Covid-19 and Social Trust

Principal Investigator: Arnstein Aassve

Covid-19 and social trust: a consensus view

Today, the mainstream epidemiological view is that in the absence of a vaccine deployable at a large scale, the most effective strategy to counter the Covid-19 pandemic is through massive and swift behavioral change. Governments around the world are engaged in the enforcement of unprecedented ‘social distancing’ measures that, according to the current consensus, would reduce the pace of the contagion. Empirical studies from previous outbreaks, such as Ebola, have shown that social and institutional trust is a key variable to explain behavioral change at the population level. Individuals, in fact, tend to adopt behaviors that are solicited by trusted public actors, whereas a lack of trust in public and health officials leads to underutilization of health services and to ignoring health advice. Against this background, in April 2020 the World Health Organization advised that, to achieve behavioral change, “it is critical to gain understanding of issues such as: trust in health authorities, recommendations and information; risk perceptions; acceptance of recommended behaviors; barriers/drivers to recommended behaviors; misperceptions and stigma”. Thus, insights from social and behavioral sciences can be “used to help align human behavior with the recommendations of epidemiologists and public health experts.” (van Bavel et al. in press). Only in this way, can the course of the pandemic be changed and dealt with.


The current ongoing study

On Monday, April 6, we launched what is, to the best of our knowledge, the first pre-registered empirical study aiming at monitoring social and institutional trust in the United States during the Covid-19 pandemic, and at assessing the relation between trust and behavioral change during the health emergency. The research team is recruiting about 100 participants per day via Amazon Mechanical Turk. Our survey asks participants to answer a set of closed- and open-ended questions concerning trust towards other people and key health and public institutions. Participants are also asked to answer questions concerning their attitudes, personal experience (including emotional reactions), level of knowledge and behavior with respect to the Covid-19 crisis. Our aim is consequently to have a longitudinal design that enables us to track and monitor how the measured variables change as a function of the unfolding of the Covid-19 pandemic.  The main goals of our study are three: first, (i) we assess the relation between social and institutional trust and behavioral change during the Covid-19 pandemics. Our very preliminary results (n=444), show that trust in different institutions and where and how often individuals acquire information have distinct effects. For example, while higher trust in the president is linked with lower level of self-reported compliance with standard WHO and CDC recommendations on frequent hand-washing and social distancing (p < 0.05), it is positively linked with self-reported prosocial behavior (e.g., volunteering; p < 0.01). As a potential signal of backfire effects, which should be investigated further through broader sampling and mediation analyses, we also found that individuals who more frequently acquired Covid-19 related information through CDC and WHO websites reported lower compliance with their recommended behavior (p < 0.01). Second, (ii) by leveraging our subjects’ answers to open-ended questions related to trust in institutions during the outbreak, we will train a natural language processing model to develop a tool deployable on large scale social media data. This would allow us to monitor changes in population levels of trust during pandemics. Third, (iii) with the longitudinal component of our survey, we aim at generating a panel over time to observe how behaviors and trust change as a function of the unfolding of the crisis. At the end of the survey, we plan to ask participants to share with us their Google Takeout data to match the self-reported and actual behaviors by leveraging their location histories (in a fully anonymized way).



Collaborations and dissemination

The team started developing the original concept of this survey back in late 2019. The original idea was to present a new approach to measure interpersonal trust and trust in different institutions. As such, this survey is a follow-up of the Trustlab project, which I implemented together with a OECD-based team in 2016, and took place in France, Italy, South-Korea, United Kingdom, USA, Slovenia and Germany. As our current survey was about to go into the field, the Covid pandemic struck. Since we are located in Lombardy Region in Italy, we got a firsthand look at what was about to happen. One of my colleagues and collaborators at Bocconi University (Alessia Melegaro), an epidemiologist specializing in influenza, was given the responsibility of providing the Covid projections for the Region. As early as mid February 2020, she informed us that Italy was facing a major crisis. At that point, we started to redesign our study to make it specific to the Covid crisis. I will build on these existing collaborations and the proximity to the decision makers in Italy to help inform policy response models across OECD countries. In order to do so, my team has devised a dissemination plan aimed at reaching two types of audiences. First, in order to reach the broader public, we will produce periodic research briefs that we will condense into blog posts published by the team’s website ( I will rely on Bocconi University’s communication team to circulate the content on social media, online and print newspapers. We will also produce articles for media outlets such as ( To engage with policy-makers, we plan to build on pre-existing links with the OECD in order to sensitize this international institution to our findings. We will also interact with the WHO by submitting our findings to the Bulletin of the World Health Organization and to the CDC’s Emerging Infectious Disease journal.



  • Aassve, A, Alfani, G, Gandolfi, F, & Le Moglie, M (2020) Epidemics and Trust: The Case of the Spanish Flu. IGIER Working Paper n.661
  • Aassve, A, Mencarini, L, Chiocchio, F, Gandolfi, F, Gatta, A, & Mattioli, F (2018), “Trustlab Italy: a New Dataset for the study of Trust, Family Demography and Personality”, DONDENA Working Paper n.115
  • Aassve, A, Billari, FC and Pessin, L. (2016) Trust and Fertility Dynamics, Social Forces, Vol 95 (2), pp 663–692
  • Van Bavel, J. J., Baicker, K., Boggio, P. S., Capraro, V., Cichocka, A., Cikara, M.,  Crockett, M. J., Crum, A. J., Douglas, K. M., Druckman, J. N. Drury, J., Dube, O.,  Ellemers, N., Finkel, E. J., Fowler, J. H., Gelfand, M., Han, S., Haslam, S. A., Jetten, J.,  Kitayama, S., Mobbs, D., Napper, L. E., Packer, D. J., Pennycook, G., Peters, E., Petty,  R. E., Rand, D. G., Reicher, S. D., Schnall, S., Shariff, A., Skitka, L. J., Smith, S. S.,  Sunstein, C. R., Tabri, N., Tucker, J. A., van der Linden, S., Van Lange, P. A. M.,  Weeden, K. A., Wohl, M. J. A., Zaki, J., Zion, S. & Willer, R. (in press). Using social and  behavioral science to support Covid-19 pandemic response. Nature Human Behavior
  • Betsch, C. (2020) How behavioral science data helps mitigate the Covid-19 crisis. Nature Human Behavior
  • WHO (2020). Survey tool and guidance. Rapid, simple, flexible behavioral insights on Covid-19. A