On gun control

Americans honor those who die in mass shootings via the strange ritual of repeating their (usually not very well thought out) opinions about gun control. In that spirit, I thought I would share mine.

First, I do not think that individuals, in general, have a right to bear arms. I think the fundamental right is to self-defense. This is a right that those on the left really ought to like, as it is all about allowing the vulnerable to protect themselves from thugs of various sorts.

I know that, at this point, I have already lost some Europeans. The sorry spectacle of Tony Martin, the fellow in England who shot the 10th person to rob his farm being sent to prison was actually defended by some. If you think the weak should just take it and take it and take it until such time as some government officials arrive, even if they never do, then we part ways right here.

Returning to my original stream of thought, in my view, the right to self-defense is limited in several senses. First, the response should be proportionate to the rights violation. If someone walks on your lawn, you do not get to shoot them, or even hit them, even though they have violated your rights. Second, you must make reasonable efforts to call in the proper authorities, and you must hand over the situation to them when they appear, assuming they actually do arrive and manage to do more than just shoot your dog and seize your cell phone for having the audacity to record them (your employees, doing their jobs) when they do appear. Third, your right to do harm ends when the rights violation ends. You are not allowed to shoot someone in the back while they run away, even if they just tried to kill you.

In addition, I think people have the right to use technology to aid them in exercising their right to self-defense. The right to self-defense would not be very useful for the weak if they were not allowed to use technology to assist their efforts. This right, too, is limited in my view. The key limit is that ex ante restrictions may be imposed to prevent the technology doing more harm than good. I think we can all agree that, for example, individuals may not possess nuclear or (more serious than mace or pepper spray) chemical weapons. They would have a fine deterrent effect, and end many rights violations, but those benefits surely do not exceed the likely expected costs of misuse. I think that many of the thoughtful subset of those who advocate gun control of various sorts think that guns fall into the same category as nuclear weapons.  I think their implicit claim about how a well-done empirically-based cost-benefit analysis would come out is incorrect, but that is an empirical question.

My framework would allow, and I would support, a law that gun owners be required to take a non-trivial course in how to use a gun, and pass a test of some sort indicating that they do indeed know how to safely own, maintain and use a gun. In this sense, I think that guns are like cars. People can use them, but because of the danger of misuse - cars are very effective weapons too - we restrict their use to individuals who have demonstrated the ability to drive, some knowledge of the rules of the road, and who do not misuse the car, for example by driving drunk. In a similar spirit, I have no problem with lifetime bans on gun ownership for individuals who use a gun to commit a crime.

That's what I think. My sense is that my view is quite different from both the common blue team view and the common red team view. I also think that focusing the discussion on what constitutes legitimate self-defense, and on empirical cost-benefit analyses, is useful to advance the discussion beyond the bumper sticker level.

I end with some remarks about the evidence and about the implicit cost-benefit analysis that underlies my view (and which reflects only a passing, rather than a deep, knowledge of some of the relevant literature). First, I think simple comparisons to Canada are not very useful. Canada is full of Canadians. They are different in their social capital and in their history in ways the challenge the meaningfulness of comparisons. There are also more  guns around in Canada than a lot of Americans think. Second, the socially optimal number of mass shootings (like the socially optimal number of most bad things) is non-zero. Third, the number of mass shootings with even very strict gun control will be non-zero. This affects the cost-benefit calculation associated with gun control of various sorts. Fourth, my sense is that we do not have very good evidence on the deterrent effects of guns. There is a literature, and some data, on this, but my sense is that the data are not that good and that both for that reason and because of the partisan nature of the issue, the estimates vary widely. Estimates based on the number of actual instances  where an individual scares away a criminal by brandishing (or more) a gun, which is what we mostly have, are, moreover, a lower bound on the extent of actual deterrence, which includes those crimes not attempted in a world without gun control but attempted in the world where it exists. Finally, the thermostat model does not apply here, as usual. What matters is not the amount of control in some naive sense but its design.

I welcome pointers to the (serious) literature (i.e. not your favorite partisan website, not the NYT editorial page, etc.) in the comments.

JW and JW can think of this as my response to their facebook post and tweet, respectively.