high trust in scientists and those with low trust

By contrast, trust in government has a more limited effect on support for NPIs (see Figure 2). Moreover, this effect varies across countries. In the US, and, to a lesser extent, in Brazil, trust in government actually plays a negative role in support for NPIs. It is notable that these are the countries in which bilateral correlation between trust in government and trust in scientists is lowest (0.19 and 0.11, respectively). This is reflective of political leaders publicly opposing NPIs recommended by scientists. In such a context, it seems, trust in scientists and trust in governments have opposite effects. 

Figure 2 Difference in support for NPIs among respondents with high trust in government and those with low trust

But how does one’s trust in other people (e.g. fellow citizens) affect support for NPIs? Somewhat surprisingly, higher trust in others actually decreases support for NPIs (see Figure 3). Respondents who trust others also tend to trust them to observe social distancing, wear masks, and abide by curfew or lockdown rules. One interpretation of these results is that, because these individuals trust other people to observe and carry out these practices spontaneously, they do not see the need for stringent government-imposed NPIs. Furthermore, respondents who trust others are also more likely to support vaccination, suggesting that they understand and care about positive externalities between people. One country that illustrates this type of trust at play is Sweden. 

Figure 3 Difference in support for NPIs among respondents with high trust in others and those with low trust

Trust levels and NPI compliance

It is important to distinguish between support for NPIs and compliance with them. After all, people may comply with a mandate despite their opposition to it; likewise, individuals might be supportive of a policy in principle, yet not comply with it in practice. Despite this distinction, we observe similar patterns for support and compliance when it comes to the correlation with trust in scientists, government, and other people. 

Individuals with higher trust in scientists are also more likely to comply with NPIs, while the patterns are mixed for trust in government and trust in others. 

Change in compliance over time

As the pandemic has lingered on for nearly two years, compliance with NPIs has shifted over time. In every country studied, such compliance decreased significantly from March to December 2020, including among individuals who place high trust in scientists (see Figure 4). There remains, however, a significant gap between people based on their trust in scientists. 

Figure 4 Evolution of compliance index from March to December 2020, by trust groups

Variations of trust levels across time may be endogenous to the evolution of the pandemic itself, which in turn is a function of how well NPIs are implemented. However, considering pre-crisis trust levels (which are available for France), we see a similar relationship between pre-existing trust in scientists and support for and compliance with NPIs. 

Experimental evidence further underscores the robustness of the relationship between trust in scientists and compliance. When randomly chosen respondents were asked whether they would agree to wear a mask at home if it were recommended by either the government, a science institution (e.g. the WHO), or an individual scientist, support was highest when such a recommendation emanated from scientists. The exception was countries where trust levels were already very low, such as France and Germany. 

Support for vaccination and trust

Across countries, the share of citizens who trust scientists is strongly correlated with the share that support NPIs and are willing to be vaccinated. There were stark cross-country differences in people’s willingness to be vaccinated in December 2020 – ranging from approximately 70% in the UK, Australia, and Brazil, to around 60% in the US, Sweden, and New Zealand, and to only 36% in France and Poland – and a strong correlation with the share who trusts scientists, as displayed in Figure 5. 

Figure 5 Cross-country comparison of willingness to be vaccinated 

In contrast, trust in the government and trust in others are only weakly correlated with willingness to be vaccinated across countries. For instance, in Brazil, the UK, and the US, there is high support for vaccines but low trust in government. Thus, in line with findings at the individual level, trust in scientists is the key driver of the acceptance of restrictions or vaccinations, rather than trust in others or in the government. 

The importance of scientific independence 

Finally, the evolution of trust in scientists lies also in the perceived level of the scientists’ independence. Typically, in countries where trust in scientists has seen the most significant decrease – such as Brazil, Italy, France, or Poland – a substantial and growing share of citizens suspect that scientists are likely to hide information. This pattern emerges precisely in those countries where the initial level of trust in government at the time of the outbreak was very low. This suggests that when people associate scientists and scientific bodies with government action and political decision-making, it erodes their trust in these scientific institutions. It is therefore crucial to guarantee confidence in scientists by preserving their independence, especially in countries with low trust in government. 


Algan, Y, D Cohen, E Davoine and S Stantcheva (2021), “Trust in Scientists in Times of Pandemic: Panel Evidence from 12 Countries”, Proceedings of the National Academy of Sciences 118(40).

Bargain, O and U Aminjonov (2020), “Trust and Compliance to Public Health Policies in Time of COVID-19”, Journal of Public Economics 192.

Barrios, J, E Benmelech, Y Hochberg, P Sapienza and L Zingales (2020), “Civic Capital and Social Distancing during the COVID-19 Pandemic”, Working Paper, Chicago University. 

Bicchieri, C, E Fatas, A Aldama, A Casas, I Deshpande, M Lauro, C Parilli, M Spohon, P Pereira and R Wen (2021), “In Science we (should) trust: expectations and compliance during the COVID-19 pandemic”, PLOS One, 4 June.

Daniele, G, A F M Martinangeli, F Passarelli, W Sas and L Windsteiger (2020), “Covid-19 and socio-political attitudes in Europe: In competence we trust”,, 1 October.

Durante, R, L Guiso and G Gulino (2021), “Asocial capital: civic culture and social distancing during the COVID-19 crisis”, Journal of Public Economics 194.



Topics:  Covid-19 Politics and economics

Tags:  COVID-19, NPIs, vaccine uptake, trust, scientists, non-pharmaceutical interventions

Yann Algan is Professor of Economics at Paris School of Economics and University Paris East

President of the Paris School of Economics, Director of CEPREMAP, CEPR Research Fellow

PhD student, University of California, Berkeley.

Professor of Political Science, Sciences Po

Professor of Economics, Harvard University; Research Fellow at the CEPR; Research Associate at the NBER


CEPR Policy Research