William M. Briggs

Statistician to the Stars!

Firearm Homicide Rates Around The World: Comparing With USA

Here’s an around the world tour of homicide rates by firearm, brought to you by our friends at the United Nations Office on Drugs and Crime. Rates are per 100,000 population, from 1995 to 2010, where it is important to note that the UN only resolves to the nearest tenth:

Figure 1

Figure 1

Each line represents one of 116 countries. Rates are from near 0 to nearly 80, making the chart difficult to display. Hold off on identifying which countries are which a moment, and notice that for many “highly rated” countries, their signals are increasing. All is not well in the world.

Here are four countries that often find mention by American politicians lamenting the rates in this country:

Figure 2

Figure 2

The black lines are the homicide rates (left vertical axis) and the red (right vertical axis) the percent of all homicides committed with a firearm. Citizens of Switzerland have to dodge twice as many bullets than denizens of Sweden. Curious that Australia (motto, “We like beer”) has always seen lower rates than Germany (motto, “We like beer, too.”).

The rates track the percentages closely. This was exactly what we discovered in the United States, where the homicide rates for all manner of weapons and killing methods tracked the overall homicide rate. The interpretation is that whatever is causing the increase or decrease in rates isn’t preferentially captured by any weapon or method. Guns per se are not therefore a special problem.

Here are three countries which closely match the USA (with the UN’s estimates of the USA numbers; for our government’s official statistics, see this post).

Figure 3

Figure 3

The USA’s raw rates are roughly 6-12 times higher than the first four countries, but right on par with that capital of fear Liechtenstein. It’s cheating to list Liechtenstein, of course, because in 2008 there was only one recorded murder, and that with a gun (bang!). But fair’s fair with Mexico, where we see the effects of the collaboration with Mexico and our safety-first government’s official “war” on drugs. Sadly, in any war, there are always “collateral damages”. Nevertheless, we can be confident that soon—any year now—drugs will cease to be used by our citizens.

A direct comparison with the USA and (say) Sweden is difficult, at least because demographics vary widely between the two countries. Sweden’s homicide rate is roughly constant all across the country, where the rates in the USA are highly State and neighborhood dependent. Compare Bedford-Stuyvesant, Brooklyn (ouch) with Charlevoix, MI (0 homicides per 100,000 for several years in a row).

Sweden is about 450,000 square-kilometers, which is 4.5% the size of the USA. Sweden has 9.5 million people, which is about the number of folks on a weekday in New York city; Sweden’s population is only 3% of the USA. We can thus find many Sweden-sized chunks of the USA which have very different rates, some lower than Sweden’s, others higher.

Sweden is also racially homogeneous. Sweden’s black population is less than 1%, while it’s around 13% in the USA. This is important because in the USA the black homicide rate is about 7.5 times the white homicide rate (see this post for a complete analysis). This disparity has been noted in other countries, too.

In the USA, the overall white homicide rate in 2010 was just under 3.5. Given about 60% of homicides where committed by firearm, this puts the white firearm homicide rate at about 2 per 100,000. The black homicide rate in 2010 was about 25, giving an estimated black firearm homicide rate around 15 per 100,000, again approximately 7 times the white rate. (This assumes that whites and blacks use firearms at the same rate.) Again, this is overall. For “real” rates, one must always know into which neighborhood one has wandered (State rates here).

Here is each country with the most recent rate listed for each, sorted from high to low, reading left to right. Because of the varying data sources and other difficulties in measuring and comparing the numbers across countries, rates differing by only a few tenths are probably indistinguishable. Also recall the UN only records to the nearest tenth. Thus Hugo Chavez’s workers paradise Venezuela and not-so-happy-after-all Jamaica should be considered about the same.

Country Rate Country Rate Country Rate Country Rate
Honduras 68.4 Liechtenstein 3.0 Ireland 0.5 New Zealand 0.2
El Salvador 39.9 Uruguay 2.8 Bosnia and Herzegovina 0.5 Mauritius 0.1
Jamaica 39.4 Peru 2.6 Serbia 0.5 Algeria 0.1
Venezuela 39.0 Sierra Leone 2.3 Zambia 0.4 Turkmenistan 0.1
Guatemala 34.8 Chile 2.2 Liberia 0.4 Mongolia 0.1
Saint Kitts and Nevis 32.4 Albania 1.8 Finland 0.4 Azerbaijan 0.1
Trinidad and Tobago 27.3 Congo 1.6 Sweden 0.4 Israel 0.1
Colombia 27.1 Maldives 1.6 Croatia 0.4 Qatar 0.1
Belize 21.8 Sri Lanka 1.5 Portugal 0.4 Belarus 0.1
Puerto Rico 18.3 Cambodia 1.4 India 0.3 Hungary 0.1
Brazil 18.1 Kazakhstan 1.3 Nepal 0.3 Poland 0.1
South Africa 17.0 Macedonia 1.2 Armenia 0.3 Norway 0.1
Dominican Republic 16.3 Bangladesh 1.1 Denmark 0.3 United Kingdom 0.1
Panama 16.2 Viet Nam 1.0 Northern Ireland 0.3 Slovenia 0.1
Bahamas 15.4 Uganda 0.9 Greece 0.3 France 0.1
Ecuador 12.7 Lebanon 0.8 Netherlands 0.3 Australia 0.1
Guyana 11.5 Turkey 0.8 Cuba 0.2 Grenada 0.0
Mexico 10.0 Switzerland 0.8 Tajikistan 0.2 Bermuda 0.0
Philippines 8.9 Bulgaria 0.7 Malaysia 0.2 Hong Kong 0.0
Paraguay 7.3 Italy 0.7 Czech Republic 0.2 Japan 0.0
Anguilla 7.1 Belgium 0.7 Moldova 0.2 Korea 0.0
Nicaragua 5.9 Egypt 0.6 Slovakia 0.2 Brunei Darussalam 0.0
Saint Vincent 5.5 Taiwan 0.6 Ukraine 0.2 Singapore 0.0
Zimbabwe 4.8 Georgia 0.6 Estonia 0.2 Bahrain 0.0
Costa Rica 4.6 Luxembourg 0.6 Latvia 0.2 Romania 0.0
USA 3.2 Canada 0.5 Lithuania 0.2 Iceland 0.0
Barbados 3.0 Kyrgyzstan 0.5 Spain 0.2 Malta 0.0
Argentina 3.0 Cyprus 0.5 Austria 0.2 Monaco 0.0
Palestinian Territory 3.0 Jordan 0.5 Germany 0.2 Solomon Islands 0.0

9 Comments

  1. Thank you this posting. I have been researching statistics on homicide and gun deaths and this puts the statistics in an understandable format.

  2. It is a good job that it does not go up to 2013 and include such places as Afghanistan, Libya and Syria.
    Or don’t they include Government or terrorist murders in the statistics?

  3. A C Osborn,

    I’m sure it would be truly unfair to include deaths caused by the Government. After all, you didn’t kill them.

  4. A C Osborn,

    “Or don’t they include Government or terrorist murders in the statistics?”

    The point of the exercise is evaluating the effect of civilian gun ownership on murder rates. If they did include either of the above we would have to figure out how to remove them for a valid analysis.

  5. Briggs

    5 February 2013 at 2:51 pm

    A C Osborn,

    If you have accurate sources of data on these things, send them my way.

  6. For Figure 1 use a semi-log plot and omit the zeros. There is always a solution for chart display.

  7. Briggs

    5 February 2013 at 3:18 pm

    William Sears,

    Tried it; just as ugly. Also tried limiting to 0-1; but too many lines. Best to look at individual countries.

  8. Sander van der Wal

    6 February 2013 at 6:41 am

    It appears that a lot of the high-rate countries are on the American Continent. South-Africa is the first country on a different continent, and the Phillipines the second one.

  9. “The rates track the percentages closely. … The interpretation is that whatever is causing the increase or decrease in rates isn’t preferentially captured by any weapon or method.”

    Surely that conclusion would be better supported by a situation where the percentages are fairly constant. (In this post, where ‘percent’ is per hundred homicides, not the previous one, where ‘percent’ was per hundred people.)

Comments are closed.

© 2014 William M. Briggs

Theme by Anders NorenUp ↑