IFG Inquiry: How much do data donors trust the German government in handling the pandemic?

Dear donors,

For about half a year now, you have not only been able to donate your vitals and sleep data, but also to respond to various regular surveys about your personal experience and behaviour during the pandemic. For example, each week we ask how much our donors trust the federal government to properly handle the pandemic. The responses are measured on a five-point scale ranging from 1 (not at all) to 5 (completely) and a total of 29,530 donors already submitted their responses. Since we recently received a request about this question according to the freedom of information (IFG) act we would like to share some information on that topic with you today.

Trust varies with case numbers

The following figure shows the mean daily responses over time. It can be seen that the average trust our donors have in the federal government has overall been moderate, i.e. between mostly and partially, with the main exception of November and December 2021. Average trust declined steadily with the increase of new infections in November, followed by a resurgence of trust in December when case numbers fell again. However, when the number of daily new infections increased again in January 2022 due to the Omicron variant, average trust in the federal government declined only slightly and levelled out with the reduction in case numbers in February 2022. The official change of government on December 8 2021, however, had little impact on the trust ratings.

Figure 1: Trust in the federal government over time. The average value of trust of our donors is shown in red. The black line shows the number of new infections averaged over seven days, and the gray dashed line indicates the official change of the German government on December 8, 2021.

The changes can also be displayed as a bar chart. For example we can compare the responses for the last week in October 2021, November 2021 and February 2022, respectively: while 60% of respondents indicated to trust the federal government mostly or completely in the last week of October 2021, this figure dropped to 25% in the last week of November 2021. In the last week of February, their share rose again: 44% of the data donors mostly or completely trusted the federal government to properly deal with the coronavirus.

**Figure 2: Trust in the federal government in October and November 2021 as well as February 2022.** Percentages are rounded to whole numbers.
Figure 2: Trust in the federal government in October and November 2021 as well as February 2022. Percentages are rounded to whole numbers.

Conclusion and Summary

Further analysis of the data showed that trust was slightly higher on average among male and elderly donors, however differences were small. In addition, participants with children under the age of 18 were somewhat less likely to trust the federal government than other donors. This might, in part, be caused by the particular burden placed on families over the course of the pandemic.

Finally, it should be noted that the trust of data donors is not representative of the trust of the whole German population. For example, results from the COSMO study show that the trust of the general population is slightly lower than what we found in our donor population. Nevertheless, the donated data is of high value as the frequency of the surveys allows short-term changes to be detected and potentially be linked to other variables, such as changes in new infections, certain policies, or even personal burden.

As always, we would like to thank all our donors for their active participation in the data donation project as without your ongoing and regular support analyses such as this one would not be possible at all. In our next blog post, we will discuss a new relationship between vital and survey data and show how your donated data can be used to estimate COVID-19 case numbers based on machine learning.

Marc Wiedermann
Marc Wiedermann
PostDoc / Data Scientist

Researcher and Data Scientist with strong interests in time series and network analysis, predictive models and low-dimensional dynamical systems for the spread of human behavior.

Philipp Sprengholz
University of Bamberg
Dirk Brockmann
Dirk Brockmann

Head of Research on Complex Systems Group