Issues of method in analyzing the policy response to emergencies.

• Author: Tushnet, Mark V.
• Position: Response to Richard A. Posner and Adrian Vermuele, Stanford Law Review, vol. 56, p. 605, 2003

INTRODUCTION I. RATCHETS II. FEAR CONCLUSION INTRODUCTION

Accommodating Emergencies (1) is a superb representative of what I think of as the second generation of responses by legal scholars to the circumstances brought into existence by the terrorist attacks of September 11, 2001. The first generation of responses consisted of two elements. Administration shills defended every jot and tittle of the policies adopted by the Bush administration. (2) The shills could not bring themselves to admit that, possibly, the policies were not well-thought through on the level of antiterrorism policy, nor that, possibly, the policies raised not insubstantial constitutional questions because the policies might be overly intrusive on constitutional values of equality and privacy. (3) The administration shills were countered--in the first generation by overly excited civil libertarians, who saw in every new Bush administration policy not merely a threat to fundamental constitutional values, but the realization of such threats, portending a descent into totalitarian tyranny, made worse by the fact that the initial steps carefully targeted only non-U.S, nationals but were justified by arguments as to which the targets' citizenship was analytically irrelevant. (4)

The second generation of scholarly responses is more temperate. It acknowledges the complexity of the constitutional questions raised by antiterrorism policy in general and by the Bush administration's policies in particular. In second-generation scholarship, the Constitution's rights provisions provide room for novel policies to deal with apparently novel threats but nonetheless cabin innovation to avoid the withering of existing civil liberties. Its structural provisions contemplate a complicated set of interactions between the President and Congress, in which the President concededly has a first-mover advantage and can set the agenda for further discussions, but in which Congress has an important role--one acknowledged by the courts--in constraining presidential initiatives, particularly those that raise the more urgent rights-based concerns. (5)

Professors Posner and Vermeule focus on two sets of arguments familiar from first-generation scholarship: that responses to emergencies inevitably ratchet down the protections a constitution provides to civil liberties and that emergencies produce bad policies because decisionmakers are dominated by fear, producing distorted judgments both about the nature of the emergency and about the efficacy of the policies offered to deal with the emergency. I agree with nearly all of what they have to say. Ratcheting down is indeed not inevitable, and the policies adopted in response to emergencies are not always ill-adapted to the problems presented. In what follows, I offer some primarily methodological quibbles with the way they try to establish those conclusions. An overview of my comments is that Posner and Vermeule are right to point out the weakness of the evidence offered to support claims about ratchets and fear, and right to ask for some general analytic reasons for thinking that emergencies produce particular kinds of policy response, but sometimes misdescribe what is wrong with the claims made about ratchets and fear.

1. RATCHETS

   In discussing the possibility that emergencies produce ratchets in public policy, Posner and Vermeule pursue two strategies. The first involves examining what actually happens in and after emergencies. Here they show that it is extremely difficult to describe the policy changes in such circumstances as "ratchets." Policy changes occur, and sometimes those changes "stick" in a ratchet-like way. Sometimes, though, the policy changes do not stick. In addition, sometimes the policy changes are in a statist direction and sometimes in a libertarian one. Finally, sometimes the policy changes seem normatively desirable and sometimes not. Posner and Vermeule's contribution here is to deescalate the discussion and to insist, sensibly, that policy changes be assessed on their merits, without clouding the discussion with reference to ratchet-like effects. (6) I have no real quarrels with their discussion of the evidence regarding ratchet-like effects. (7)

   Posner and Vermeule's second strategy is to sketch reasons for thinking that ratchet-like effects are unlikely or at least reasons for thinking that proponents of the thesis that such effects are likely have not made out a persuasive case. This strategy involves seeking to identify general social processes, not restricted to the context of emergency, that would help explain decisionmaking in and after emergencies. Posner and Vermeule take a methodologically individualist approach to social phenomena. (8) Large-scale effects--such as the adoption of policies dealing with terrorism--must be explained by identifying mechanisms by which the choices individuals make are aggregated to produce the identified effects. Posner and Vermeule argue that those who claim there is a ratchet in policies developed in response to emergencies have not identified a plausible mechanism by which the ratchet arises.

   My methodological concern is, perhaps, minor. Methodological individualism is probably the only serious candidate as the basis for social analysis. (9) But philosophers of the social sciences typically note that methodological individualism is a reductionist program. That is, methodological individualism seeks individual-level explanations, but--the philosophers say--it is neither unusual nor troublesome for the search to be unsuccessful. Philosophers of the social sciences who agree that explanations must ultimately provide individual-level explanations also agree that, quite often, perfectly decent explanations--that is, explanations that serve an analyst's immediate purposes--are not going to satisfy the rigorous demands of methodological individualism understood as a program. Virtually everyone agrees, for example, that there is something about a nation's "culture" that has something to do with the policies the nation's decisionmakers pursue and yet reducing "culture" to the individual level proves to be enormously difficult. For that reason, methodological individualists do not reject explanations that invoke culture, nor do they always fault such explanations for failing to provide microfoundations. (10) Acknowledging the difficulty of reducing culture to individuals, they note that the commitments of the social sciences as sciences are to some eventual reduction and that the cultural explanations, to the extent they are plausible, should become part of an agenda for future research.

   Yet, these philosophers of the social sciences also note what might be ineliminable problems with the proposed agenda. The shorthand for one important difficulty is the "small-n" problem--that is, once an analyst incorporates an appropriately large number of general variables into the analysis, what initially was a discussion of many instances becomes a discussion of a smaller number of cases. This problem crops up in Posner and Vermeule's article. They properly point out that claims about ratchets are claims that can be assessed only by careful empirical examination of "comparative politics." (11) They criticize those who purport to identify ratchets for "casual citation[s of] a few salient examples,"...