3 ms·

Sure, if you compare to the tail end of the crack wars. Compare to 1950 or 1900 or 1980 or... What does this prove exactly? The perception of "safety" is als

Sure, if you compare to the tail end of the crack wars. Compare to 1950 or 1900 or 1980 or... What does this prove exactly? The perception of "safety" is also not only a function of "crimes per capita". "Blocks safe for a woman to walk through alone at night" or "total victimization rate" make just about as much sense. "Obviousness of safe / unsafe areas" is also a good one - if you have to have local knowledge of what is a "good neighbourhood" or not in order to avoid being victimized, obviously non-locals will develop a perception of the area as unsafe. Wow, from the downvotes people seem really invested in particular choices of time horizon & measures of safety. Who knew?

seanflyon 12y ago
Your safety is a measure of how likely you are to be attacked, which is fundamentally a per capita measurement. If there are twice as many people and 10% more attacks, that means that the average person is less likely to be attacked.
- fiatmoney 12y ago
  No. Crimes per capita can obviously be concentrated in a smaller segment of the population (for instance, those that live in high crime areas). Half the population twice as frequently, or vice versa, looks the same in aggregate but not in perception. And geographic concerns do matter. "I must live in this expensive area to avoid being mugged" drives a perception of unsafety even if you are not being mugged.
anigbrowl 12y ago
You're partly right. The 1950s were particularly safe and crime is somewhat higher now than it was then (though not uniformly; you are safer in Houston today than in 1950, surprisingly). On the other hand, it was a bit higher at the beginning of the last century, which is not too surprising as the solve rate was much lower and so were sentences (an average of 10 years for murder in 1910, whereas by 2003 it's 85). A good reference on this subject is The collapse of American Criminal Justice by William Stuntz.

seanflyon 12y ago

Your safety is a measure of how likely you are to be attacked, which is fundamentally a per capita measurement. If there are twice as many people and 10% more attacks, that means that the average person is less likely to be attacked.

fiatmoney 12y ago
No. Crimes per capita can obviously be concentrated in a smaller segment of the population (for instance, those that live in high crime areas). Half the population twice as frequently, or vice versa, looks the same in aggregate but not in perception. And geographic concerns do matter. "I must live in this expensive area to avoid being mugged" drives a perception of unsafety even if you are not being mugged.

anigbrowl 12y ago

You're partly right. The 1950s were particularly safe and crime is somewhat higher now than it was then (though not uniformly; you are safer in Houston today than in 1950, surprisingly). On the other hand, it was a bit higher at the beginning of the last century, which is not too surprising as the solve rate was much lower and so were sentences (an average of 10 years for murder in 1910, whereas by 2003 it's 85). A good reference on this subject is The collapse of American Criminal Justice by William Stuntz.