I hear an argument on a pretty consistent basis that goes a little something like this:
Republican: "Democrat run cities have crime rates that are 10 times higher than Republican run cities. If we put a Democrat in charge every city in America is going to become Chicago!"
Democrat: "Republicans don't govern anywhere where people actually live! You can't say that you're better on crime when you aren't even in the game."
Out of sheer curiosity I wanted to try to determine who was right in the argument above. I spent a few nights and weekends personally compiling data to see what I could figure out.
THE DATA COLLECTION
I first pulled a list of the top 250 cities in America by total population. I also pulled in their population density (people per sq mi).
I stopped at 250 because if you get any further down the list than that you start to run into places where the "Mayor" is a part time job that is little more than a nice title as it relates to crime policy. There are exceptions to that rule, there are some small town mayors that are involved in a substantive way, but not most, and continuing past 250 hampers the data with a lot of dead weight more than it helps reveal the overall trends we're trying to investigate.
Cities that made the cut towards the bottom are places like Palm Bay FL, Clearwater FL, Provo UT, Beaumont TX, Independence MO. These cities have around 100k-120k people in them, which is generally where the list ends. Any smaller than that and we're talking population totals that can fit inside a college football stadium (that's actually true).
I then spent longer than I wanted finding the mayor for each of the 250 cities, and their party affiliation. This is more difficult than it sounds. As you work down the list you run into plenty of races where Mayors run unopposed, and because of that they don't declare for a particular party. You also run into a significant number of races where candidates aren't allowed to declare a party by local law, where they have to remain non-partisan on paper.
I did my best to research, I'm not kidding, every single instance when that occurred to try to logically determine which party a candidate likely belonged to based on endorsements, affiliations, even sometimes speech transcripts. Where there wasn't a clear sign of partisanship I left them in the "Independent" category with others who explicitly categorize themselves as "Independent."
So now I had a list of the top 250 American cities with their Mayors and associated party affiliations.
My next step was to look up, one by one, each city's crime rate in the FBI "National Incident-Based Reporting System" (NIBRS). Here you can find the crime rate for every city in America (who has reported their data on time), split between violent crimes per 100k people & property crimes per 100k people.
Finally I wanted to pull in information, again city by city using Census data, on Poverty Rate & Diversity. It's "common knowledge" that these are things that effect crime rate, and I wanted to see if that actually holds true. I went city by city through the most recent Census data and pulled in each city's poverty rate, and % of the population belonging to the largest 3 ethnic groups in America; White, Black, and Hispanic/White Hispanic.
Side note here there is a question as to where the Census categorization "White Hispanic" should fall when discussing diversity. For this analysis I erred on the side of categorizing Hispanic/White Hispanic together, but it's worth noting that in many regions "White Hispanic" is an on-paper designation where in everyday life the average person wouldn't recognize them as anything but White. The result of this is that it over-represents some cities as being more diverse than they really are in practice.
THE RESULTS
This is where we get into the "you can make data say anything you want" territory. I'm going to show how Republicans would spin the data, then how Democrats would spin the data, and finally I'll explain how I see things.
Republican Talking Points:
- Democrat cities have +167% more violent crime, and +81% more property crime than Republican cities.
- Democrat cities have around a +5% higher Poverty Rate.
- Democrat policies lead to higher poverty and much higher crime rates, numbers don't lie.
Democrat Talking Points:
- 8 of the top 10 cities (80%) are Democrat run. 16 of the top 20 cities (80%) are Democrat run. 35 of the top 50 cities (70%) are Democrat run. Republicans don't run big cities, they have 1 city, San Diego CA, that has more than 1M people that live in the city proper and that's a California Republican.
- Democrats run the cities where people actually live. The average population of Democrat run cities is +66% higher than the average Republican city. Of the total population in the sample, 68% live under a Democrat Mayor and only 29% live under a Republican Mayor.
- Democrats run cities where people are more closely packed together, averaging +49% in population density over Republican cities.
- Democrats run more diverse cities, with the average Black population % being more than double that of Republican cities.
- More people, being closely packed together, and being more diverse in background and opinion, is the reason for the difference in crime rate. Large cities also attract those looking for work, so the poverty rate is slightly higher. Democrat policy does its best to contain issues related to these factors, and since Republicans don't deal with them there's no way to tell if they'd be better or worse at it.
MY ANALYSIS
The problem is trying to determine causality for the increased crime rate in Democrat run cities. In technology it'd be called a "Root Cause Analysis." People like to say, "the numbers speak for themselves," they don't. Your worldview determines what you see in the numbers. "Is it the policy or the circumstance that causes the difference in crime rate?" That is the question at the core of this argument. That's what needs to be determined in the Root Cause Analysis.
In my opinion there is some truth to both the Republican and Democratic talking points on this subject. Democrats have to solve for problems that Republicans simply don't have to deal with.
It's like sitting on your couch watching an NFL QB throw an interception. You from your couch yell, "How did you not see that guy! You had a wide open receiver on the other side, how'd you miss that?!" In reality if you were placed in the same situation as that NFL QB you might or might not do the exact same thing or much worse.
Republicans for the most part are arm-chair QBing Democrat Mayors here.
At the same time a +167% increase in violent crime in Democrat run cities is nothing to ignore. It's hard to argue that a QB who has thrown 5 INTs in a game should stay in the game. Regardless of whether or not you think their back-up is worse, you kind of have to give him a shot.
Since we can't simply flip all Mayors in America for 2 years and see what the results are I attempted to normalize the data. The most direct way to do it is to simply pull out and analyze the middle 3rd of the list.
This eliminates the top 3rd which are massively populated and mostly Democratically run cities, and eliminates the bottom 3rd which are sparsely populated and mostly Republican run cities.
What we're left with is a list of 83 cities that are relatively the same size while the sample remains large enough to identify trends.
Quick Notes:
- Democrats still hold a +15% advantage in number of cities, and total population, but when looking at "Average Population by City" in the middle 3rd it's almost identical.
- Democrats are still more closely packed together, but only by a +6% difference vs. Republicans.
- The Diversity gap remains about the same.
- Avg. Poverty Rate goes up +0.5% in Democrat cities, and goes up +2.7% in Republican cities, closing the total gap by about half.
- The Violent Crime gap drops from +167% to +20%
- The Property Crime gap drops from +81% to +11%
What we see immediately is that circumstance does play a huge factor in the difference in crime rates between Democrat and Republican run cities, BUT it doesn't explain all of it. +20% violent crime rate isn't "10 times higher in Democrat cities" but it's definitely significant.
Circumstance also plays into Poverty Rate. The Republican Poverty number moves up more significantly as they're losing the bottom 3rd sparsely populated cities that were keeping their average low.
That the Poverty and Black % numbers move together is very interesting to me. It leads me to believe those two things are intrinsically linked and aren't all that changed by geography or city size. I didn't pull poverty data by city by race, I'll save that for another rainy day, but my educated guess is that Black people are over-represented in the % of poor people no matter where they live. Every city has its "bad neighborhoods" and that's where the black people live.
My conclusion is that Democratic policy is likely responsible for some of the difference in crime rate, but circumstance looks like the primary driving factor.
However, even boiling it down to the middle 3rd doesn't eliminate all possible factors effecting the numbers. Something as simple as the Black % being higher in Democratic cities, and law enforcement being more likely to report something as a crime when dealing with a Black person, could alter the numbers if you were able to normalize for it. Add 10 other variables like that in and you may see a completely different picture, you never know. Again, numbers do not in fact speak for themselves, there are typically complex variables in all statistics that sway the results that you should at least know about when forming an opinion.
This is why I take the time to do the research for myself when I'm able, to educate myself as much as possible on what is actually going on in the world, and I don't listen to talking points I hear on TV. I'd encourage everyone else to take the same approach, at least in not listening to people in the media who simply spit out sound-bites using data without giving you the complete picture as I've attempted to do here.
Finally, in an effort to provide complete transparency I've attached all my research data below if you'd like to pick through it yourself, and I'd encourage you to do so.
Stay Curious. Please Share.
Comments