hoist them on their own petard

Everyone knows I am a huge fan of using graphics to illustrate hard-to-describe concepts, so it should come as no surprise to anyone that I love the hell out of Howard Nemerov’s post comparing firearm ownership against a variety of other metrics in countries throughout the world. Of course, one of the best parts about this post is the source for all of his data – the United Nations, a notoriously anti-self-defense, anti-firearms organization that is intentionally and willfully supporting policies that run directly contrary to the information they themselves collected. Diogenes’ lamp, anyone?

Mr. Nemerov links to all the various sources he used for his number-crunching, but I am going to borrow his graphics and type up short synopses below them:

On average, throughout the world, as the private ownership of firearms increases, the number of homicides decreases (the correlation, -0.10, is admittedly small).	On average, throughout the world, there is no correlation between civilian firearm ownership and suicides (the value works out to be 0.04). In this case, the likelihood of causality in either direction is excruciatingly small.
And now we get into the real fun… The UN keeps metrics on how “free” a country is, and as the number of firearms in civilian hands increases, the “freedom” of that country likewise increases, with a correlation value of -0.34 – given the 200ish degrees of freedom involved, that number may be considered “significant”. (Note: lower “freedom” scores are better, counterintuitive though that may be.)	Likewise, based on another set of UN metrics, as the quantity of firearms in civilian hands increases, so too does the economic “freedom” of that country (Spearman’s value of 0.48 – again “significant”).
And significant (0.39) correlation strikes again, in that as private ownership of firearms increases, governmental corruption decreases (Higher numbers are better when it comes to corruption, again, counterintuitively.).

So what does all this mean? Well, in the small scale, as Mr. Nemerov points out himself, it means the United Nations is pointedly and intentionally ignoring its own statistics and information when proposing what to do about firearms – their decidedly anti-firearm stance, coupled with their claimed desire to ensure the freedom of all peoples in the world, simply do not make sense when put together. On the larger scale, unfortunately, the dataset these graphs are derived from is largely suspect, to the point of being useless – different countries track crimes in different fashions (up to and including omitting them entirely if no suspect is convicted, and, of course, ignoring state-sponsored murders), the “freedom” and “corruption” information is purely subjective (both what is gauged and how), and the source (the UN) is known to … adapt … information to be more salutary to themselves or countries they support.

And, finally, on the largest scale, my rights are not subject to any statistics, studies, or calculations – you can no more enslave me if the data indicated doing so would make the country safer than you can strip me of my right to self-defense for the same reason. This information, however, so very kindly provided to us by one of the largest anti-firearms organizations in the world, does give us bounteous cold, hard facts with which to disprove the specious talking points of anti-rights nuts – nuts sometimes even working for the very same organization responsible for this data! And, if nothing else, in a complicated world where even the mention of “statistics” or “probabilities” can cause a person’s eyes to glaze over, graphics matter when it comes to expressing complicated subjects, and Mr. Nemerov’s charts will help immensely in that regard.

Of course, on the “entertaining” side of the spectrum, one has to wonder what the average anti-rights cultist‘s reaction will be to this news… On the one hand, they cannot libellously discredit the source (given how they all love the United Nations for its anti-firearms stance) or the data (given that it was gathered by the UN), but on the other hand, these statistics indicate that their arguments are on very rocky ground indeed. One almost has to feel sorry for them… almost.

(As a reminder to all readers, correlation does not imply causation, though causation requires correlation. Put another way, positive correlation indicates that as value X increases, so too will value Y (with the inverse holding true for negative correlation), but not necessarily because of value X.

When it comes to significance of correlation, I cheated by using a table (note this is a two-tailed probability, so follow the instructions on the page), but you can solve it by hand if you want (and have the full dataset). And despite its name, “significant” correlation does not necessarily prove, or even demonstrate, causality – it simply indicates “that the probability of the relationship you have found being a chance event” is rather small.)

(Courtesy of Joe Huffman.)