Survey Results Suggest the Importance of Clearly Communicating Coronavirus Risk and Behavior — Including as States Re-Open

As news of the coronavirus swept the nation in March 2020, the risks of COVID-19 infection and infection-fatality were still unclear. Perhaps as a result, a new study found respondents showed large disagreements between the risks of COVID-19 infection. Yet, those who perceived a higher risk indicated they were more likely to engage in handwashing and other protective behaviors. And participants seemed more willing to act on their risk perceptions as the pandemic progressed and the risks became more real.

The study was led by Wändi Bruine de Bruin of the USC Schaeffer Center for Health Policy & Economics and the Price School of Public Policy and Daniel Bennett of USC Dornsife and published in the American Journal of Preventive Medicine. Bennett and Bruine de Bruin believe that the results may help public health and policy leaders in developing effective risk communication and information sharing.

“If people think that the risks of infection and associated consequences are low, then they may not feel the need to implement the recommended protective behaviors,” said Bruine de Bruin who co-directs the USC Schaeffer Center’s Behavioral Sciences Program and is a USC Provost Professor of Public Policy, Psychology, and Behavioral Sciences.

Coronavirus Risk Led to Behavioral Changes, Including Handwashing, Social Distancing, and Travel Plans

For the March survey, 6,684 individuals participated, with roughly half completing it before March 13 when the White House declared COVID-19 a national emergency. Half the respondents completed it later in March.

Most but not all respondents reported taking actions to protect themselves. Across the entire sample, 90% reported handwashing, 58% reported avoiding high-risk individuals, 57% reported avoiding crowds, and 37% reported canceling or postponing travel.

The median survey respondent saw a 10% chance of catching the coronavirus. If infected, the median respondent perceived a 5% risk of dying.

But the authors point out that respondents showed large disagreement in perceived infection risk. And this disagreement correlated with differences in reported rates of protective behavior.

For example, reports of handwashing increased from 83% among the quarter of respondents who perceived the lowest risk of COVID-19 infection to 94% among the quarter who perceived the highest risk. Similarly, reports of avoiding public spaces or crowds as a protective measure increased from 45% for those reporting low perceived risk to 67% for those reporting highest risk.

“Every day we assess risks and decide how much to protect ourselves,” said Bennett. “It’s not clear how people make these decisions for a new and unfamiliar disease.” Bennett is an assistant professor (research) of economics at the Center for Economic and Social Research who studies how people make health decisions.

Perceived Risk and Correlations with Protective Behaviors Intensified Later in March

Because the survey was released mid-March, with some individuals responding earlier than others, the researchers were able to examine how risk perceptions and protective behaviors changed as the initial stage of the epidemic unfolded. All told, half the respondents completed the survey after March 13, when the White House issued a national emergency and banned European travel while several states announced school closures and bans of large gatherings.

Late responders showed greater likelihood to engage in protective behaviors:

  • Reports of handwashing increased from 86% among earlier responders to 93% among those who responded after March 13.
  • Avoiding public spaces or crowds increased from 43% to 71%
  • Avoiding high-risk individuals increased from 46% to 71%
  • Canceling or postponing travel increased from 24% to 49%.

Furthermore, association between perceived COVID-19 infection risk and protective behaviors was stronger for later responders.

“Early on in March, the risks of infection were perceived as relatively lower compared to later in March. But even people who saw greater risks were not yet as willing to act upon them,” said Bruine de Bruin. “That’s likely because the mixed messages people were receiving made the risks still seem relatively uncertain, not as imminent, and affecting mostly other countries. As the epidemic progressed, those perceptions changed.”

“In other words, our findings show that risk perceptions and behavior are connected. Moving forward, and when opening the country back up, policymakers need to provide clear and consistent messages about what the risks are and what people can do about them,” explained Bruine de Bruin.

About the survey: The sample of 6,684 U.S. residents participated in the survey from March 10 to March 31. The margin of sampling error is +/-2%. Survey respondents are members of USC Dornsife’s Center for Economic and Social Research’s Understanding America Study (UAS) probability-based internet panel. The survey was funded by USC. Information about the survey and the UAS panel, including methodology, question wording and results is available upon request to and will be posted online when the survey is complete.