> You are comparing two different numbers, homicides with percent of prison population for all crimes? What do you find so confusing? I am using homicide as a

> You are comparing two different numbers, homicides with percent of prison population for all crimes? What do you find so confusing? I am using homicide as a proxy for all violent crime. Homicide because it is much harder to bias through over-policing, which is the usual rebuttal to using any kind of police data. You can also look at the victimization survey, which gives numbers for other crimes, without involving the police. It gives similar ratios of offense rates. > The opening paragraph on that first link you shared seems to directly contradict what you are saying It doesn't contradict me at all. It's comparing the population size to the rate of arrest. I am comparing the rate of imprisonment to the rate of violent crime committed. Of course if you only compare population size, without accounting for behavior, the imprisonment will be "disproportionate", just as people who study linguistics are "disproportionately" represented among translators.