How often do guns prevent crimes? It sounds like a simple question, and those who like to write reports on the danger of guns think so. Even those who defend gun rights often fall for this question. But the truth is, there is simply no easy way to tell.

Before I begin let me state my position, though in this particular essay I will not bother to justify it, simply state it to make my position clear. I believe we possess a right to own and carry firearms, even concealed, without government permission. In fact, I believe we have a right to own any and all weapons owned by the military. My rationale is simple. Joining the police or military doe snot grant one any special status, so anyone who owns weapons does so as a citizen, and if a soldier can have an M16 as a citizen, so can I. Otherwise we must argue that soldiers have special rights, and if they do, how did they get them? I can't grant anyone rights I don't have. So, I must have those rights, or have had them at one time. Rights being inalienable, I must still possess them. And, as a consequence, I can only assume any weapon that is legal for another to possess is just as legal for me. ("A Right Is A Right", "A Simple Proposal", "Symmetry and Asymmetry in Government")

But as I said, that is not relevant here. In fact, it may not be relevant to many who would argue with my conclusion. It seems that most gun control arguments don't pay the slightest bit of attention to rights. Instead, they act as if rights are irrelevant, as if the government's omnipotence is a given, and instead argue from some sort of vague "pragmatism'. ("The Shortcomings of Pragmatism", "Pragmatism Revisited", "Pragmatism Revistied, Again") I disagree with this argument as well ("Gun Control, The FDA and Regulating the Law Abiding", "Nuclear Disarmament and Gun Control "), but that is irrelevant here. What I intend to argue is that their justification is based on entirely faulty evidence.

Basically, the argument offered almost always takes the form of the cliched statement "you are more likely to be shot with your own gun than defend yourself" or perhaps "more accidental shootings take place than crimes are prevented." These arguments, when they bother to provide numbers and not simply make unsupported assertions, use simplistic assumptions, drawing on the number of individuals killed and comparing them to the number of criminals shot in the course of entering a home. It sounds plausible, but in truth the numbers are all but meaningless.

I won't spend much time on the first number. In general those killed in their home during a crime are just that, but even then there is some distortion possible. For example, is an armed drug dealer whose home is invaded by thieves or competitors and then dies in the ensuing shoot out really a home defender to count in the figures? In one sense he is, but as he is killed in the course of a dispute between criminal co-conspirators, that case hardly fits the pattern of an innocent individual shot with their own gun that the gun control proponents hope to evoke. Still, as I said, for our purposes let us assume the numerator is entirely valid, that the number killed with their own gun is valid.

That leaves us with the number of crimes prevented. And there we find the number is largely unknowable.

First, there are the overt cases of self defense which are not counted. Yes, those cases were criminals are shot is an easy number to compute. But that is only a small part of the number. What about those times a homeowner brandished a gun but did not use it? Maybe they are counted, maybe not. In some cases the homeowner may not have told the police, in others may not have filed a police report, and so many may have evaded our notice.

And then there are other cases. What about cases where a homeowner has a gun and so gets the confidence to go investigate a noise, scaring off criminals without ever showing the gun. Dud the gun stop the crime? Will it be counted? And what about the cases where having a gun allowed the homeowner to go outside to check on a noise, preventing a criminal from ever entering? In that case there was never a crime to report, and so it passes under our radar.

And it gets worse. What about states that passed concealed carry laws? Those who used to work as muggers, now knowing they could be shot, often change to other crimes. But how do we know how many muggings were prevented? We know, for example, that in Florida, when citizens were allowed to carry in their cars criminals changed to robbing exclusively rental cars. Ye,s it created a wave of crimes against out of state visitors which made the news, but what did not make the news was the other side of the equation, the even more massive decline in automotive crimes against state residents. Those crimes were stopped, and yet the numbers were never counted.

At this point some will argue that we can still count the number of crimes, we have only to compare the crime rate with the previous year. Or maybe with a comparable state. But, as with most such studies, that is a very dubious source. First, and most obviously, there are very few comparable states. Unless a state is nearly identical in terms of demographics, economics, laws, even geography, it is a very loose approximation at best. Nor does the past year make a good comparison. I recall when living in Baltimore, year after year, the murder rate would change despite a lack of changes in the police force's policies. However, when it rose above 300 per year, everyone would become indignant, and when it fell below 300 they would rejoice. What everyone forgot was that in neither case was it anything more than the outcome of chance. The police followed the same practices year after year, and yet they received both the blame and the credit, when the numbers changed form nothing but the random chance of how many chose to murder someone that year.

And that is the real problem with any attempt to create quantitative evidence. The same way we cannot tell how much business would have been supported by the tax dollars spent by the government, we cannot tell how many crimes were stopped by guns. Nor does it really matter.At least not in my eyes. The real question is not whether or not it is safe to let citizens have guns, but whether or not the function of government is to protect us from ourselves, or whether it is a tool we use to collectively protect our rights. ("Man's Nature and Government", "The Citizen Dichotomy", "My Vision of Government", "My Vision of Government Part II", "Prelude", "Liberalism, Its Origins and Consequences - Preface",  "The Problem of Pornography", "Free Speech, Absolute Rights and the Absurdity of "Balancing Tests"") Only once you establish the omnipotence of government do questions of practicality become important. And as far as I know, none of the gun control advocates have bothered to explicitly tell us that they think the government has the right to do whatever it wants to save us from our own stupidity. And until they do, this whole argument is pointless.

Still, since it will continue to be argued by both sides, I suppose it would help to bear in mind the fact that even the reported number of crimes prevented is but a fraction of the true number, that many crimes are prevented at a point too early for anyone to know. And any attempt to measure the number of such crimes prevented is, at the very best, a quite loose estimate. All of which makes the argument much more difficult than either side imagines.

POSTSCRIPT

I want to clarify my closing statement. In saying the argument is difficult, I do not mean the argument over gun control itself. I believe the argument is simple, that the rights of individuals are absolute and inalienable. What I meant by that sentence was, once the argument has degenerated into one over whether or not it is practical to allow citizens to keep guns, the proof is very difficult to assess. Since I believe the argument should end simply at the discussion of rights, I don't particularly care about the argument over the relative safety, but as some seem to think it important, I thought I should point out how hard an argument it is to evaluate honestly.