What’s up with New York? New York State accounts for 6 percent of the US population but 30 percent of confirmed covid diagnoses and 40 percent of covid-related deaths. They got hit early and hard, but the other states are catching up: last week, 21% of newly diagnosed cases and 21% of new deaths came from NY.
What’s up with testing?
Hypothesis 1: State A and State B have equal populations. State A tests everyone with a fever, whereas State B tests everyone with a fever and a dry cough. All else equal, State A would test more people overall, while State B would have a higher hit rate for the smaller number of tests it conducts.
Hypothesis 2: Again, assume two states with equal populations. State A’s hospitals report a surge in patients testing positive for covid, whereas State B experiences no such surge. All else equal, State A would test more people, and have a higher hit rate, than would State B.
Which hypothesis is better supported by the data? Turns out it’s Hypothesis 2. Using Worldometer’s data, I correlated the states’ testing rate per million of population with their case rate per million. The correlation coefficient is 0.65 — a pretty strong positive relationship between the two variables. I.e., the more a state tests, the higher the rate of test-positive rate. That’s not what I expected.