(((tedfrank)))

(((tedfrank))) @tedfrank

@charlesmurray Isn’t this necessarily going to be correlative? If crime goes down, the ratio will go up and vice versa, simply because prisoners will be a lagging variable. I agree about causation, but I’m not convinced this is a good way to demonstrate it because of the confounding math. — PolitiTweet.org

(((tedfrank))) @tedfrank

@charlesmurray I agree, but the ratio will be misleading when crime is going down. What does the data look like when it's prisoners per 100,000 adults? That should show the same effect without double-counting. — PolitiTweet.org