Are sanctuary cities more violent?

Statistical Ideas – by Salil Mahta

With the battle growing in sanctuary cities to deviate from President Trump’s strategy on immigration law, it is worth seeing another topic that we have been following closely: violent crime rate across major U.S. cities.  This of course comes as President Trump menacingly engages with Chicago Mayor Rahm Emmanuel, in a twitter exchange that reiterates the use of the term “carnage” (the President’s quotes, not ours) to suggest the warfare-style chaos that is occurring there.  

If Chicago doesn't fix the horrible "carnage" going on, 228 shootings in 2017 with 42 killings (up 24% from 2016), I will send in the Feds!

— Donald J. Trump (@realDonaldTrump) January 25, 2017

By logic (no formal definition exists per ICE) sanctuary cities are generally liberal (Democratic) cities or regions that oppose the current conservative policies of oppressing undocumented immigrants.  The 14 most-populated U.S. cities represent 30 million Americans (or 9% of the 319 million U.S. population).  Half of these large cities happen to be in states bordering Mexico.  And of these 14 cities, 9 are sanctuary cities (64%) and 5 are non-sanctuary cities (36%).  As we prove below, one is significantly safer from violent crimes in sanctuary cities, but for a couple notable exceptions (i.e., Chicago, Philadelphia).  Below we’ll distinguish between these two types of cities:

• Nearly 2 thousand murders occur annually in these sanctuary cities (>70%), while less than a thousand murders occur in the non-sanctuary cities (<30%).

• While each taken life is too many, mortality statistics drive these large death numbers into probability context.  So with the much larger population from the sanctuary cities (collectively or on average), the homicide rate (per 100k) there is “just” 5, while it is 9 in non-sanctuary cities!

• It is true that the murder rates have come in most cities across the U.S., but again these rates are unacceptably high, particularly as some cities are many times more ferocious versus their peers!
• The blended murder rate from the 14 large cities meanwhile is in-between at 8, and this is also just less than twice the national average is 4.

Now we can see a chart of these 14 cities on this map below, where the size of the circular-marker is related to the population of the city, and the text color of the murder rate is blue for sanctuary cities; red for non-sanctuary cities.  The blue on the map regions represents areas who mostly voted for Hillary Clinton (as noted above this was mostly limited to mega-cities), while red indicates the rest of the country where Donald Trump completely dominated the popular vote.

Using a mathematical practice similar to boot-strapping, we show further below, the population-weighted murder rate distribution for the 14 cities.  Also 4 of the 9 (44%) sanctuary cities have had an above-average murder rate among the large cities, while 2 of the 5 (40%) of the non-sanctuary cities had an above-average murder rate.  Though this difference -skewing towards sanctuary cities- is not statistically significant (less than ½ standard deviation, or σ).

Meanwhile the murder rate difference of 4 (9 v 5), between non-sanctuary cities and sanctuary cities, is highly statistically significant, given the population in millions discussed early in this article.  But the chance any given non-sanctuary city is more murderous versus a sanctuary city is not statistically significant (less than 1 σ). Predominantly with such benign jumbo-metropolises, such as New York with their “low” murder rate of 4 (despite sky-high homelessness), versus Houston (the county’s largest non-sanctuary city) with their murder rate of 11.

The bottom line is there is a minor bias towards more violence in non-sanctuary cities, areas generally aligned to conservative policies and gun-friendly.  This is where Americans will typically have a higher probability of being slain (expressly young Black males in the inner-cities who are gunned down by others in the same community, as opposed to the police – many directors of which nationally follow this site). Though the large cities are not easily separable into such mass generalizations, this is also why it happens to be easy for President Donald Trump to censure the worst areas of the country.  Since there we have heterogeneously diverse characteristics of violence, from our large sanctuary cities.

http://statisticalideas.blogspot.com/2017/01/are-sanctuary-cities-more-violent.html