Dan’s “Rational Security” post and Jason’s provoking democratic searches response seem to me to have occupied part of the field of what I wanted to say here, which is that random suspicionless searches can be left to democratic controls without imperiling the entire constitutional order.
A bigger issue for Dan and privacy absolutists: not all anti-terror policies lead to Korematsu! Although I’m significantly more sympathetic to slippery slope arguments than I used to be, thanks to Volokh, I think Dan’s argument here is off-target. The differences between the internment cases, involving racially suspect classifications, and the searches here are evident. Most significantly, in a factual finding that commentators on this site appear to be ignoring, these really are random searches; the police aren’t permitted discretion to search any particular suspect class. Dan argues nonetheless that checking bags of subway entrants is a first step toward totalitarianism. I’d like to hear more about the mechanisms of this particular slippery slope. But until I do, my intuition is that a policy that burdens equally all residents is significantly less troubling than one that does not.
Dan also, I think, ignores my point that the court really didn’t defer to the government here, at least as deference is normally understood. Sure, the court is tougher on plaintiff’s witnesses, but that is because they didn’t have the relevant expertise. What is the court to do if the plaintiff doesn’t show up with the right folks, hire an independent security consultant? That isn’t how our system works.
I take Dan’s big point to be that this is an unwise policy. It (according to him) misallocates scarce dollars on a policy that will not have significant deterrent effects. I disagree that simply because the chance of search are low and terrorists might be able to evade the cordon we can conclude that there is no or low deterrence. But putting that aside, there is a space between what the Fourth Amendment permits and what smart police policy ought to be. (Thanks to my colleague Craig Green for reminding me of this). To conflate the two, i.e., to require the police to justify anti-terror searches as the least intrusive method necessary, or the most effective strategy policy possible, would simply be to substitute the anti-terror judgments of one group of elites (judges and scholars) for another (elected officials and police authorities). What is the normative argument for that result?