First-Order and Second-Order Judgments

There is an issue that comes up repeatedly in discussions of national security law. Suppose we define “law” broadly to include the optimal allocation of institutional authority to establish and execute national security policy. On what grounds can academic commentators who lack expertise in national security policy argue for any particular allocation of such authority, in particular circumstances? If one is agnostic about the merits of first-order policies that one cannot judge, for want of expertise, how can one make a second-order argument that some institution is relatively more likely to make “good” national security policy? To know what counts as “good”, wouldn’t one have to judge the merits of first-order policies? In our book, Eric and I are agnostic about the merits of many national-security policies the executive has pursued during emergencies, in particular the post-9/11 emergency. But we urge the second-order claim that the executive does and should receive even more deference during emergencies than during normal times. Is this inconsistent? (Lou Fisher’s post can be read to suggest that it is).

On closer inspection, however, this looks like a pseudo-puzzle. In fact it is routine to make second-order judgments when, and indeed because, one cannot make first-order judgments. I may have no idea whether the diet prescribed by my Harvard-trained doctor is superior to that suggested by the diet guru Dr. Atkins; but I can make a coherent judgment to trust the former’s credentials and expertise over the latter’s. Where one cannot judge outputs or results, one can still judge inputs, such as training, resources and expertise. Moreover, it is often easier to make relative judgments than absolute ones. (How tall is the Sears Tower? I have no idea. But I am confident it is taller than the Washington Monument, whose height I do not know either). And the thesis we defend in our book is entirely relative: as one moves from normal times to emergencies, the executive’s comparative advantages increase and other institutions should shift more authority to the President. “The deference thesis does not hold that courts and legislators have no role at all. The view is that courts and legislators should be more deferential than they are during normal times; how much more deferential is always a hard question and depends on the scale and type of the emergency.” (Terror in the Balance, p. 6). Because it is so hard to know how much deference is correct, we plump for the historical level of deference, which has been very great during emergencies; those who would argue that the historical level has been too great have a kind of burden of proof.

Our particular second-order judgment about the relative benefits of shifting power to the executive during emergencies may be right, or wrong, on the merits. But there is nothing inconsistent in combining (1) agnosticism about first-order judgments of national security policy with (2) clear second-order judgments about relative institutional capacities. Indeed, to the extent (1) is correct, (2) is all the more pressing.