With Turkana and Julie Gulden's diaries on the recommended list I believe today is a great day to present some analysis I have done.
The few here that know me know I love playing with data from multiple sources. I found the Census data for 2008 with the estimates of populations by Hispanic and Mexican heritage for 1100 cities in the United States.
I also found the FBI's 2007 crime statistics (89Mb download) for 8659 cities within the country.
I was able to match 828 communities where complete data was available in both databases. The communities with a population under 15000 were either not listed in the census data or had incomplete census data.
There were a few anomalies. The most glaring was Chicago Illinois which reported everything except violent crimes. I did my best to eliminate those incongruities.
Here are the charts I developed from 829 of the 2270 communities in the country with populations over 15000 looking at the FBI's 2007 "violent crime" and "murder and nonnegligent manslaughter" statistics sorted by The Census Bureau's 2008 overall population estimate, the estimate of overall population of Hispanic decent, the estimate of overall population of Mexican decent, the estimate of of non-citizen population, and the estimate of citizens born in the country.
The blue scatter plots are the 828 data points I had available. The yellow line is overall mean value. And the red line is an undifferentiated trend line.
I didn't include any values since I was emphasizing the trend. Nationwide the increase in crime rates due to population increase is actually reduced by the number of Hispanics or non-citizens and populations of Mexican decent heavily reduce the rise in crime rates due to population.
The blue dots throughout the graphs indicate the crime rate on the y axis.
The x axis represents increasing populations.
The straight yellow line is the overall mean average.
The straight red lines represent the rate of increase in crime rate due to increasing population.
The greater the slope of the red line the greater the crime rate increases based on the observed population.
Violent crime rises as expected when sorted by overall population.
When the crime rate is sorted by overall Hispanic population there is a measurably lower rate of increase.
When sorted by population of persons of Mexican decent there is a much lower rate of increase in crime rates.
When sorted by population of non citizens there is also a measurably lower rate of increase in crime rates compared to the general population.
When sorted by population of U.S born citizens there is a crime rate increase measurably higher than the overall population, 30% faster than when considering Hispanic population, and nine times the rate of increase when compared to population of Mexican decent.
When you look at the murder and nonnegligent manslaughter rate by increasing population you see an increase as expected.
When you isolate out murder and nonnegligent manslaughter rate by increasing population of Hispanics you see a lower rate of increase.
When you isolate out murder and nonnegligent manslaughter rate by increasing population of persons of Mexican decent you see a significantly lower rate of increase.
The murder and nonnegligent manslaughter rate increases at when isolated for increases in non citizen population is actually 50% of the rate of increase for the general population.
The murder and nonnegligent manslaughter rate when sorted by population of U.S born citizens shows a lower increase then the overall population, bit it is still 70% faster than that for Hispanic populations, and almost seven times the rate of increase from populations of Mexican decent.