3 ms·

> New York City is literally statistically safer for murder per capita than many of these smaller cities. This little trick is a pet peeve of mine. Per capita

> New York City is literally statistically safer for murder per capita than many of these smaller cities. This little trick is a pet peeve of mine. Per capita crime statistics mean nothing to the people living in a city. It doesn't matter that each person you run into each day is slightly less likely to be a criminal if you run into 100x more people in NYC than in a small suburb. What matters is the question "how likely am I to be the victim of a crime?". The answer is: a lot more likely in NYC than a small city.

quicklime 6y ago
Doesn't "murders per capita" answer that question? It's "murder victims per capita", not "murderers per capita", so it doesn't matter if you run into 100x more people in NYC.
- ponker 6y ago
  And even if it was murderers per capita, it would require the assumption that each murderer is murdering a constant percentage, regardless of location, of the people that he sees.
jcranmer 6y ago
The more I think about it, the less sense your argument makes. Property crimes like burglary and arson are committed against your property, and so the natural denominator is per household size. As long as household size is pretty invariable among cities, then per capita is going to correlate quite highly with household size. How about crimes like murders? Most murders are people killing someone they know. Your social circle is limited more by human capabilities, so it's not going to increase just because you live in a denser city. So again, your likelihood of being murdered will correlate with murders per capita. Another take on the statistics: if your argument is that you are more likely to run into the number of criminals because of a higher density of interactions, then consequently the number of crimes per criminal should also go up. I don't have any hard numbers on these statistics, but I would be extremely shocked if they bore you out here. A final comment is that actual crime statistics is measured by the number of crimes committed per residents living there. There is actually some bias here: some of the highest numbers here are measured in places like Times Square which combines extremely high tourist traffic (and hence high pickpocket rates) with extremely low residential numbers. But if you talk about entire city-level rates, those effects are going to be muted. If you again assume that the number of victims of a single crime is constant, then the number of crimes committed per population, and that the victim of a crime is chosen uniformly at random, then the crime rate per capita is pretty much the inverse of the likelihood of being a victim of a crime.
ponker 6y ago
Your math is deeply wrong. You’re only in more jeopardy in NYC if you manage to become 100 people in NYC.

quicklime 6y ago

Doesn't "murders per capita" answer that question? It's "murder victims per capita", not "murderers per capita", so it doesn't matter if you run into 100x more people in NYC.

ponker 6y ago
And even if it was murderers per capita, it would require the assumption that each murderer is murdering a constant percentage, regardless of location, of the people that he sees.

jcranmer 6y ago

The more I think about it, the less sense your argument makes. Property crimes like burglary and arson are committed against your property, and so the natural denominator is per household size. As long as household size is pretty invariable among cities, then per capita is going to correlate quite highly with household size. How about crimes like murders? Most murders are people killing someone they know. Your social circle is limited more by human capabilities, so it's not going to increase just because you live in a denser city. So again, your likelihood of being murdered will correlate with murders per capita. Another take on the statistics: if your argument is that you are more likely to run into the number of criminals because of a higher density of interactions, then consequently the number of crimes per criminal should also go up. I don't have any hard numbers on these statistics, but I would be extremely shocked if they bore you out here. A final comment is that actual crime statistics is measured by the number of crimes committed per residents living there. There is actually some bias here: some of the highest numbers here are measured in places like Times Square which combines extremely high tourist traffic (and hence high pickpocket rates) with extremely low residential numbers. But if you talk about entire city-level rates, those effects are going to be muted. If you again assume that the number of victims of a single crime is constant, then the number of crimes committed per population, and that the victim of a crime is chosen uniformly at random, then the crime rate per capita is pretty much the inverse of the likelihood of being a victim of a crime.

ponker 6y ago

Your math is deeply wrong. You’re only in more jeopardy in NYC if you manage to become 100 people in NYC.