In our earlier analysis we compared unemployment, with Republican states doing better, and COVID deaths, with Democratic states doing better. In this next part of our multi-part series, we compare the combination of the two.
During the first part of our analysis, we collected the monthly unemployment numbers from the BLS for every state from the beginning of the pandemic. We then collected the data for the governor of each state and the party of each governor. We took the average and population-weighted average of both Democrat and Republican states and plotted them on a time series.
For the second part of our analysis, we collected the monthly death per million numbers from each state’s COVID website and the CDC for every state from the beginning of the pandemic. We then collected the data for the governor of each state and the party of each governor. We took the average death per million of both Democrat and Republican states and plotted them on a time series.
Today we combine both into a single score. If the pandemic has taught us anything it is the different people respond differently to the threat of a transmissible virus. We wanted to create a grading system that would help to compare the differences in strategies employed by different states. To do this we took our monthly results from both studies and multiplied them by each other. This would help answer if the low illness rates were worth increased unemployment or not between red and blue states. Both deaths per million and low unemployment have values with lower values being more preferred. This allows us to easily multiply the two together in order the yield a single score. This also more closely resembles the multifactor nature of policy-making much better than any single variable. With a lower number being preferred individually this means that the product of the two or a lower score is better.
First, we plotted out every state and color-coded them based on their governor’s party. This graphic reveals to us less than any previous graphic of this nature. There are some high blue and high red points.
Next, we took the mean and standard deviation of the scores of every state categorized by party. Here we can see that red states appeared to be much less variability than blue states in their performance. Republican states rarely exceeded blue states by more than 3 or 4 points, however, blue states drastically spiked over their red state scores, particularly early in the pandemic, by ~15 points. Blue states also performed slightly worse in the Winter of 2020 with red states performing worse in the Fall of 2021.
We then took the average score of all months between red and blue states and found that red states had a better average score; 5.7 vs blue states 7.8. Red states, on a monthly average, performed 25% better.
Finally, we decomposed these scores into their seasons. Here we found the largest variance to be in Spring 2020. Our previous analysis confirms this as it revealed that red states did not experience the virus until later into the year in 2020. After Spring 2020 blue states also saw a large variance in Summer 2020, again, the previous analysis confirms this as we know red states did not experience the virus until later into the year in 2020. Lastly, blue states also underperformed in Winter 2020.
We found that combining multiple factors, unemployment, and COVID illness, into a single score that Republican states outperformed Democratic states over the entire pandemic by ~25%. Much of this comes from relative underperformance in Spring and Summer 2020 was not offset by Republican underperformance in Fall 2021.