Mark Kleiman and the Volokh Conspiracy's pseudonymous "Philippe de Croy" agree that the recent "code orange" terrorist alert issued by the Department of Homeland Security was a pointless exercise. As "De Croy" put it, "the costs created by the warnings -- in anxiety, distraction, and wasted time -- greatly exceeded any expected benefits from them."

I think both are seriously underestimating those benefits. First of all, warnings like the current one might actually be correct, in the sense of having been been triggered by observations that are in fact signs of a real, specific, imminent attack. In that case, it is quite possible that a terrorist cell, suspecting that the authorities may know too much about them, might decide to postpone the attack for fear that their plans may be compromised. (Note that this result may occur even if the authorities don't actually have any information about the particular plan in question, but have issued the alert based on an entirely distinct set of clues.) The alert could thus, all by itself, end up saving many lives. Moreover, such a change of plans on the part of the terrorists--or even consideration thereof--may involve activities, such as contact with leaders abroad, financial rearrangements, or movement of personnel, that may expose them to detection.

Of course, the terrorists may decide not to change their plans at all. In that case, an alert probably increases (albeit only slightly) the likelihood that the attack will be detected or foiled by security officers or ordinary citizens exercising an extra degree of vigilance. Certainly it's hard to argue that in such a case, the alert was pointless.

But what if there's no attack planned at all? Even then, alerts can be useful (if they're not too frequent). Again, they provide an opportunity to "tweak" the terrorist network and see how it responds. Events (activations of particular communications channels, for instance) that correlate well with alerts can be noted and investigated for possible intelligence value.

Finally--and here is where I disagree most directly with Kleiman and "de Croy"--there is the matter of the effect of alerts on public attitudes. Both gentlemen agree that, as Kleiman puts it, "it would help if the population cultivated an attitude of calm rather than one of panic." Well, I can think of nothing more likely to reduce the public's level of panic than frequent warnings that force people to confront, accept, and eventually learn to live with the everpresent possibility of another attack. As I noted previously in a discussion of the DC sniper, when it comes to terrorist threats--or low-grade dangers in general, for that matter--familiarity tends to breed contempt. Holding off on warnings may, in the short run, reduce society's overall level of worry, but at the cost of making the next attack (God forbid) even more traumatic. The small (and surely declining) flurries of panic we experience now when alerts are sounded should pay off later in the form of more orderly public responses to terror threats in the future.