Deference to the executive in the United States after September 11: Congress, the courts, and the Office of Legal Counsel.

• Author: Posner, Eric A.
• Position: Thirtieth Annual Federalist Society National Student Symposium

According to the "deference thesis," legislatures, courts, and other government institutions should defer to the executive's policy decisions during national security emergencies. (1) In this Essay, I will address two criticisms of the deference thesis. The first argument, which has been developed most powerfully by Professor Stephen Holmes, is that rules dominate standards at moments of crisis. (2) An executive that is unconstrained, that is, not bound by rules, will make worse policy choices than an executive that is bound by rules. (3) This type of argument is usually made in the context of urging legislatures and courts to constrain the executive during emergencies. (4) Some commentators, however, doubt whether it is possible for legislatures and courts to constrain the executive during emergencies. (5) These doubts have led to a second argument that the executive should be bound by institutions within the executive branch such as (in the United States) the Office of Legal Counsel, (6) or through the construction of new institutions that review the executive branch's actions. (7) Both arguments criticize the deference thesis but propose different solutions. The first argument proposes that Congress and the judiciary give the executive less deference; the second proposes that officials within the executive branch give the President less deference. Thus, we can distinguish external constraints on the executive and internal constraints on the President.

Both arguments are flawed. The external constraints argument gets the normal analysis backwards: rules are better for routine, recurring situations. Although some emergencies are, in fact, routine, the type of emergency that calls for deference is not. The internal constraints argument, as normally presented, makes the fatal assumption that the President can be bound by his own agents against his own perceived interest, and relies on other questionable premises about the structure of government in the United States.

1. THE DEFERENCE THESIS

   The deference thesis states that during emergencies the legislature and judiciary should defer to the executive. (8) It assumes that the executive is controlled by the President, but to the extent that the President could be bound by agents within the executive, the deference thesis also holds that those agents should follow the President's orders, not the other way around. In normal times, the three branches of government share power. For example, if the executive believes that a new, dangerous drug has become available, but possession of the drug is not yet illegal, the executive may not act on its own to detain and prosecute those who deal and use the drug. The legislature must first enact a statute that outlaws the drug. The executive also depends on the legislature for financial appropriations and other forms of support. The executive also faces constraints from the courts. If the executive arrests drug dealers and seeks to imprison them, it must first obtain the approval of courts. The courts ensure that the executive does not go beyond the bounds of the new law, does not violate earlier-enacted laws that have not been superseded by the new law, and does not violate the Constitution.

   In emergencies, the executive often will contemplate actions that do not have clear legislative authority and might be constitutionally dubious. For example, after September 11, the U.S. government engaged in immigration sweeps, detained people without charges, used coercive interrogation, and engaged in warrantless wiretapping of American citizens. (9) Many, if not all, of these actions would have been considered violations of the law and the U.S. Constitution if they had been undertaken against normal criminal suspects the day before the attacks. After September 11, both the legislature and the courts gave the executive some deference. The legislature gave explicit authorities to the executive that it had initially lacked; (10) the courts did not block actions that they would have blocked during normal times. (11) But neither body was entirely passive. Congress objected to coercive interrogation and did not give the executive all the authorities that it requested. (12) After a slow start, the courts also resisted some of the assertions the executive made. There is some dispute about whether this resistance was meaningful and caused the executive to change policy or merely reacted to the same stimuli that caused the executive to moderate certain policies independently. (13) In any event, no one disputes that the courts gave the executive a nearly free pass over at least the first five to seven years of the conflict with al Qaeda.

   The deference thesis, then, can be strong-form or weak-form. This ambiguity has had unfortunate consequences for debates about post-September 11 legal policies. Few people believe that the courts should impose exactly the same restrictions on the executive during an emergency as during normal times. Indeed, doctrine itself instructs courts to balance the security value of a course of action and its cost to civil liberties, implying that certain actions might be legally justified to counter high-stakes threats but not to counter low-stakes threats. (14) Nor does anyone believe that the executive should be completely unconstrained.

   The debate is best understood in the context of the U.S. government's post-September 11 policies. Defenders of these policies frequently invoked the deference thesis--not so much as a way of justifying any particular policy, but as a way of insisting that the executive should be given the benefit of the doubt, at least in the short term. (15) The deference thesis rests on basic intuitions about institutional competence: that the executive can act more decisively and with greater secrecy than Congress or the courts because it is a hierarchical body and commands forces that are trained and experienced in countering security threats. The other branches lack expertise. Although they may have good ideas from time to time, and are free to volunteer them, the ability of the executive to respond to security threats would be unacceptably hampered if Congress and the courts had the power to block it to any significant degree.

   Secrecy is an important part of the argument. Policymaking depends on information, and information during emergencies often must be kept secret. Congress and the courts are by nature and tradition open bodies; if they were to act in secret, their value would be diminished. Meanwhile, the argument continues, the fear of an out-of-control executive who would engage in abuses unless it was constrained by the other branches is exaggerated. The President has strong electoral and other political incentives to act in the public interest (at least, in the United States). Even if the executive can conceal various "inputs" into counterterrorism policy, it cannot conceal the "output"--the existence, or not, of terrorist attacks that kill civilians.

   Thus, it was possible for defenders of the Bush Administration's counterterrorism policies to express discomfort with certain policy choices, while arguing nonetheless that Congress and the courts should not try to block executive policymaking for the duration of the emergency--at least not as a matter of presumption. Critics of the Bush Administration argued that deference was not warranted--or at least not more than a limited amount of deference was warranted, although again these subtleties often were lost in the debate--for a variety of reasons. I now turn to these arguments.

2. EXTERNAL CONSTRAINTS: THE PROTOCOL ANALOGY

   1. Medical Protocols

      In an article published a few years ago, Professor Holmes uses the arresting image of the medical protocol as a device for criticizing the deference thesis--or, more broadly, the thesis that the executive should be "unconstrained" during emergencies. Holmes describes his own experience in an emergency room, where his daughter had been brought with a serious injury:

      At a crucial moment, two nurses rushed into her hospital room to prepare for a transfusion. One clutched a plastic pouch of blood and the other held aloft my daughter's medical chart. The first recited the words on the bag, "Type A blood," and the other read aloud from the file, "Alexa Holmes, Type A blood." They then proceeded, following a prepared and carefully rehearsed script to switch props and roles, the first nurse reading from the dossier, "Alexa Holmes, Type A blood," and the second reading from the bag, "Type A blood." (16) To the layman, the repetitive actions of the nurses seem senseless. Why are they repeating themselves when the patient might die unless she receives the blood transfusion immediately? Surely, the nurses should depart from the script rather than follow it in a time of extreme medical urgency. Yet the protocol makes good sense. Experience has taught medical personnel that basic errors--the transfusion of the wrong blood--occur frequently, and that they can be avoided through the use of simple protocols. Although following the protocol uses valuable time, in practice the increased risk to the patient as a result of the loss of time is less than the risk caused by the errors that protocols are designed to prevent. (17)

      The larger and more striking point of the example is that, even during emergencies, when the stakes are high and time is of the essence, agents should follow rules rather than improvise. In this way, agents should be constrained. (18) This argument has potentially radical implications. Recall that the conventional objection to deference is that the risk of executive abuse exceeds the benefits of giving the executive a free hand to counter al Qaeda. Professor Holmes argues--although at times he hedges--that in fact the benefits of giving the President a free hand are zero: A constrained executive, like a constrained medical technician, is more effective than an unconstrained executive...