"People of any given race are much more likely to have a crime committed against them by someone of their own race, than by someone of any other race." Is ther

"People of any given race are much more likely to have a crime committed against them by someone of their own race, than by someone of any other race." Is there a correction factor for housing and employment segregation? The analysis seems to rely on a large municipal area with overlapping cultures rather than the usual situation on the ground which is highly segregated for a variety of socioeconomic reasons. So given that, it is extremely likely that, say, home domestic disturbances or workplace violence will be extremely highly correlated with victim and attacker having same race. However whats behaviorally relevant is the totally different statistics of "I'm a X and I see a Y walking down a dark street toward me, how should I react based on X and Y crime statistics?" "you must be 4 times more afraid everytime you see a" That whole clause is a statistical error as the odds of becoming a victim of a X have a lot more to do with the odds of a given X being a criminal, than they do with relative ratios of X on Y violence vs Y on X violence. Outside of really weird situations like active race riots or race based protests or similar lawlessness. There is the meta question of "how do I need to signal to others in public to remain politically correct" vs "how do I actually need to act to survive" and discouragingly that gulf has been widening over past years mostly by changes in required political signalling behavior despite a remarkable drop in actual crime rates. Or at least reported crime rates.