While I agree that the # of tests has to be taken into account, I think your analogy is a little off. Unlike the cars, the people getting tested is not a random sample. At least some of the people have good reason to get tested (i.e. they have symptoms and had close contact with a Covid + person). In this case, an increase of tests is in…
While I agree that the # of tests has to be taken into account, I think your analogy is a little off. Unlike the cars, the people getting tested is not a random sample. At least some of the people have good reason to get tested (i.e. they have symptoms and had close contact with a Covid + person). In this case, an increase of tests is indicative of an increase in cases. However, there are currently tons of tests being taken for no good reason (college students, travelers, etc.) and we are therefore catching lots of asymptomatic +'s , which we would not have capture a year ago. So, I think the "normalizing" you have performed is likely too aggressive, but the general point stands nonetheless.
While I agree that the # of tests has to be taken into account, I think your analogy is a little off. Unlike the cars, the people getting tested is not a random sample. At least some of the people have good reason to get tested (i.e. they have symptoms and had close contact with a Covid + person). In this case, an increase of tests is indicative of an increase in cases. However, there are currently tons of tests being taken for no good reason (college students, travelers, etc.) and we are therefore catching lots of asymptomatic +'s , which we would not have capture a year ago. So, I think the "normalizing" you have performed is likely too aggressive, but the general point stands nonetheless.