New research report examines predictors of trust in governmental health communication on Covid-19

David Schieferdecker, Natalie Berger, and I published a new research report in the European Journal of Health Communication (available in open access here). Using representative survey data from Germany our study reveals significant shifts in public trust regarding Covid-19 information provided by the government. Initially, trust levels were high but declined sharply as the pandemic progressed, never fully returning to their early heights. This drop may reflect the fading of a “rally around the flag” effect—where initial crises spur trust in governmental action. As people grew accustomed to the pandemic, trust waned, and frustration with changing policies such as lockdowns and slow vaccine rollouts (see Figure 1). Alarmingly, by the second year of the pandemic, less than 50% of respondents trusted the government’s health information. This has serious implications for the effectiveness of public health campaigns, as trust is critical for public engagement with preventive measures.

Figure 1. Average levels of trust in Covid-19-related information by the federal government in Germany, December 2020 to September 2021. Note: The figure shows the mean level of trust by wave with 95% confidence intervals in brackets. License: CC BY 4.0.

We found that older individuals and those more afraid of contracting Covid-19 tended to trust government information more. However, merely being part of a vulnerable group due to health conditions did not always translate into higher trust levels, pointing to the complexity of risk perception (see Figure 2). The study also highlights how political ideology influenced trust levels. Supporters of the right-wing populist AfD party exhibited the lowest levels of trust, reflecting a broader trend where populist voters, often skeptical of political elites, extend this distrust to health information.

Figure 2. Models of trust in Covid-19-related information by the government. Note: The figure shows coefficient estimates with 95% confidence intervals for two multilevel linear models with random intercepts for participants. The moderation model includes an interaction term between vote choice and fear of infection. Reference categories: Gender = Male, Age = 35-64 years, Education = Low (below pre-university education), Risk group member = No, Vote choice = Established parties. Wave is coded from (1) December 2020 to (4) September 2021; household income is coded from (1) under EUR 500 to (12) EUR 10,000 or more; fear of infection is coded from (1) not afraid at all to (4) very afraid. License: CC BY 4.0.

Finally, fear of infection had a stronger impact on trust in governmental communication among AfD supporters compared to voters of established parties (see Figure 3). While voters of established parties generally trusted government action regardless of personal fear, AfD supporters who feared the virus were more likely to rely on government information. We suggest that AfD supporters with greater fear of the virus may have been more likely to turn to government information, as their typical media sources might not have offered adequate guidance on managing risks.

Figure 3. Adjusted predictions for voters of different parties with different levels of fear
of infection. Note: The figure shows predicted values for trust in information from the government by different levels of fear of infection for voters of different parties, with 95% confidence intervals. License: CC BY 4.0.

We further discuss the implications of these findings and make recommendations at the end of the paper.

How to cite

Joly, P., Schieferdecker, D., & Berger, N. (2024). Trust in Governmental Health Communication on Covid-19: Does Vulnerability Moderate the Effect of Partisanship? European Journal of Health Communication, 5(3), 19-32. https://doi.org/10.47368/ejhc.2024.302

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.