Because of the wide variations in population density within the US, one really needs to discuss "Homicides Per Capita".
What is striking about the US is the wide variation in "homicides per Capita" by location. Here is some information I developed last June in the course of a discussion with Clayton and Ralph Luker on HNN. ---------------------------------- The FBI's database "Uniform Crime Reports" is here http://www.fbi.gov/ucr/01cius.htm . I pulled the table at the bottom "Offenses known to Law Enforcement -By Cities 10,000 and over,2001 " into a spreadsheet and added a computed column to computer the ratio "Murders/voluntary manslaughter per 100000".
What's surprising is the many places have ZERO deaths per 100000 people in 2001, and many more only had 1-5 deaths per 100000. However, there are hot spots of homicides. Gary Indiana and Fairfield Alabama rated highest at 79 deaths per 100000 and 97 per 100000 respectively. (That's not low if continued over time. Gary's rate gives you 790 deaths over 10 years time--meaning almost 1 person per 120. Imagine your high school class of 240 student losing 2 people within 10 years after graduation.)
Some other high areas: Alabama: Birmingham (30 /100000), Fairfield (97), Washington, DC (41), Compton-California (48), Opa Locka- Florida (65), New Orleans(44), Detroit (41), St Louis (42), New York (43), Youngstown-Ohio (42), West Columbia-South Carolina (45),
Mr Luker's Atlanta,Georgia had a death rate of 34 per 100000.
By contrast, most states which did not have cities with deaths rates reaching 30 per 100000. Some states, like North Dakota , mostly have cities with death rates lower than Europes, i.e, in the 0 to 1 range.
The death rate in Clayton Cramer's Boise Idaho is 1.05 per 100000.
Most of the local townships around my home had Zero deaths per 100000 but Norristown (a small, poor, largely black town located about 5 miles away has roughly 15 per 100000) and Philadelphia (located about 17 miles away) has a death rate of about 20 per 100000.
Maybe Clayton and I should lend Mr Luker some of our firearms.
I also checked the other file " Table 8a, Offenses Known to Law Enforcment by City with populations less than 10,000". What is striking is that the vast majority of these 4660 small towns reported NO deaths by murder/voluntary manslaughter in 2001.
Here again, there are some towns with high rates but in most case we are talking about 2-3 deaths which give a high ratio because of the small population. I ignored cases of 1 death and checked on those towns with 2 or more deaths. Some examples: Clayton, Alabama (135 /100000), Irwindale Calif (135), Edgewood Florida (2 killed in population of 1950, giving ratio of 102 per 100000), Woodland Township (169- 2 out of 1180), Milan New Mexico 157 (3 out of 1901) , Selma NorthCarolina 67 (4 out of 6015), Burns Oregon 96 (3 out of 3100), Jamestown Tennessee 107 (2 out of 1855), Pound Virginia 181 (2 out of 1106 --small town in Appalachians on Kentucky border), South Boston Va 58 (5 out of 8622)
All of the above suggests that targeted job, education, and law enforcement assistance to the hot spots could reduce deaths greatly and that areas with high gun possession don't necessarily have high or even moderate homicide rates. ----------- PS Note that Fairfield Alabama is on the outskirts of Birmingham Alabama. The declining US Steel has been the major employer in both Birmingham and Gary Indiana for a long time.
Birmingham also has a very high murder rate, as noted in my original post above, but the high number of murders take place over a much larger population. Fairfield's population is about 12,000.
