Figures never lie, but…they do tell stories

If you ever thought communicating risks is a fact-based, “neutral” endeavor, current coverage of the coronavirus pandemic should convince you otherwise. Which numbers you choose to present and the context you provide to help your audience interpret those figures greatly influences their conclusions.

Here are three lessons drawn from recent media coverage of Covid-19.

1. Present the figures that are most useful for your audience.

The most closely tracked index of the severity of the pandemic is the total number of cases. As of April 13, 2020, the US has the highest number of cases of Covid-19 in the world.

This frightens us, as it should. I’ve read about what happened in Italy and this leads me to expect worse in the US. But this statement is misleading because it doesn’t account for the relative size of the country. Cases per million people is a more useful gauge of the relative burden of the pandemic in different countries. By that measure, the US fares a bit better than most European countries. And China looks pretty good—its incidence per million people is only 57, compared to 1,172 in the US, 2,968 in Switzerland, and 3,625 in Spain.

2. Provide comparison points to help people interpret figures.

How contagious is the coronavirus? The current best guess is that every patient transmits the novel coronavirus to 1.5-3.5 people. This figure—the reproduction rate—makes perfect sense to epidemiologists but needs to be interpreted for laypeople. One way to do that is to present it in contexts of risks they already understand, e.g. to the best of our knowledge, the coronavirus is probably a little more contagious than the flu (where each patient typically infects 1.3 people) but nowhere near as contagious as measles (where each patient typically infects 12-18 people). 

3. When numbers are unreliable, it’s better not to use them at all.

Another statistic that appeared everywhere recently was that an estimated 20-60 percent of the US population would get Covid-19.  Presenting such a broad range is useless and perhaps counter-productive, because it essentially allows people to pick the figure that is most in line with their own biases. When the numbers are so elastic, it is better to talk about scenarios (best, worst, most likely). 

In our own communications, data and statistics are often useful for helping the public to understand risks to their land. Just make sure you’re using the right numbers to tell the right story.