Risk Communication is a Risky Business: The Importance of Communicating Timelines during Periods of Suspended Animation
I conduct research in the area of crisis informatics. Public health pandemics are different enough from my primary domain for research — the informatics of natural hazards — that I’ve been getting a sense of the situation by putting my own household in order while figuring how I might formally research this new crisis that has descended on us all. Why do I make this distinction between public health pandemics and other kinds of crises? It is because we know that online information behaviors are different across different kinds of crisis events. People behave quite differently online in the face of a hurricane versus a school shooter, for example, because the problems to be solved by those affected are different. Pandemics are going to be different yet again in terms of online information behaviors, and so there is a lot for me, even as a 15 year veteran of crisis informatics research, to learn. What we know about the digital experience in other kinds of crises might not fully generalize to a pandemic.
One of the subareas for which I conduct research is risk communication, or how official risk messages are issued to the public, and how the public responds to them. This is an already deep literature and, most recently, my colleagues and I have been adding to that scientific knowledge by examining how graphical risk communication diffuses quantitatively and qualitatively across social media during hurricane events. This pandemic is teaching us yet more, and it is on this topic that I would like to offer my first reflections on this crisis.
Of the many things we know about risk communication, I’ll share a few points. First, we know that risk communication is very hard to do. This is because to convey risk means to convey uncertainty, and to convey uncertainty means to engage in probabilities. Probabilities are hard to explain succinctly to a wide audience who are being reached by broadcast media of various kinds, all of whom might have a different relationship to the risk. Another important point is that risk communication needs to avoid foreclosing uncertainty; it also needs to avoid keeping uncertain that which can be closed. This means that messaging cannot be unequivocal unless uncertainty has lifted. But succinctness, which is also valued in risk communication, favors unequivocal statements. This is why I like to say that risk communication is a risky business.
In the case of the COVID-19 threat, social distancing and sequestering — which is not far from the sheltering-in-place practice that is employed in some natural hazards — is an activity we engage in even though we cannot see the hazard first-hand, and that is what makes us more uncertain. The end point for sheltering-in-place for natural hazards might not always be entirely clear — but we are pretty sure that we’ll know it when we see it. For this pandemic, the threat is invisible, and so there is no way to then know on our own when it is “safe to come out.”
Social distancing is also upsetting because it breaks normal routines, even under the best of conditions (eg., when one’s livelihood is not under direct threat). I am seeing my children, spouse, neighbors, students, and myself feel anxious about having to make sudden changes in our plans. But the worst of it is that we don’t know when it will end.
At this stage of the pandemic, no one does. However, one of the problems with risk communication as I see it in this case — and in the US at least — is insufficient externalization of assumptions about how instructions map to timelines. For example, rather than just closing schools and universities as has happened rather suddenly this week, we could say with a bit more elaboration:
“Our university is moving to all on-line remote teaching for the rest of the semester. The assumption we are working with at the time of this writing suggests that 5 weeks is sufficient. However, with only 2 more weeks of the semester beyond that, we want to support students and instructors, and limit switching back to in-person settings once this new commitment to teaching pedagogy is made. We want the transition to summer session to be as smooth as possible, given this disruption today.”
And for the more general case:
“We are now working with the assumption that X weeks of social distancing could significantly reduce the rate of spread. However, we expect that in one week of regionally-managed social distancing, we will have more empirical knowledge about the rate of contagion, and will update this estimate then.”
Risk communication to socially distance without an end in sight is individually distressing and societally devastating. Some temporal bounds, even if tenuous, with clear explanation for why those bounds make sense at the moment of communication, could be deeply reassuring, and improve mental health, market health, and compliance with social distancing practices. Suspended animation might be more readily recast as being efficient, less scary, and less risky to other parts of our lives such that we might be less reticent to participate in this critical collective effort.
Leysia Palen has been conducting crisis informatics research since 2004 as a Professor of Information Science and Computer Science at the University of Colorado Boulder. She and her research can be found at https://cmci.colorado.edu/~palen/.
This is the first article in a series on research and reflection during the novel coronavirus pandemic.
