COVID-19 and the humble pie chart

Over the summer, as the Delta variant of COVID-19 began to spread, there was plenty of conversation about how certain states were contributing outsized numbers of new COVID cases to the U.S. total. There are different ways to examine this phenomenon, such as looking at per-capita cases or comparing data to the 14 days prior. 

With the right settings, however, the often-maligned pie chart can give you a good snapshot of this phenomenon. The pie charts below are set to show the top five states by new cases in the month. All other states are listed in the red “other” bucket. 

So, at a quick glance, the smaller the red section, the more concentrated the new cases are in the top five states. You can see that back in June 2020, new cases were very concentrated, with the top five representing almost 50% of all new cases in the U.S. The top five were also the larger states (California, Texas, Florida, etc.), which makes sense since we are not normalizing for population.   

Over the course of 2020 and early 2021, the red “other” stayed fairly large (around 60%), meaning new cases were less concentrated. When we get to the Delta surge in July 2021, however, the concentration is quite high again (top five breaking 50% of all new cases). While the large states are still there, we also see Missouri and Louisiana enter into the top five. In September 2021, we began to see states like Tennessee and Georgia as well.   

In addition to the quick overview, the interactivity of Domo cards lets me hover on the “other” section and quickly see a list of the top states outside of the top five. This exercise made me reconsider if the humble pie chart can be a powerful data viz after all.