[iudqnolq]

You're right, but not for the reason I think you intend to imply. Many Americans think Democratic cities are high-crime and Republican suburbs and rural areas are safe. If you look at the data this is completely wrong.

So places actually hurt by crime tend to support the police. Because the medium place actually hurt by crime is somewhere like suburban Louisiana.

[xyzzyz]

> Many Americans think Democratic cities are high-crime and Republican suburbs and rural areas are safe. If you look at the data this is completely wrong.

It is not "completely wrong". It is only somewhat wrong if taken literally as stated, because due to extremely strong political correctness norms, the crucial variable is typically left unmentioned in public. However, most Americans who have the above quoted belief, also understand the impact of that crucial variable. That crucial variable is, of course, how many black Americans live in the area. Once you control for that, you will indeed observe that Democratic cities have higher crime than Republican suburbs. This is why your example was Louisiana, which has very high black population, and so the suburbs have much higher percentage of black residents than other places around the country.

The reason why people think Democratic-ran cities have high crime compared to Republican suburbs, despite the country-level correlation analysis makes nonclear or even wrong, is that they do not do country wide analysis, but instead they look at a particular city close to them, and the actual suburbs and rural areas around it, and if you do that, the pattern actually is obvious, including in Louisiana. The fact that it breaks down on the national level is just Simpson's paradox.

[iudqnolq]

You're wrong.

I change my example to New Mexico. 80% White. Very high crime in both metro and nonmetro areas, with nonmetro higher.

I'm not sure exactly what data analysis you're describing. You might find this breakdown of death by assault by state and metro/nonmetro. This dataset doesn't have perpetrator race, so I restricted victim race to white. In the US perpetrator and victim race correlate strongly.

https://wonder.cdc.gov/controller/saved/D140/D315F773

[rayiner]

Within any given state, Republican suburbs and rural areas are almost always safer than democrat run cities. For example, in Louisiana, the homicide rate in New Orleans is 41 per 100,000. In neighboring St. Charles Parish it’s 1/10 that. The same thing is true for say Massachusetts. The homicide rate in NYC is 5.7 per 100,000. The homicide rate in Putnam County is 0.4 per 100,000.

The statistic you’re referring to misleadingly disregards regional patterns in violent crime. Southerners are more violent than folks in the northeast, and midwesterners are somewhere in between. But within each region the bill of violent crime is still concentrated in the, usually blue, cities in those states.

[tptacek]

All you're observing here is that metro areas get more conservative as you move from dense downtowns to almost-rural exurbia, and crime tends to track density.