In my last post I looked at state-level data, which showed a significant correlation between population density in each state and deaths per million people. In this post, I look at the county-level data in four states. I did not normalize deaths per county for population, which could make a difference, but just the number of deaths per county makes for an interesting comparison. See the graph below for a comparison of death by county from COVID for four states. The x-axis is population density (people per square mile) and the y-axis is total number of deaths per county due to the virus.
The R^2 values for each trend line provide insight into how much of the variation in deaths per county is explained by differences in population density. This number is lowest for Virginia (just 7%) but pretty high for TN (approximately 61%).
I do not have many explanations for differences at this point, but will discuss variations in each state that seem to stand out. One thing to keep in mind is that some of the variation could be influenced by variations in reporting and testing availability. The lower the level you go, the more noise and uncertainty there is in the data.
- Tennessee: Nashville is in Davidson County and Memphis is in Shelby County. In Sumner County (the county to the north and east of Nashville) there was an outbreak at a nursing home that accounted for several deaths.
2. Florida: Pinellas County is the Tampa/St. Petersburg area.
3. In Virginia, both Arlington and Fairfax are in northern Virginia, near D.C.
4. In Alabama, Madison County has reported few deaths, but Mobile has had many. Mobile has a Mardi Gras celebration that could have aided in the spread of COVID.
I would be interested in people’s comments on these variations, and any insights you might have. Please comment.