US Crime: How do we rank?

Image Credit: https://www.unodc.org/documents/data-and-analysis/gsh/Booklet1.pdf Page 25

Before we even start getting into the nitty-gritty of why our crime stats look the way they do, we need to establish a baseline of where the US stands in terms of crime versus the rest of the world.

Once we understand that, we can then place crime and associated deadliness into the appropriate context.

So just how do we stack up versus the rest of the world?

Crime Rates

Overall crime rate is calculated by dividing the total number of reported crimes of any kind by the total population (per 100,000 people) — worldpopulationreview

The data here is pulled from worldpopulationreview. They note that there is a strong correlation between age and crime with most crimes being committed by younger cohorts.

We can extrapolate this to mean that unusual countries where a greater percentage of their population is younger will demonstrate greater crime rates. I’m ignoring this aspect in this section but will explore it later.

Interestingly, the site notes that the world’s lowest crime rates are seen in countries that have very effective law enforcement such as Switzerland, Denmark, Norway, and Japan.

The US ranks 56th out of 137 countries with a crime rate of 47.81 whereas a higher number rank is better.

Let’s put that into perspective:

• Afghanistan ranks 4th having a crime rate of 76.31
• Mexico ranks 39th having a crime rate of 54.19
• Iraq ranks 53rd having a crime rate of 48.42
• The United States ranks 56th having a crime rate of 47.81
• The United Kingdom ranks 64th having a crime rate of 46.07
• Canada ranks 82nd having a crime rate of 41.89
• Russia ranks 86th having a crime rate of 39.99
• China ranks 109th having a crime rate of 30.14
• Japan ranks 130th having a crime rate of 22.19

I chose these countries based on the general salience I perceive these countries have in general media. This data is also dated for 2020 and crime has definitely generally trended up since then.

So, the United States has a lower crime than Afghanistan, Mexico, and Iraq, but a higher crime rate (in some cases far higher) compared to other world Superpowers.

Not exactly a great look, but of course, there are a few factors that aren’t controlled for such as the above-mentioned age of the population and the accuracy of the reported crime rates in each country.

It’s important to note that we’re not taking into account the severity of the crimes or the type of crimes being committed, but just observing aggregate numbers.

Age

I’ll quickly take a peek at the age cohorts of each country given that we noted above that younger (population age-wise) countries have higher crime rates.

Let’s again turn to worldpopulationreview.

The US ranks 60th out of 225 countries with a median age of 38.5 whereas having a lower rank represents an older country.

Let’s put that into perspective:

• Afghanistan ranks 200th having a median age of 19.5
• Iraq ranks 184th having a median age of 21.2
• Mexico ranks 130th having a median age of 29.3
• China ranks 61st having a median age of 38.4
• The United States ranks 60th having a median age of 38.5
• Russia ranks 51st having a median age of 40.3
• The United Kingdom ranks 49th having a median age of 40.6
• Canada ranks 39th having a median age of 41.8
• Japan ranks 2nd having a median age of 48.6

Following the basic heuristic of a younger country having higher crime rates, we actually see that these rankings resemble the rankings of crime rates which sanity checks that statement.

I actually wonder if having excessive crime rates increases the mortality rates and decrease the lifespan of a population thus causing a greater correlation of lower age to higher crime.

Notably, Afghanistan has a very young population and a corresponding very high crime rate, while on the flip side, Japan has one of the oldest populations and a very low crime rate.

Here’s a visual example of just how extreme Japan’s aging is:

Image Credit: Victoriali2

Basic Linear Modeling

The following models are not rigorous and should not be used as is for scientific purposes, but rather as a simple visual starting point.

Out of sheer curiosity, I decided to plot the above crime rate data against the world population just to observe just how strong the stated relationship was.

Nothing fancy here, just plain old Excel.

We actually see a surprising coefficient of determination (R² value) here at .3 which when observing social behavior (like crime) indicates a somewhat correlated relationship.

As mentioned as the median age of a country increases, the more likely they are to have lower crime rates, which is reflected here.

To properly graph this data rigorously, we’d likely want to strip out outliers (countries with an excessive crime rate or far below average crime rates), weight data so we trust the data from more populated countries, and strip out data from low population countries.

Not to mention crime is definitely not just dependent on age and involves a multitude of additional factors including social safety nets, economic stability, the effectiveness of law enforcement, and population happiness among others.

Following the above formula, Crime Rate = -1 * (Median Age) + 78, we see that the US should have a crime rate close to 39.5 crimes per 100,000 people.

The US instead has a crime rate of 47.81 crimes per 100,000 people.

I also wanted to check if there was a relationship between crimes and population sizes of the country, though there’s nothing that indicates that would be the case.

Maybe population density proxying urban or rural environments could be more likely to have a relationship.

Due to the vastly disparate population sizes of countries from a few hundred thousand to over a billion people, I compressed the scale using the log of the population.

Visually, we see that there’s a slight uptrend, but don’t be misled by the apparent upward trendline. If we take a look at the R² value here, it’s effectively negligible at 0.0295, indicating no relationship.

Remember, correlation is not causation and thus the existence of any relationship indicated in any linear model, especially one as simple as this does not mean that one factor caused another.

A commonly used example is that crimes go up with the number of ice cream sales. It doesn’t take much to see this makes no sense. The true “cause” to both is likely heat and it just so happens that crime and ice cream sales are correlated.

Closing Thoughts

It came as quite a surprise to me (though maybe it shouldn’t have) that the United States is by no means the safest country in the world compared to other global superpowers.

I thought we would at least rank competitively albeit on the lower end of the global superpowers.

Even after “adjusting” for the median age of the population we see that the US has a higher than expected crime rate.

Though it’s worth taking with a grain of salt given my aforementioned caveats.

This begs the question, why?

Follow along as I dig into the factors that make up the crime rate.