Criticism of COVID-19 Models by Politicians May Erode Public Trust in Science
Specifically, Democratic political leaders' critiques of such models was shown to undermine public support for the models' use, while Republican opinions had little impact
Criticisms of COVID-19 models by Democratic elites in May 2020 appeared to undermine public support for the models' use—and trust in science more broadly—according to a series of survey experiments conducted with the participation of more than 6,000 Americans.
However, whether Republican elites criticized or supported the models appeared to have little effect. Sarah Kreps and Douglas Kriner suggest the lack of response to Republican messaging could be due to the party's split messaging on science-backed guidance for this issue. When Democrats criticized COVID-19 models, however, it strongly contradicted the public's expectations.
"The fact is that whether or not political leaders' science communication sways people, they have an ethical obligation to treat the science with care, acknowledging uncertainty while asserting that we are constantly updating with new understandings and data about the virus," says Kreps, the co-author of the study.
Related Article: Computer Model Indicates When COVID-19 Measures Can Be Lifted
Since models are built on abstractions and incomplete data that make them inherently uncertain, and research on the novel coronavirus is still in its infancy, models predicting the virus' spread have sometimes been inaccurate. It has remained unclear how competing communications about uncertainty in COVID-19 models affects public support for and trust in science.
To better understand the effects of science communications in the context of the pandemic, Kreps and Kriner developed five survey experiments and used them to assess shifting public attitudes toward references to COVID-19 models from prominent Democrats and Republicans. The surveys were designed to test responses to both the cue giver (the Democrat or Republican) and to whether his or her statement ignored, acknowledged, highlighted, or weaponized model uncertainty.
Based on their findings, Kreps suggests scientists should avoid emphasizing dire implications associated with epidemiological models while sidestepping uncertainty altogether, since this approach could backfire if projections prove incorrect.
"Instead, they should acknowledge that models are simplifications of reality and our best estimate based on a lot of moving parts," she says. "Politicians can help convey to the public what we know and what we still don't know about the virus, and stress the need to adapt policies in response to new information," Kriner adds.