Quote (thundercock @ Aug 25 2020 03:41pm)
Ok, let's go with the stop and frisk example. Suppose that a white person has a 1 in 1000 chance of carrying an unlicensed firearm in public. Now suppose a black person has a 1 in 500 chance. If you frisk the population RANDOMLY, you will see that the black people are packing heat twice as much as a white person. Great! Now, what do you think will happen to the data if you sample blacks at twice the rate as whites?
We saw the result of this with stop and frisk. White stops were the most successful, likely because most whites were excluded from the population by default. If you were stopped as a white, you fit other criteria that were far more indicative of an underlying crime.
Quote (Thor123422 @ Aug 25 2020 03:14pm)
Take two neighborhoods with equal crime rates, now at twice as many police to the second neighborhood. You've just used policing to double the recorded crime rate in a neighborhood with the crime rate being equal
Murder rates allow for some clarity here, as it's easier to identify a homicide than most other forms of crime. Let's assume that 100% of unsolved murders are committed by non-blacks (undoubtedly a poor assumption) and crunch the numbers to see whether the crime rate is higher in black(er) communities, or whether that narrative is just a function of over-policing.