How will the coronavirus pandemic affect trust in science, and in scientists?

In a new and very interesting article published in the Journal of Public Economics (ungated earlier version here), Barry Eichengreen (University of California, Berkeley), Cevat Giray Aksoy (European Bank for Reconstruction and Development), and Orkun Saka (University of Sussex) looked at the relationship between exposure to past epidemics and trust in science and in scientists. They use data from the 2018 Wellcome Global Monitor, which included several questions about how much people trust government and corporate scientists, and how much they trust science in general. They also measure each survey participant's exposure to past epidemics while they were aged 18-25 (what is termed the 'impressionable years'), using the EM-DAT disaster database (and the article itself has an exhaustive table of all of the epidemics they included in the appendix), and measuring exposure as the number of people in the country as a share of population in each year.

They find that:

...such exposure is negatively associated with trust in scientists and, specifically, with views of their integrity and trustworthiness. Specifically, an individual with the highest exposure to an epidemic (relative to zero exposure) is 11 percentage points less likely to have trust in the scientist (the respective average of this variable in our sample is 76%).

Interestingly, they also find that:

The effect we find is not a general decline in trust in science, but only in scientists.

It also isn't a decline in trust of elites, because there is no similar negative relationship with trust in doctors and nurses, or traditional healers. So, it really is just a loss of trust in scientists. Eichengreen et al. posit that this arises because:

Members of the public who are not familiar with the scientific process may interpret the conflicting views of scientists and criticism of some studies by the authors of others as signs of bias or dishonesty. This paper cannot analyze the first argument, due to lack of data on scientific communication during past epidemics. But we provide suggestive evidence for the second, showing that individuals with little scientific training drive the negative relationship between past epidemic exposure and trust in scientists.

Do the results matter though? It turns out that they might be quite consequential:

...we show that past epidemic exposure negatively shapes respondents’ long-term attitudes towards vaccination and reduces the likelihood that their children are vaccinated against childhood diseases.

The danger here is that we are in a world-changing pandemic right now. The generation in their 'impressionable years' right now are not only being exposed to a greater risk of harm, but they are also far more exposed to the scientific debate than ever before, due to pervasive social media. If the Eichengreen et al. paper results can be extrapolated, then this generation is going to have an incredibly negative view of scientists (but not of science more generally). The way that scientific debate has been played out over Twitter has not done any scientists any favours. It is difficult to see how this can be mitigated. At the very least, as Eichengreen et al. conclude:

Addressing concerns about corporate agendas, personal bias and disagreement in scientific communication is even more important in this light. Our results suggest that it is especially important to tailor any such response to the concerns expressed by members of the generation (‘‘Generation Z”) currently in their impressionable years.