Today in Your Fact is Not In Fact a Fact:

This chart is making the rounds, purporting to demonstrate that there is no relationship between gun ownership in Canada and rates of homicide.

So let's look at this chart.

The thing that leaps out is it compares a rate of occurrences (homicides) with the raw number of guns owned, not guns owned per capita.

Why would anyone make such a comparison? Is anyone suggesting that if I own two guns, I am twice as likely to kill someone as if I owned one?

Obviously, a better comparison would be per capita ... but it would make more sense to compare the rate of gun *ownership* with the rate of homicide.

So let's try that.

We don't have a rate of gun ownership. We have a rate of PAL and RPAL licensing, so we'll use that.

This chart compares the rate of PAL/RPAL licensing with the rates of homicide overall and the rate of firearms homicides. (Data sources: Statscan, the Gun Blog)

The results are ... less convenient

(Note: in that chart I am showing homicide rate per *billion* simply to fit everything on the same scale. This doesn't matter; what matters is the slope.)

So the slopes seem to coincide rather neatly, which explains why someone might choose not to use this data if he had, say, some kind of point to "prove."

But this should not be taken to show that these rates are in fact related.

The homicide rate is driven by all kinds of things. Guns are not the driver of violence in society. So this chart should not be taken to prove that firearms availability drives homicide rates.

But it should be taken to demonstrate that the chart making the rounds is bullshit.

The bullshit chart claims greater firearms availability does not correlate with greater firearms homicide. And that's just nonsense.

Firearms availability correlates with firearms homicide for the same reason that miles driven correlates with accident rates. This is basic.

So what the bullshit chart is really trying to do is to throw the most basic ideas into dispute, which is the essential goal of gun trollery. And it should be ignored.