Inconvenient Data: There’s No Relationship Between Gun Ownership and Murder Rates

blackboard data board calculation numbers statistics

Gun prohibitionists claim that that gun control works. They say gun control laws reduce the number of guns possessed by law-abiding gun owners. They say that’s a good thing because they want you to believe that guns cause crime. In particular, they say that guns result in violent crimes like murder.

I don’t know if that’s true, but I do know that we can find out. Countries have different cultures and different rules about owning guns. That leads to different rates of gun ownership. Different countries also have different murder rates. Put those variables together and we can see if guns really do cause crime.

The data isn’t as good as we’d like it to be. The last comprehensive international data on gun ownership was from the Small Arms Survey of 2017. People can update their estimates based on trends, but that isn’t the same as real survey data taken in the last few years.

The data on murder we have is from the United Nations Office on Drugs and Crime’s Global Study on Homicide. The problem is about half the countries are missing when we look at the data on murder rates. Japan and France are glaring examples.

I asked a demographer/sociologist about that missing data. She said that many countries do a poor job of sorting out murder, suicide, and defensive homicides. In a perfect study, we’d have worldwide survey data taken in the same year. We don’t have that, but it turns out that the small errors don’t really matter much.

Let’s look at the original claims by gun prohibitionists. They say that guns cause crime. It’s like a mantra they repeat over and over again: It’s the guns. More guns cause more crime, they say, and fewer guns would mean less crime. It’s as if the mere existence of firearms compels criminals to rob, assault, rape and murder their victims.

A simplified version of what the Gun Control Industry would have you believe would look like this. Here is what the murder rate would look like when we plot it against the rate of gun ownership.

The problem with their argument is that isn’t what we see when we look at real world data.

To be fair, we wouldn’t expect real data to look like that. It wouldn’t be that clean, that direct a relationship. Lots of factors influence the murder rate. The most obvious is that we’ve been killing each other since long before guns were invented.

There is a baseline rate of murder in all societies, even if there are no guns to be found. Sociologists have looked at rates of crime to find that income levels, upward income mobility, marriage rates, age of family formation, and rule of law are just a few of the most significant factors that affect murder rates. We find that other factors dominate the effect of gun ownership when we look at the 70 countries for which we have data.

Here is a plot of the gun ownership rate (guns per 100 people) versus the murder rate (murders per 100,000 people) . . .

gun ownership vs. murder rate chart plot

What’s noticeable here is that we see a slight negative correlation between murder rates and gun ownership. That’s probably an effect of income. You need to be a relatively affluent society to own lots of guns. That wealth comes with high rates of employment and the rule of law. That doesn’t look significant to me.

Say what you will — and the gun banners certainly will — but the data doesn’t show a correlation of gun ownership and murder rates. We also see a number of countries in the data with very low rates of gun ownership and frighteningly high murder rates. We also see countries with high rates of gun ownership and low murder rates.

Violent crime has many causes and looking at the data doesn’t show a relationship between gun ownership and murder rates. The obvious question is what that means.

I’m a retired engineer, not criminologist or a sociologist. I spent an hour putting this data together. If I can figure out that civilian gun ownership has an insignificant effect on the murder rates in various countries, then so did the researchers who are paid by anti-gun billionaires. This data is out there for anyone to see. They’ve seen it and it means they lied to please the people who are paying them.

Spend enough time wading through enough datasets and include (or exclude) the right variables and you can “find” whatever effects and relationships you want. Shame on them. Also, shame on us for giving them any attention and not checking the data for ourselves.

Please, take a look at the data on your own if you think that I’m wrong. Here are the datasets (here and here) I used. Ask yourself how hard you’d need to torture the data to show that guns cause crime. Then ask yourself why you’d want to do that.


This article originally appeared at the Slow Facts blog and is reprinted here with permission. 

10 Responses

  1. The anti-gun crowd does not care about homicide or even violent felonies committed without a firearm. That is, the argument is “fewer guns leads to fewer gun homicides”. While that’s still a fallacy (see Dr. Lott’s work over the years), remember, the “party of science and facts” don’t want to hear your facts.

    Of course, if you did the unthinkable and created charts based on economic indicators, or heaven forbid, race……

    With reference to nations such as Japan, etc: Very few countries in the world are as open to a veritable melting pot of cultures. Japan, Australia are often propped up as examples, however, their homogenous society coupled with their land-borders with other countries is non-existent. Further, as you stated, murder/suicide/… numbers are generally suspect both in numbers and categorization.

  2. They always see correlation where there is none and no correlation where there is plenty.
    Not that correlation is causation but still.

  3. There was a similar study done in Canada in the 1980s that came to the same conclusion. I’m trying to find a link, and will post when I can.

  4. Those links show the whole story. The people who say that if we get rid of guns crime will stop. Well, those facts don’t bare that out. Those people by getting rid of guns just want to oppress the people. The criminals don’t care about the laws. They will keep shooting people. Chicago has the strictest gun laws, however they lead the country in murders, can anyone explain that?

  5. One set of statistics always stands out when it comes to “gun control”. Every country that instituted gun control wound up banning all firearms. Once those countries banned all firearms, they began murdering their own people. Since the beginning of the 20th Century until Present, over 250 MILLION innocent people have been murdered by their own governments due to “gun control”. Red China has murdered over 100 MILLION and still counting. Soviet Russia murdered over 50 MILLION (mostly Ukrainians) and still counting. Nazi Germany murdered over 15 MILLION. Imperial Japan (from 1931-1945) murdered over 12 MILLION (many of them Chinese). North Korea has murdered over 5 MILLION. Afghanistan under Communist rule before the Soviet Invasion in the 1980s due to the people revolting against Communists there, murdered over 5 MILLION. Communist Vietnam and Communist Cambodia (under the Khmer Rouge) murdered over 2 MILLION each. In the United States, when the U.S. Army disarmed 300 Lakota Indians at Wounded Knee, South Dakota, on December 29, 1890, they murdered all of them; mostly women, children and seniors, with Gatling guns, cannons and repeating rifles.

    These statistics are out on the internet for all to see, and have never been refuted.

  6. It’s not proper to do a straight line fit with a major outlier like the US. So the US should be removed from the data. Then the conclusion is “no correlation”.

    It is true that correlation is not causation, but if causation does not produce correlation, then the effect is not distinguishable from random noise. So lack of correlation indicates an unimportant effect.

    You get a similar result if you plot homicide rate or suicide rate vs. strictness of gun laws.

    1. Without taking time to do a linear regression, just eyeballing the orange dots here, it looks to me like removing the US “dot” makes the slope of the trend line a very whole lot steeper. More guns, fewer murders.

      Why so, one wonders. A conjecture, for discussion: If guns are available to women, women can deter bigger stronger more violent men from murder with knifes or clubs or just fists. If guns are NOT available, then women wake up dead.

      If the data were sufficiently detailed one might plot ONLY the rate of murdered women against the rate of gun ownership, and see if the trend persists.

  7. Many people have problems understanding the data because it is too convoluted to reduce to a bumper sticker slogan.
    Murders, Justifiable Homicide, Suicide and Accidents all get thrown together, not to mention that the focus on Gun Deaths ignores all other murders although the victims are no less dead.

  8. I noticed a few issues with the data sets. First, the murder rate map is almost empty. This is because government types are loath to discuss anything that reflects poorly on them, so they either don’t report the data publicly, or they don’t even bother to collect it. Second, the United States’ Southern border is occupied by the most violent nation for which there is data, Mexico. Given the infiltration of the cartels across that border, it’s reasonable to suggest that the activities of said cartels has been exacerbating murder rates in the U.S.. According to this, without the influence of Mexican and Central American criminals, the United States would be a lot closer to the bottom of the graph. Then, there’s the fact that, excepting the brief reprieve from the Trump Administration, our own government has been industriously importing Third World criminals (many countries have been openly emptying their prisons and asylums into the U.S. with the blessing of Clinton/Obama/Biden). That must also be accounted for.

Leave a Reply

Your email address will not be published. Required fields are marked *