I blogged about this nearly six years ago, but I thought I’d mention it again, given the recent discussions about “smart guns.” It’s not the most important factor in considering whether to rely on a smart gun as one’s only gun (or whether the law should require that all guns be “smart guns,” once such guns become reliable and inexpensive enough). But it strikes me as an interesting inquiry, both for its own sake and for the broader light it might shed on how we think about risk-benefit analysis.

A high-altitude nuclear detonation can generate an electromagnetic pulse that will basically destroy unshielded electronic circuitry in line of sight — potentially over hundreds of thousands of square miles. There has also been talk of e-bombs, which are nonnuclear devices that could create an EMP over a much smaller area.

This is a serious risk. There has been a congressionally established Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack, which issued a detailed report that recommended a good deal of spending on such matters. The Department of Homeland Security has been investing time and money in the matter. It is, of course, eminently possible that there will be no such attack at any time in our lifetimes, and any particular time the attack is highly unlikely. But at the same time it is also possible that there will be such an attack at some time in the next several decades.

This, it turns out, is one of the little-known twists in the debate about “smart guns.” Smart guns, in theory, would only be usable by their authorized owner. This would be done using some technology, likely chip-driven technology — possibly some radio transponder that reacts to a special ring that the owner wears, or possibly even fingerprint recognition (though that would have to be mighty quick and reliable).

I don’t support laws that mandate smart guns, chiefly because there’s no reason to think that such guns will be reliable enough any time soon. But I certainly see the advantage of such guns, as a means of preventing the 100 or so fatal gun accidents and the greater number of nonfatal gun accidents involving kids that happen each year in the U.S.

If I had a child, and smart guns were reliable enough, I might well be willing to spend some extra money to get a smart gun instead of my current gun. And if (as I asked you to assume) such smart guns became generally about as reliable and about as costly as ordinary guns, I think smart gun mandates might be constitutional under the theory that they do not materially interfere with the right to keep and bear arms in self-defense. (Imagine a requirement that printing presses or cellphones contain anti-theft technology; I don’t think this would, by itself, violate the freedom of the press or of speech, if the requirement didn’t make the devices materially less reliable or more expensive.)

But the concern about electromagnetic pulses puts a different cast on things. Naturally, I don’t expect an e-bomb being set off in L.A. any time soon; but I also don’t expect a fatal gun accident in my house any time soon, since those are rare events, too.

But I do know that there’s a nontrivial chance that in my lifetime, there will be some terrorist or military attack on the place that I live. When that happens, there might well be serious social disruption caused by the attack, and extra need for me to be able to protect myself and my family. It would be just the wrong time to be armed with something that used to be a gun but that’s now just an expensive lump of metal.

We’ve generally lived our lives in environments of peace and civil order, but there’s no guarantee that this will continue; in fact, judging by recent human history, there’s reason to think that there’s a significant (10 percent? 20 percent? who knows?) probability that at least some time in our lives, our homeland will be attacked, possibly with sophisticated anti-electronic weapons, and civil order will break down. And when that happens, we’ll both be in special need of personal defense weapons, and in special need of personal defense weapons that haven’t had their innards fried to a crisp. And our police departments, to the extent that they’re still on the job after such an attack, will also be in need of such weapons.

Naturally, this is just one cost that one has to consider — both in one’s personal buying decisions and in deciding what the constitutional rule ought to be — and as I mentioned the benefits of smart guns, if they become highly reliable, are nontrivial. Moreover, the cost might be minimizable, for instance if the guns end up being properly shielded (though I understand that creating such shielding is not easy, which is one reason that e-bombs are potentially powerful weapons), or if the guns are set up so that when the “smart” technology fails, the result is a working dumb gun rather than an inoperable one. And of course what makes this interesting is that the magnitude of the risk is not just low but hard to determine. At the same time, I think it’s hard to deny that this is a cost that we ought to consider in some way (or to have a good justification for not considering it)?


  1. How should individual citizens consider this cost in deciding about their own personal self-defense (especially if they think they can’t afford both a “smart gun” and a backup purely mechanical gun)?
  2. How should police forces consider this cost in deciding whether to arm their police officers with “smart guns” that might fail in the unlikely but potentially very serious event of an EMP attack — a situation where armed police presence might be especially important?
  3. How should legislators and voters consider it in deciding whether to require smart gun technology (again, once it becomes generally reliable and inexpensive enough — though perhaps EMP resistance should be seen as part of the “reliability” inquiry)?
  4. How should judges consider it in evaluating the constitutionality of smart gun requirements?
  5. Are there technological features that should be considered, for instance a provision that smart guns be designed so that they are properly shielded, if that’s possible, or that, if their electronics are destroyed, the weapons fail into a mechanically operating mode rather than into an inoperative mode?

Again, note that the government is already taking EMP seriously enough to study it, and is at least considering spending serious money to harden critical infrastructure against it. The question is how we should consider EMP risks when it comes to a different form of social infrastructure — both citizens’ own self-defense tools and such tools owned by police departments.