Hyperdepoliticization

By: Edward Rubin*

The effort to insulate public administration from the tempestuous intensity of political controversy is a dominant theme in the history of American government.  At the federal level, it is older than the administrative state itself, having been put in issue by the transition from the Federalists to the Jeffersonian Democrats in 1800[1]—the events that later led to Marbury v. Madison[2]—and by the Jacksonian Democrats’ introduction of the spoils system in 1828.[3]  When the federal government instituted what is generally regarded as its first true regulatory program—the control of railroad shipping rates—it quickly moved to insulate the agency responsible for administering the program by rendering it independent of presidential control.[4]  Independent agencies—arguably one of America’s only three truly original governmental inventions, together with judicial review and national parks—proliferated in the Progressive Era and have since been joined by many others, adding up to about seventy in all.[5]  Their constitutionality was confirmed by the Supreme Court in its 1935 decision Humphrey’s Executor v. United States[6] and is now regarded as a settled matter, despite a certain amount of unitary executive grousing.[7]  This insulation can be referred to by the awkward name of depoliticization.[8]

When we speak of depoliticizing an administrative agency, we mean insulating it from direct presidential control.  The resulting distinction between executive and independent agencies, while looming large from the President’s perspective,[9] is essentially invisible to Congress.  All the mechanisms of control that Congress exercises over administrative agencies are equally available whether the agency is executive or independent.[10]  Amendments of authorizing statutes, adjustments of budgetary allotments, oversight hearings, casework, and staff level contacts can all be deployed against any agency with equal force.  Yet, Congress is hardly apolitical.  The President, however beholden to interest groups, at least thinks of himself, and describes himself, as representing all the people.  Members of Congress, with their narrower constituencies, are often viewed as motivated by exclusively political considerations.[11]

This observation suggests the possibility that a further level of insulation could be effectuated by shielding a particular administrative agency from congressional as well as presidential control.  Doing so could be described by the even more unwieldy term of hyperdepoliticization.  Unlike the ordinary depoliticization process, hyperdepoliticization would remove all direct political influences and place the agency in a structural position resembling a court, rather than an ordinary independent agency.  In considering this possibility, a variety of empirical questions naturally present themselves.  Would it truly shield the agency from partisan politics; would it improve the quality of agency decision making?  Would it render the agency unaccountable in some meaningful and measurable way?[12]

Hyperdepoliticization also presents some more theoretical issues that are absent from garden variety depoliticization.  In fact, the issues it raises are so theoretical that they belong as much to the field of philosophy as they do to law or political science.  When Congress insulates an agency from presidential control, it is simply distributing authority among its various subordinates in a particular way.  This is, of course, what any institution-wielding hierarchical authority will do and thus raises no theoretical complexities.  Those who favor a unitary executive argue that Congress may not establish insulation of this sort; that is, all authority to “take Care that the Laws be faithfully executed”[13] must be initially granted to the President.[14]  But this argument is based on the text and history of the U.S. Constitution.  No one has argued that there is any theoretical difficulty in limiting the way that a superior grants authority to its various subordinates.

Congress, however, being a legislature, has no superior in a democratic system.  There are constitutional limits to its authority, to be sure, and the courts are superior to Congress in defining and enforcing those limits.[15]  But within those limits—that is, when it is acting constitutionally—Congress is the supreme authority.  Thus, the only institution that possesses the authority to insulate an agency from Congress is Congress.  To insulate an agency from all political pressure by elected officials, therefore, Congress must not only place limits on a subordinate (the President, in the agency context) but also on itself.[16]  A technique of this sort raises legal theory problems because we have no established methodology for doing so.  Legislatures act primarily by passing statutes; every statute can be amended by a subsequent statute enacted in exactly the same manner as the original statute.  The technique also creates political theory problems because the legislature is the people’s primary representative in a democratic system.  If a legislature places limits on itself, it is, in effect, binding future legislatures, which means that it is disabling the people at that future time from implementing their preferred policies.  That appears to violate the basic premise of democracy.  These legal and political complexities, moreover, are not specific to the United States.  They are not even specific to the standard type of democratic polity where a legislature represents the people.  If the people’s only representative were a single executive officer—a structure that no democratic nation has adopted but that can be found in local government—the same problems would appear because that executive could only hyperdepoliticize its subordinates by placing limits on itself.[17]

These legal and political problems with hyperdepoliticization lead directly to the philosophical problem: how precisely does a conscious actor place limits on its future conduct?  Limits are regarded as a form of imperative order and such orders are typically imposed on a subordinate—that is, someone to whom the actor is authorized to give orders.  But a conscious actor is generally on the same level at both a given time and a subsequent time.  Why should the actor at the current point in time (“t₂”)regard itself as the subordinate of its earlier self and obey a decision that it reached at an earlier point in time (“t₁”)?  This is a major question in modern analytic philosophy,[18] and it is directly applicable to a legislature’s efforts to place limits on its future conduct.  Typically, philosophical problems only become issues in legal and political theory by analogy.  It may be illuminating to think about questions of free will, the nature of causality, or the reliability of induction when analyzing the work of a legislature, but it is also possible to ignore these complexities and deal only with political practicalities.  The philosophical problem involved in a conscious actor binding its future conduct is immediately presented in the case of hyperdepoliticization.  It is essentially the same problem in the political context as it is in the philosophic one and cannot be ignored.

Perhaps the best-known philosophic discussion of a conscious actor’s self-commitment is John Elster’s.[19]  It is of particular interest here because Elster begins with the rationality of individual actors then applies his conclusions to political settings at considerable length.[20]  The first Part of this Essay explores the issue of hyperdepoliticization by analyzing Elster’s discussion of self-commitment.  On the basis of the themes developed through that exploration, the second Part then discusses specific instances of hyperdepoliticization by Congress.

I.  Self-Commitment: Ulysses, Jon Elster, and the Power of Principle

A.     Elster’s Theory of the Constitution

Elster uses a famous incident from Homer’s The Odyssey as an image of a rational actor’s effort to commit himself to a future course of conduct.[21]  Book XII ofThe Odyssey recounts the passage of Ulysses and his crew by the Sirens’ Island.[22]  Circe warned Ulysses that anyone who hears their song will be irresistibly drawn to them and killed, so Ulysses instructs his sailors to stop up their ears with wax and then says to them, as Elster quotes, “You must bind me hard and fast, so that I cannot stir from the spot where you will stand me . . . and if I beg you to release me, you must tighten and add to my bonds.”[23]  Elster sees Ulysses’s action as an example of imperfect, or partial, rationality.[24]  “Ulysses was not fully rational,” he writes, “for a rational creature would not have to resort to this device.”[25]  But he was not entirely irrational either, Elster continues, because he was able to achieve by indirect means what a rational creature could have achieved directly.[26]  “His predicament—being weak and knowing it—points to the need for a theory of imperfect rationality,” Elster concludes.[27]

This discussion seems to be a promising approach to the issue of hyperdepoliticization.  The idea of a rational actor committing himself to a course of action that he then pursues, without any external force that can require his obedience, is directly applicable to the sort of discipline that a legislature must follow in order to hyperdepoliticize an administrative agency.  Elster recognizes this application, and in a subsequent work that also employs the image of Ulysses and the Sirens, he analyzes and discusses it at length.[28]  But the application that he chooses does not involve the legislature at all; rather, he argues that the process of self-commitment—of tying oneself to the mast—is analogous to the establishment of a constitution that binds subsequent political action.[29]  A constitution, Elster argues, is a means by which the political system binds itself to desirable policies so that it can resist the temptation to abandon or compromise those policies in times of crisis.[30]

If one is interested in the legislature’s use of self-commitment to achieve hyperdepoliticization, Elster’s decision to analogize his rational actor to the political system generally, rather than the legislature, is unfortunate.  This may seem to be a somewhat parochial concern however; clearly, the analogy Elster chose enables him to pursue much larger questions than a single and somewhat unusual legislative strategy.  The problem is that Elster’s analogy is simply incorrect.  It reveals a misunderstanding of political constitutions and possibly a misunderstanding of his own theory as well.

The basic problem is that the whole notion of self-commitment depends upon the existence of an identifiable self.  Defining the self is a complex enterprise that has occupied much of twentieth century philosophy,[31] but we can greatly simplify the problem for present purposes.  If we consider the issue of self-commitment, the concept of a self must involve an entity that is capable of making decisions regarding its own future course of action.  Any competent human being fits this definition.  Groups made up of separate human beings can readily qualify for this definition of a self as well; all that is needed is three rules: one identifying the group, a second specifying a process by which the group can reach a decision, and a third indicating how such decisions are to be announced.  Obviously, every functioning legislature meets this definition of a self because it needs these same three rules to carry out its ordinary functions.  None of the complexities that occur when applying the concept of intent to a collective body like a legislature apply to the question of treating the legislature as a decision maker.[32]  An external observer can only know that an individual has reached a decision when the individual takes some sort of action, and this action frequently consists of spoken or written words.  A collective body can act in exactly the same way; either some individual is designated to speak for the group or a certain proportion of the group’s members indicate assent to a document that has been presented to them.  Determining the intent of the legislature may be a thorny problem, but there is rarely any ambiguity about whether the legislature has in fact made a decision such as the enactment of a statute.

When Elster discusses constitution making in Ulysses Unbound, however, he implicitly incorporates the idea that a society is committing itself to a course of action.[33]  The problem here is that society is not a self, as just defined, but an academic or rhetorical abstraction.  A society cannot make decisions and therefore cannot bind itself in the required manner; only individuals or institutions within a society can do so.  This conceptual inaccuracy on Elster’s part is mystifying.  Elster is acutely aware of the complexities involved in analogizing collective entities to individuals, and refers to this problem explicitly in Ulysses Unbound.[34]  One section of his chapter about constitutions is entitled “Disanalogies with Individual Precommitment.”[35]  Even more strikingly, the first of the three essays that comprise his volume entitled Ulysses and the Sirens (Ulysses and the Sirens itself being the second essay) is a critique of analogies between intentional and functional explanations.[36]  “[I]n spite of certain superficial analogies between the social and biological sciences,” he writes, “there are fundamental differences that make it unlikely that either can have much to learn from the other.”[37]  Yet, Elster’s discussion of constitutions often rests on a similarly defective analogy of an individual and the political system or society in general.  The disanalogies he notes in Ulysses Unbound between the individual and constitutional case is that a constitution may bind “others” or it may not be binding at all.[38]  He does not address the basic problem that a constitution typically does not bind the constitution maker.

With respect to the American example, to which Elster devotes considerable attention in his essay, the important point is that the U.S. Constitution was drafted by a specially convened Convention[39] and then ratified by a specially designed voting procedure in the thirteen states.[40]  The decision-making self in this case was the Convention or, more problematically, the Convention and the eligible voters in each state.  Having drafted or enacted the Constitution, these entities then passed out of existence.  Thus, the U.S.  Constitution is simply not a case of self-commitment; rather, it is a case of a superior giving a written order to a subordinate,[41] and thus it is structurally identical to an ordinary statute in which the legislature places limits on a subordinate such as an administrative agency.  For constitution making to be an act of self-commitment, the Constitutional Convention would have had to place limits on what it could do in the future.  The Convention (or the Convention and the ratification votes in the states) did not do this; it simply passed out of existence.  In this sense, the U.S. Constitution resembles the Spartan Constitution after the death of Lycurgus, the self (an individual in this case) who drafted it.[42]

The Constitution itself provides a mechanism—a new constitutional convention—by which the same process that led to its creation and adoption can be revived.[43]  Does this make the initial drafting or adoption an act of self-commitment?  The answer is no for two reasons.  First, the Constitution, significantly, places no limits on what the new convention can do, aside from maintaining the equal representation of states in the Senate,[44] which is hardly the self-commitment to which Elster refers.  Second, and more significantly, the definition of self precludes the drafting and adoption of the Constitution from being an act of self-commitment.  Because the self (that is, the decision-making entity that drafted the Constitution) passed out of existence after the Constitution was adopted, a new constitutional convention, organized according to its provisions, would be a different self.[45]  If the Convention engaged in an act of self-commitment, it did so by disbanding, not by propounding the constitutional provisions.  The same can be said, of course, of the ordinary amendment process that Article V provides.  First, there are no substantive limits on the process itself, and second, the procedures specified are limits on the legislature that was created under the Constitution and on the states that became subject to it, not on the entity that drafted the Constitution in the first place.[46]

Can the willingness of U.S. government officials, or the people generally, to be bound by the Constitution be regarded as an act of self-commitment?  The Constitution, after all, is just a piece of paper and has no force unless people are willing to treat it as worthy of obedience.  But this is not what Elster means by self-commitment and probably not what most people mean by it.  Elster’s definition is that the actor himself makes a decision at t₁ that will bind him at t₂, not that he decides at t₂ to obey some external entity that existed at t₁ but has now passed out of existence.[47]  One can define self-commitment more broadly if one wishes, and perhaps even so broadly that it includes any act of obedience, but this would not distinguish the Constitution from any other governmental rule.  If obeying the Constitution is a form of self-commitment, then so is obeying a statutory rule or administrative regulation.  It is true, as the existentialists remind us, that we are always free to disobey, but this inspiring insight tells us very little about the purpose of national constitutions.[48]

Suppose, by way of an analogy, that the owner and chief executive officer of a company decides to relinquish all her ownership and control one year in the future.  Because she is aware that she has made some bad decisions by acting on her own, she decides to establish a managing committee to run the company and place all the stock she owns in escrow.  However, because she is aware that she built the company acting on her own, she also decides that if the committee fails to show a profit for three successive years, it will be required to appoint a single executive officer to replace it, with exactly the same authority as she originally possessed, and that this officer will be given all her stock.  We would certainly be willing to describe her decision to relinquish control and ownership of the company as an act of self-commitment.  In doing so, she might bolster her resolve with the mechanisms Elster discusses,[49] such as entering into a binding contract with the members of the future managing committee; telling her family and friends she will resign, so that her family will be angry and her friends contemptuous if she reneges; or creating rewards for herself, such as arranging for that cruise to Tahiti that she has always wanted to take.  But the document by which she organizes the company after she relinquishes control cannot be regarded as an act of self-commitment; it binds others, not herself.  Nor does it become an act of self-commitment because it provides for the possible re-establishment of her position.  She will be dead or in Tahiti by that time, and her successors will make the decision.

Going beyond the United States to either other nations or hypothetical situations, it appears that a constitution or constitutional amendment approaches true self-commitment as the process of adoption involves fewer and fewer actors other than the legislature.  Once the legislature is empowered to enact a new constitution or amend the existing one without the participation of any other decision maker, then it is genuinely binding itself.[50]  It is declaring that it will not alter the constitutional provisions in the future, although it has the authority to do so.  Thus, at t₂, consistent with Elster’s definition, the legislature will willingly change ordinary statutes but refuse to change the ones it has identified as constitutional.  When the legislature must act in concert with other institutions to amend the constitutional provisions, it is no longer exhibiting the same level of restraint, and it is no longer acting on the basis of its own previous commitment.

In the situation where the legislature can act by itself, constitutional provisions, however defined, merge with ordinary legislation as a structural matter, and differ only on the basis of the enacting legislature’s declared intent to bind the future.  We have thus arrived at the issue of hyperdepoliticization.  In the American case, where Congress cannot enact constitutional provisions on its own, it can bind itself only by enacting a statute.  Such a statute can be a true self-commitment in the sense that Congress at t₂ is the same institution as Congress at t_1, with the same authority and the same freedom of action.  But the creation of a constitution by a body that then passes out of existence is not an act of self-commitment at all, and the only thing that Elster’s discussion of constitution making tells us is that he misunderstands the political process.

There have been a number of subsequent efforts to rescue the idea of a constitution as the self-commitment of a decision-making entity.  Stephen Holmes, for example, traces the issue back to Bodin’s theory of sovereignty, which insists that each polity must possess a single, unconstrained ruler that, within its jurisdiction, is subject to no earthly power.[51]  Such an authority, Bodin says, exercises all the powers of governance, but one of those must be the power to bind itself—to make definitive commitments.[52]  This is, of course, a familiar concept, which economists generally describe as the power to make credible commitments.[53]  Non-lawyers are often surprised by the notion that a full panoply of legal rights includes the right to be sued; but this right, which is denied to slaves or legal infants, grants the rights holder the power of commitment.  Madison’s concept of a constitution, Holmes argues, is a commitment in this sense, a voluntary constraining of sovereign power through the creation of rights that others may enforce against that power.[54]

Holmes provides a useful account of late eighteenth-century thought, but it only solves an eighteenth-century problem about the way an unconstrained sovereign, conceived as a necessity of governance, could be constrained.[55]  Using it as a theory of constitution making depends on the rather abstract notion that “the people” are sovereign.  In the American case, it asserts that this sovereign entity existed, or was somehow expressed, in the Constitutional Convention and the ratification votes in nine of the thirteen states.[56]  This may or may not be convincing, but it does not really rescue the self-commitment concept.  The “sovereign people,” having expressed itself through a decision-making institution, has not passed out of existence, but it is now dormant.[57]  To make a decision which might potentially be bound by the first one, this sovereign—the people—would need to generate a new entity through which it can act.[58]  Even if one accepts all the political mysticism involved in this conception, it seems clear that no such entity exists in our governmental system.  What does exist is the government this sovereign created, a set of institutions that are not, either individually or collectively, the sovereign.  In fact, if we interpret the American ratification process as the sovereign, then it seems that our sovereign went into retirement without making any commitments or placing any limits on itself if it chose to return, aside from that business of equal state representation in the Senate.

Jed Rubenfeld proposes a related idea that attempts to replace the uncertain and somewhat old-fashioned idea of sovereignty.[59]  Like persons, he argues, whose individuality only makes sense when considered as something that exists over a period of time, a people (that is, a group of persons who govern themselves) necessarily exists over time, not at a single moment.[60]  Such a temporally extended entity, whether an individual or a nation, does not make commitments through the ephemeral act of speaking, which is the way we customarily describe the decision-making process, but through the durable act of writing.[61]  A written constitution can thus be regarded, according to Rubenfeld, as a commitment made by the people as a temporally extended entity.[62]  It is obeyed because obeying the durable or written commitments that one makes is the way such an entity—either a person or a nation—expresses its self-assertion of its own identity, and thus establishes its freedom.  Those commitments then determine how particular constitutional cases are to be decided.[63]

The difficulty with this theory for present purposes is its ethereal circularity.  Rubenfeld’s notion of a people is not fatally vague—he answers the obvious objections effectively[64]—but it only exists, or only acts, by making commitments.  The only way we know what those commitments are, apparently, is that the people, as an entity, have made them.  The circularity occurs because “the people,” however meaningful this notion may be as an explanation of national identity, is not a decision-making mechanism in the same sense as the institutions supposedly bound by its commitments—the chief executive, the legislature, or the judiciary.  Rather, it is a self-contained process of identity formation that acts only by asserting that identity.  Governmental institutions may be guided by this process, but they are not bound by it; indeed, they are not only free to disobey, but they can destroy or dissolve the self-committing entity by doing so.  Perhaps Rubenfeld has resolved the difficulty that Elster and Holmes confront when speaking about self-commitment where there are, in fact, two different selves, but only by positing a single self that functions as cultural expression, a “brooding omnipresence in the sky,”[65]  not as a government decision maker.  If a person had a mystical experience that changed her sense of self, but that only affected her decisions because she interpreted herself in a different way and felt free to make different choices which denied that changed sense of self, we would generally not describe that as a case of self-commitment.

B.     Elster’s Analysis of Rational Decision Making

Allowing for the fact that Elster analogizes Ulysses’s self-commitment to the wrong political situation, we might nonetheless look to his well-known discussion once we correct the error and recognize that the problem he describes applies to statutory hyperdepoliticization and not to constitution making.  Unfortunately, most of his discussion is unhelpful because his answer also misunderstands the problem he presents.  In essence, he spends most of his time fighting his own hypothetical, looking for empirical ways out of a philosophical dilemma.  The task of correcting the conceptual difficulties thus created could be left for philosophers to resolve were it not for the fact that the particular reasons why Elster’s analysis is philosophically flawed are immediately implicated in applying that analysis to the decision making of a collective body like a legislature.

We can begin by being precise about the use of the myth as an analogy.  Ulysses has himself bound to the mast so that he will not respond to the Sirens when he hears their song.  The song thus represents a current situation (that is, the situation at t₂) and the bonds that hold him represent some sort of commitment that disables him from responding to that situation.  His instruction to his crew represents an earlier decision—that is, a decision taken at t_1.  The question is whether a person actually can effect such a restriction of his current decision making, and why he would do so.  Now suppose that the crew members had decided to tie Ulysses to the mast on their own.  We would no longer have a philosophic problem regarding the nature of Ulysses’s decision making.  Being subject to an external constraint, his failure to respond to the Sirens’ song would tell us nothing about his own decision making as a rational being.

Elster does not devote any time to discussing this situation, of course,[66] but he spends a great deal of time discussing a variety of situations at t₂ where the actor’s decision making is impaired and the actor is aware of the impairment.[67]  Self-commitment in this context can thus be described as “Peter sober” binding “Peter drunk.”[68]  In most of the situations discussed by Elster, the impairment does not result from imbibing a substance but from weakness of the will; that is, the actor knows, in the current situation (i.e., t₂) that yielding to his current inclinations will lead to undesirable consequences, but his physical or psychological desires incline him to give those consequences less weight than he gave them at t₁, when the temptation was not present.[69]  Smoking, eating fattening foods, and deciding to take mind-altering substances, all of which Elster discusses at length, are examples of such weakness of the will, and Elster assimilates future discounting (favoring the present over the future) to this weakness as well.[70]

These considerations are of undoubted psychological interest, and they certainly apply to Ulysses, whose will was overborne by a supernatural force.  But they sidestep, and in fact obscure, the philosophical problem—namely, the problem with the nature of rationality itself.  That problem arises when the actor at t_2, although fully aware of having decided to make a different decision at t_1, is not aware of any impairment in his current rational capacities.  That is, he is convinced that he is fully rational at t₂ and that what he would have perceived as a temptation at t₁ is now a valid ground for action.

To clarify this point, suppose the actor at t₂ has been drinking and now is fairly certain he is drunk.  He might well decide that he is in no condition to make decisions, and, confronted with a situation that requires him to choose between his current inclination and a course of action he had previously reached, he might well decide to favor his previously chosen course of action. But this is not truly a case of fulfilling a precommitment.  Rather, it is a case of recognizing a temporary cognitive impairment and deferring to a decision reached at a time when one was not impaired.  In fact, it is more similar to an actor’s decision to follow the advice of a more knowledgeable person—like taking medicine prescribed by a doctor.  Alternatively, we can say that self-recognized impairment is not a case of self-commitment because the actor is no longer a single self.  The person he is now, at t₂, is not quite the same person that he was at t₁.  This is not because his personality changed between the two times—that is a situation that must be accommodated within the framework of self-commitment over time—but rather because he is impaired and recognizes that he is impaired_.  The situation is recognized in ordinary usage when we say that someone, being impaired, is “not himself.”

All discussion of cases of impairment at t₂—not only the fairly obvious case of being drunk, but the related cases of being subject to a temptation that undermines one’s will—suffer from this same limitation.  They pose only a psychological problem, not a philosophical one.  Consequently, the techniques that Elster discusses for ensuring that the t₁ commitment is followed at t₂ belong to the realm of psychology.  They are stratagems by which the actor at t₁ can place constraints on itself at t₂ because the t₁ actor is convinced that it is in a better position to decide.  Among those that Elster discusses are throwing away the key to the place where an addictive substance is stored, giving away the key to a trustworthy person, imposing costs on oneself at t₂ for disobedience, and creating rewards for oneself at t₂ for obedience.[71]  The problem is that these stratagems will only be legitimate from the perspective of the decision maker at t₂ if that decision maker recognizes that it is impaired.  Otherwise, they will seem like nothing more than stratagems—ill-considered efforts by oneself to bind one’s future actions at a time when one’s understanding of those future actions was inadequate.

The true philosophic question arises when the actor is not aware that he is suffering from any impairment at t₂.  That is, he does not regard himself as “Peter drunk” or otherwise functioning at a lower cognitive level than usual, nor does he regard himself as subject to a temptation that is undermining his will.  Rather, he sees himself as mentally equal at t₂ to what he was at t₁.[72]  How can the actor at t₁ bind himself at t₂ under such circumstances?  What mechanisms can he use that will overcome the phenomenological experience of being here now, at t₂, and fully capable of making an autonomous decision?  This is the way Gregory Kavka sets up the philosophic problem in his well-known toxin hypothetical.[73]  An eccentric billionaire offers the actor a million dollars now if he will intend to drink a repulsive but not dangerous toxin in the future.[74]  When the future comes, however, the actor has already received the money, so there is no reason for him to fulfill his promise and drink the toxin.[75]  Therefore, the actor cannot honestly intend to drink the toxin, and cannot accept the million dollars, even though he would gladly undergo the relatively minor discomfort that the toxin causes in exchange for the money.[76]  One may or may not find this convoluted hypothetical illuminating; for present purposes, the important point is that the actor’s dilemma is caused by his rationality, not his irrationality.  He is perfectly in control of his faculties and his will at t₂, which is precisely the reason he has trouble intending to drink the toxin at t₁.  This is the philosophic problem: why should a rational actor at t₂ be bound by its decision at t₁—not how a rational actor at t₁ binds an irrational actor at t₂.

To clarify this point, consider the dieting situation that Elster discusses.[77]  The actor at t₁ has decided that she wants to lose weight and will stop eating dessert.[78]  At t₂, however, she finds herself at a restaurant famous for its desserts and wants to order chocolate cake.[79]  She might perceive herself as subject to an overwhelming temptation to eat the cake.  That means that she recognizes that her decision-making process is impaired, and she then has a reason to favor her decision at t₁ to diet over her current desire to eat the cake.  Whether she will obey that decision, which she now regards as superior (ethically or cognitively), or whether she will succumb to her visceral desire to eat the cake because her will (that is, her recognition that her t₁ decision is superior) has been overcome is a matter of psychology.  Alternatively, however, she might perceive herself at the restaurant as fully in control and decide to eat the cake because she now realizes that gustatory pleasure is more important to her than weight loss—that is, that she will have a better life if she eats dessert than she will if she achieves some particular body weight.  In that case, when she has the same view of her decision-making ability at t₂ as she had at t₁, is there any argument that could convince her to favor her decision at t₁ and thus maintain her self-commitment?  This is the philosophic problem.

One reason Elster fails to confront this problem is that all his examples involve a t₁ decision that strikes him, and will generally strike the reader, as sagacious, and a t₂ decision that strikes him and the reader as impaired.[80]  That is unquestionably the situation in Ulysses’s case, as confirmed by his statements at t₃, when he is past the Sirens.[81]  Consider, however, the reverse.[82]  A young man, age eighteen, has grown up in a desperately unhappy home where his parents fought with each other all the time and mistreated their children.  He commits himself to never marrying or having children of his own, and enforces this commitment by becoming a Catholic priest, or getting a vasectomy, or convincing his parents to write a will that disinherits him if he gets married, or telling his friends to shun him if he does so.  As he matures, however, he comes into contact with happily married couples, discovers he likes children, perceives himself as a kind, loving person, and has this assessment confirmed by his sexual partners, his friends, and, of course, his therapist.  Now, at t₂  (age twenty-eight), his t₁ decision no longer seems like a commitment that he wants to keep.  Perhaps the stratagems he adopted at t₁ will be effective, but he will feel that they are not legitimate and will try to reverse them.  He may leave the priesthood, try to reverse his vasectomy, relinquish his inheritance, and make new friends.  In other words, he will regard his t₁ decision as impaired, and the observer, in this case, is likely to agree.  This illustrates the problem Elster avoids: under circumstances such as these, what sort of commitment would a t₂ actor be willing to obey?

The reason to distinguish between philosophical and psychological versions of the precommitment problem in discussing hyperdepoliticization is that only the philosophical version of the problem can be of much value in this context.  Congress, as a collective entity, cannot be viewed as Peter drunk, an impaired decision maker that should not take action or that should only take action with another actor’s approval.  Legally, the legislature is the primary policy maker of a modern democratic state.  If its members regard it as impaired, they are abandoning their basic legal responsibility—the policy-making function which they were elected to perform.  If outsiders regard it as impaired, they would, in essence, be rejecting democratic government.  That is in fact the attitude that the military has adopted in various nations when it carries out a coup,[83] but it does not seem like one that ought to be encouraged.  Politically, members of the legislature who concluded that the legislature is impaired would be denying their constituents representation, and outsiders would be counseling fatalistic inaction.

In other words, when the legislature is analogized to a rationally acting individual in order to gain insight into the problem of self-commitment, the analogy is useful only in its truly philosophical form, not in the quasi-psychological form that Elster discusses.  It simply does not make sense to say that a modern legislature is drunk, suffers from weakness of the will, or is subject to any of the other psychological conditions that counsel deference at t₂ to a previous decision.[84]  The legal and political considerations that come into play when the decision maker is the legislature of a democratic regime direct us away from such analogies (however tempting they may be when the legislature is controlled by the opposing party) and toward the truly philosophical question of whether a fully competent decision maker should ever regard itself as bound by a previous decision, even a decision that explicitly attempts to bind the future.

In Ulysses Unbound, where Elster presents his application of the Sirens issue to the political system, he offers one possible way in which an entity like Congress might perceive itself as impaired.[85]  He argues that the Constitution serves as a precommitment strategy that will enable us to maintain our principles at times of political crisis such as war, when there is a temptation to compromise our civil liberties.[86]  Lawrence Tribe makes the same argument in the introduction to his constitutional law treatise,[87] and while he does not cite Elster, he relies heavily on a psychology experiment that is featured heavily in Elster’s work as well.[88]  There is so much wrong with this argument that I will need a separate essay to address it.  For the present, two observations will suffice. First, individuals are often impaired, in measurable ways, when subject to stress—a physiological reaction that brain researchers are beginning to understand.[89]  An analogy between this process and the behavior of an institution like a legislature is obviously highly speculative and unlikely to yield usable results.  Second, neither a legislature, through any institutional expression, nor its individual members are likely to concede that the legislature’s decision-making process is impaired in times of crisis.  The more common assertion is that such times require action, and that the institution, far from deferring to some previous commitment, should fulfill its crucial role by responding to the crisis.  Members of the legislature and external observers may conclude that a particular period of crisis is a poor time to enact a particular statute—that they should not spend time revising the criminal code during a war or that they should not restructure an emergency response agency until the particular emergency that agency is dealing with has passed.  But this simply asserts that there are better or worse times to make particular decisions, not that the decision maker itself is impaired in some essential way.

C.     The Real Story of Ulysses

To summarize, Elster’s discussion of Ulysses and the Sirens turns out to be of limited value in analyzing the problem of a decision maker’s precommitment because he conflates the philosophical problem that is directly implicated for any conscious decision maker, individual or institutional, with psychological problems applicable only to individuals.  If we want to focus on the philosophical problem we must look to other sources.  One possible source is a famous incident in Greek mythology, the time when Ulysses binds himself to the mast in order to resist the Sirens’ song.  This may seem to be a less-than-promising approach, because the foregoing section has just argued that Elster’s use of this particular incident provides little guidance in addressing the philosophical problem.  But we should not dismiss the myth so quickly.  In addition to misunderstanding the philosophical problem, Elster misunderstands The Odyssey.  If we look at the entire poem and not, as he does, at the incident with the Sirens in isolation, we can in fact find useful guidance on the self-commitment problem.

The hero of the poem is Ulysses, and to understand the poem we must understand its hero’s character.  We have a natural tendency to view Ulysses as a modern man, given his isolation and the cleverness he displays throughout the poem’s action.  This tendency is certainly encouraged by the famous Homeric epithet used to describe him, which is variously translated as a man “of twists and turns,”[90] “of many ways,”[91] “for Wisdom’s various arts renown’d,”[92] “resourceful,”[93]“ready at need,”[94] or “who was never at a loss.”[95]  Elster subscribes to this view.  His entire account emphasizes Ulysses’s autonomy—his ability to make decisions freely and implement them effectively.[96]

Elster’s interest in rationality supports this interpretation, which serves as the starting point for his entire inquiry.  A rational decision is necessarily an autonomous one, and Elster is interested in the reasons why it might be rational for a decision maker at t₁ to restrict his own decision at t₂.[97]  It is, moreover, the way modern people like heroes.  J.B. Schneewind identifies autonomy as an essential element of modernity and traces its emergence to the rejection of the tradition-bound idea that morality is grounded on obedience to some higher power or authority.[98]  Kant, whom Schneewind treats as the culmination of this process, explicitly identifies autonomy as the basis of self-legislating a categorical imperative.[99]  Kant says: “Autonomy of the will is the property that the will has of being a law to itself . . . .  The principle of autonomy is this: Always choose in such a way that in the same volition the maxims of the choice are at the same time present as universal law.”[100]

Contemporary and inspiring as this may be, it is not a convincing interpretation of The Odyssey.  While Ulysses displays a considerable amount of resourcefulness, it would be more accurate to describe him as the plaything of the gods than as a paragon of autonomy.  When the epic begins, Calypso has been holding him on her island as her boy toy for seven years, and his only recourse is pathetic sorrow: “With streaming eyes in briny torrents drown’d, And inly pining for his native shore.”[101]  He has literally been blown back and forth across the seas by Poseidon, who will shipwreck him again after he leaves Calypso’s island.  He escapes from Calypso through the intervention of the gods,[102] not his own efforts, just as he had Athena or Circe’s help in escaping from his other perils, and he will need Athena’s help once more in order to regain his kingdom from Penelope’s rebellious suitors.

In fact, the problem with Elster’s interpretation of Ulysses runs somewhat deeper and points toward a potential solution to the self-commitment problem.  Ulysses is not simply subject to the desires or caprices of the gods; he is respectful and obedient to them.  The Odyssey, like The Iliad, begins with an invocation to the Muse that states the theme of the entire poem.[103]  It refers to Ulysses’s vaunted cleverness, but then continues with one more feature of his personality, described (in the Rieu translation Elster uses) as follows: “But he failed to save [his] comrades, in spite of all his efforts.  It was their own sin that brought them to their doom, for in their folly they devoured the oxen of Hyperion the Sun, and the god saw to it that they should never return.”[104]  This is the only incident mentioned in the invocation, and its importance is further emphasized because it explains why Ulysses failed to save his crew, a serious mark against him in a tribal society where a leader was expected to protect his followers.[105]  It reveals that Ulysses was not only a clever man but a devout man, a man who was distinguished from others by his obedience to the gods.[106]

This quality of obedience plays an important role in the incident with the Sirens.  Elster presents Ulysses’s decision to have himself bound to the mast, instead of simply stopping up his ears with wax like his crew members, as a pure act of self-commitment on Ulysses’s part.[107]  He begins by quoting Ulysses’s instructions to his men: “You must bind me hard and fast, so that I cannot stir from the spot where you will stand me . . . and if I beg you to release me, you must tighten and add to my bonds.”[108]  In fact, the idea of binding himself to the mast is not his idea, but has been suggested to him by the goddess Circe.[109]  When she warns him, at the beginning of Book XII, about the dangers he is going to face in the next stage of his journey, she says (again in Rieu’s translation):

soften some beeswax and plug [the crew members’] ears with it; but if you wish to listen yourself, make them bind you hand and foot and on board and stand you up on the step of the mast . . . .  But if you start begging the men to release you, they must add to the bonds that already hold you fast.”[110]

To be sure, Circe does not order Ulysses to bind himself to the mast, but she certainly suggests this stratagem, and she tells him exactly how to do it.  He not only follows her suggestion but instructs his crew in words very similar to Circe’s.[111]  Thus, the story of the Sirens illustrates Ulysses’s devoutness, his willingness to listen to the gods, rather than his autonomy and resourcefulness.  To be sure, once he hears the Sirens’ song, he is no longer in control of himself, and it is his stratagem, the bonds, that restrain him.  But he has adopted the stratagem at the suggestion of a god.  The point is not that Ulysses is such a clever person that he thinks of binding himself to the mast, as a form of self-commitment, but that he is so obedient to Circe that he follows her instructions, rather than assuming that he could resist the Sirens’ song on his own.

This point is emphasized by the incident involving Hyperion’s oxen, which follows almost immediately.[112]  Ulysses’s devoutness is in fact the dominant, if not exclusive, theme of this extensively related incident.  Having been warned by Circe not to injure any of the oxen at the same time that she tells him how to escape the Sirens, Ulysses is inclined to bypass Hyperion’s island entirely.[113]  After his crew pleads with him, on the grounds that they do not share his iron constitution, Ulysses relents but asks that “every man of you . . . give me his solemn promise that if we come across a herd of cattle or some great flock of sheep, he will not kill a head of either in a wanton fit of folly.”[114]

At first, the men have no trouble resisting the temptation because they have plentiful supplies on their ship, but Zeus sends a furious storm that confines them to the island, and they soon exhaust their supplies and begin to suffer from starvation.[115]  In response, and pursuant to his responsibilities as leader, Ulysses goes “off inland to pray to the gods in the hope that one of them might show [him] a way of escape.”[116]  Finding a sheltered spot, he washes his hands and makes “supplications to the whole company on Olympus.”[117]  But the gods answer only by casting him into a pleasant sleep.  While he is away, one of his crew members persuades the others that they should eat the oxen, arguing that it is better to take their chances and later build a temple to Hyperion in expiation than to die of hunger.[118]  As Ulysses returns and smells the cooking meat, he exclaims “in horror”: “Father Zeus . . . and you other blessed gods who live for ever!  So it was to ruin me that you lulled me into that cruel sleep, while the men I left conceived and did this hideous thing!”[119]

The parallel to the biblical story of Moses on Mount Sinai and the golden calf is notable.[120]  While the theology differs, one idea is identical: a true leader is not someone who rebels against the divine will but rather one who is particularly reverent and obedient to it.  Moses and Ulysses share the same sense of outraged horror when they return from their isolated vigils to find that their followers have disobeyed the higher powers.  They themselves would never have done so, because their devotion to these powers is too strong.  As leaders, moreover, they would never have allowed their followers to disobey.[121]  But they needed to withdraw in order to express their devotion, and those weaker than themselves succumbed to the temptation in their absence.[122]

In other words, a basic feature of Ulysses’s personality is his obedience to the gods, and binding himself to the mast is an expression of this trait.  This feature can be generalized by recognizing obedience to the gods as a case of being guided by a higher principle.  A higher principle is simply a goal, purpose, or basis for action that applies in many different situations and is justified in terms that go beyond the circumstances of those situations.  Ulysses binds himself to the mast because Circe has recommended that he do so, but he follows her advice because he is guided by the higher principle of obedience to the gods.  Because it functions as a higher principle for him, he knows what to do when he is starving in the midst of cattle that a god has told him not to eat.  Thus, he knows what to do, and we, Homer’s audience, know what he will do when Athena tells him to disguise himself as a beggar before reentering his palace.  He simply follows the same principle.

D.    The Role of Principle

Treating the incident with the Sirens as part of the larger poem, rather than an isolated incident as Elster does, provides guidance in addressing the philosophical problem that Elster sidestepped.  Ulysses binds himself to the mast because he follows the higher principle of obedience to the gods.  Such devotion to a higher principle provides a general reason why an unimpaired decision maker at t₂ (not Ulysses, of course, who is definitely “drunk” at t₂) would follow a commitment it has reached at t₁.  The reason is that it is committed to the principle at both times; thus, it can state a rationale for making the commitment at t₁ that is convincing at t₂because of its own force.  The t₂ decision maker does not need to engage in the possibly demeaning and often unproductive inquiry about whether its decision-making process is impaired at that second time; all it needs to do is to recall the principled rationale it articulated at t₁.

Ulysses’s choice of his higher principle—obedience to the gods—separates him from the conceptual framework of the modern world.  This is hardly surprising, since the poem is at least 2,600 years old.[123]  But his actions can be readily adapted to the modern context by changing the basis for the choice of principle from obedience to instrumental rationality.[124]  When the actor reaches his decision, he does so because that decision will implement some articulated goal he has chosen autonomously, not because he is being obedient to a higher power.  While the content is different, the structure of decision is the same.  In each case, the decision maker appeals to a principle that is more general than any immediate reason for action.  That principle, moreover, can be clearly articulated in those terms, and understood at t₁,t₂,or any other time.[125]

In other words, hyperdepoliticization is a particular technique of governance.  The legislature at t₁ can use this technique when it implements an identified principle or purpose.  That will not necessarily be a common occurrence.  To reiterate, legislatures represent the people, they are supposed to respond to present circumstances, they are supposed to hold administrative agencies accountable, and they are supposed to decide on funding levels from one year to the next.  In most cases, they only need to enact ordinary legislation and can rely on their successors to adapt or amend it as the situation requires.  But there will be some cases where the principle that the legislature wants to implement is best achieved by hyperdepoliticization—that is, by enacting a measure that precludes its successors at t₂ from intervening in the usual manner.  The reason for the t₂ legislature to obey this measure—to refrain from intervening although there is no power within the government that can compel it to do so—is not that it perceives itself to be impaired but that it understands and accepts the principle which the hyperdepoliticization technique is being used to implement.  It agrees that the t₁ legislature identified a valid principle and has correctly discerned that this principle is best implemented if it willingly restrains itself from acting.

The idea that self-commitment means commitment to a higher principle facilitates the analogy between an individual decision maker and a legislature.  To begin with, as noted above, Elster’s discussion emerges from an analysis of rationality, but it is unclear that this term can be coherently applied to a collective body like a legislature.  There is no difficulty, however, in asking whether a collective body has made an instrumentally rational decision, because the criterion of rationality is applied to the decision itself, not the subjective intention of the decision maker.  Of course, people may disagree about whether the decision is in fact instrumentally rational.  To pursue that debate, they will need to identify the goal the decision is intended to achieve, determine whether the decision itself is likely to achieve that goal, and decide whether it is the best method for doing so.  These can be difficult questions, but they are essentially the same whether the decision is made by an individual or a collective body.

In addition, and perhaps more importantly, basing precommitment on the appeal to principle explains why a democratic legislature, the specific kind of decision maker at issue in the hyperdepoliticization case, is justified in making and obeying precommitments.  As already noted, legislatures are not usefully regarded as subject to the vicissitudes of individual behavior.  Thus, they are unlikely to view themselves as suffering from impaired decision-making capability at any given time.  With respect to the purely philosophical aspect of self-commitment, a legislature has at least two other reasons for noncompliance with a past decision that are not analogous to individual decision makers.  First, legislatures, being primary policy makers, are not supposed to take orders from anyone else, a status very few individuals enjoy.  This might be regarded as a matter of psychology, but it is also an aspect of decision-making methodology; unlike a court, or even the chief executive, the legislature is set up to make its own decisions, independent of other governmental actors.  Second, the legislature does not simply act for itself.  It is the people’s representative—the current people’s representative.  Even if it were inclined to honor its precommitments, it might regard such obedience as a failure of its obligation to its constituents, who do not necessarily share those precommitments.

If, however, the t₂ legislature is obeying the t₁ legislature’s decision because that decision is articulated and justified by principle, these reasons for resistance fade.  The t₂ legislature can base its obedience on the same rationale as the t₁ legislature did.  This, then, is the rationale for hyperdepoliticization.  At t_1, the legislature should attempt to precommit the action of future legislatures if it concludes that such precommitment is needed to effectuate a principle to which it subscribes.  At t_2,the legislature should obey its t₁ commitment if that commitment reflects a general principle to which the legislature still subscribes and if the principle will be effectively implemented by following the prior commitment.  It is when these conditions apply that a legislature should insulate an administrative actor from the legislature itself, as well as from the chief executive.

Commitment to principle will often be amplified, or reinforced, by the advantages of consistency.  The decision maker at t₁ can argue that, in addition to the particular principle it articulates, whatever persuasive force that principle may have at t₂, the only way to carry out a principle is to act consistently over time.  The idea is related to Gautier’s and Rubenfeld’s arguments,[126] but without any larger assertion that consistent action is in any way intrinsic to morality, personhood, or existence as an entity.  It is simply, as Holmes observes, a way for a decision maker to increase its power, its ability to make credible t₁ commitments. [127]  This is not to say that one must always obey the hobgoblin of consistency and never act in response to present circumstances.  A prior decision is generally not worth following if it is not announced as a commitment.  A decision maker derives no additional power from simply being rigid; in fact, it probably loses power.  The added power that consistency confers will generally be available only when the earlier decision is one that was stated as a commitment; that is, something that the t₁decision maker wanted subsequent decision makers to maintain.  In that case, the t₂ decision maker has a reason to follow the t₁ decision that supports the persuasive force of the principle underlying that decision.

II.  Current Cases of Hyperdepoliticization

Although hyperdepoliticization is not a recognized technique of governance at present, Congress has in fact deployed it in a variety of situations.  The discussion in the following Part is intended to be exemplary, not comprehensive.  At the outset, and even for the purpose of giving examples, it is useful to distinguish between what can be called grounded hyperdepoliticization and what can be called atmospheric hyperdepoliticization.[128]  Grounded hyperdepoliticization occurs when the legislature adopts definitive provisions for the purpose of disabling itself from future action, and can thus be regarded as the classic case.  Atmospheric hyperdepoliticization occurs when the legislature enacts a statute that attempts or aspires to insulate an agency from political pressures in general, and may apply to pressure from the legislature as well as to the more obvious case of pressure from the chief executive.  This Part offers two examples of grounded hyperdepoliticization.  The first involves the closure of military bases and was implemented by a statute; the second involves control of the money supply and was implemented by inaction—that is, the explicit decision not to alter an initiative adopted by the relevant agency.  The Part concludes with one fairly common example of atmospheric hyperdepoliticization—the creation of what can be called a deeply embedded agency.

A.     Grounded Hyperdepoliticization by Statute: The Closure of Military Bases

A notable case of self-imposed constraint by Congress involves the closure of military bases.  A military base produces major economic benefits for the place where it is located, and the expenses of operating the base are borne almost entirely by the federal government, whose revenue is generated by the nation as a whole.  As a result, members of Congress, as well as other politicians, lobby assiduously to obtain military bases for their state or district and treat their ability to do so as a major political success.[129]  However gratifying it is to obtain a base, it is more painful to have the base close once the economy of a particular locality has become dependent on the facility’s continued existence.  Thus, efforts to close any particular base tend to run into determined opposition from the local community and, in most cases, from its political representatives.[130]  As a result of this political dynamic, military bases proliferated throughout the course of American history.  By the 1980s, the military justification for their number and location (namely, to defend strategic points in the nation) had worn thin, and the Department of Defense was eager to close the ones that no longer served any other purpose.[131]

Congress responded to this dilemma by enacting successive statutes—the Base Closure and Realignment Act of 1988,[132] the Defense Base Closure and Realignment Act of 1990,[133] and the base closure provisions of the 2002 National Defense Authorization Act.[134]  The largest number of base closures occurred under the 1990 Act.[135]  This Act makes use of an independent commission, called the Base Realignment and Closure Commission (“BRAC”), created by the 1988 Act.[136]  The members of the Commission were originally appointed by the Secretary of Defense, but the 1990 Act provides that they are to be appointed by the President, with the advice and consent of the Senate, and on a bipartisan basis.[137]  Under the procedure specified in the Act, the military services submit proposals for base closures to the Secretary of Defense.[138]  The Secretary can add or delete bases from the list, and then he or she transmits the list to the BRAC Commission, which can once again add or delete bases that the Secretary has designated.[139]  Both the Secretary and the Commission are required to consider specified criteria that include the military value of the base, a cost benefit analysis of its closure, and the economic and environmental impacts that its closure would produce.[140]  Once the Commission reaches its decision, the list is then forwarded to the President, who—significantly—can approve or disapprove the list as a whole but cannot make additions or deletions.[141]  Presidential approval constitutes authority for the Secretary of Defense to close all the bases on the list.[142]  Congress retains the authority to halt the base closures by passing a Joint Resolution of Disapproval within forty-five days of the President’s approval, but it must once again act on the entire list and cannot make any additions or deletions.[143] Having failed to close any major bases during the Reagan administration, this procedure enabled the military to close sixteen major bases and “realign” eleven others under the 1988 Commission, and then close 102 major bases and realign or redirect eighty-three others under the four Commissions authorized pursuant to the 1990 Act.[144]

Considered simply as a procedure, the Base Closure Acts closely resemble Elster’s account of individual self-commitment—a definitive decision at t₁ and the development of an effective stratagem.  That stratagem can be called aggregation—creating a definitive list of base closures at the administrative level and then disabling the elected politicians who are inevitably involved (the President, who as Commander-in-Chief, probably cannot be excluded, plus Congress itself) from altering the list.  It has no simple analogue in individual behavior, however, because it addresses the problem that individual members of a collective body have different interests from the body as a whole.

The key to the procedure’s obvious success seems to be the inability of any elected official to take action with respect to a particular military base.[145]  One way to interpret this success is that the procedure raises the cost of influencing an elected official to a level that no private actor, locality, or even state can afford; in order to save one’s own preferred base, one must convince either the President or Congress to terminate a process that promises to save taxpayers billions of dollars and that represents the conclusions of an independent commission acting on recommendations from the military itself.

Kenneth Mayer, in an effort to rescue the cynicism of public-choice analysis from an impressively public-regarding statute, observes that the process “obscured the causal chain,”[146] or, in the language of other public-choice scholars, shifts blame from the legislature to an agency.[147]  By delegating the decision to an appointed, expert administrative body, legislators can argue to their constituents that they were not responsible for the actual decision that closed their locality’s or their constituents’ favored base.  In addition, according to Mayer, “the timing of the process was explicitly designed to minimize political harm.”[148]  The three rounds of base closures under the 1990 Act all required the President to act on the BRAC’s recommendation by July 15 of an odd-numbered year, ensuring a significant time lag between Congress’s refusal to disapprove the list and the next congressional election.[149]

But public-choice analysis cannot quite explain the interesting fact that Congress managed to overcome a powerful political dynamic through the mechanism of hyperdepoliticization.  The reason Congress enacted this legislation, and the reason it allowed it to continue while hundreds of locally treasured bases were closed or realigned, is that its precommitment strategy was based on principle.  The principle was of course efficiency, something that is regarded in our society as one of the greatest virtues of governmental action.

Efficiency can conflict with other values, such as human rights or social equality, but no such values were implicated here.  Any possibility that it would conflict with national security was eliminated by the fact that the Department of Defense, a widely respected agency not known for its self-abnegation about budgetary allocations, virtually begged Congress to close military bases.[150]  The Act’s procedure was designed to ensure that the particular base closures would be efficient because each closure had to be approved by three levels of Defense Department decision makers—the individual services, the Secretary, and a specially constituted committee.  In other words, the decision maker not only devised a workable strategy but also based that rationale on a widely accepted principle that the strategy uncontroversially effectuated.  That is the reason the statute, and its strategy, was considered binding by subsequent Congresses, the t₂ decision makers.  The reason that they did not amend the statute, or pass the Joint Resolutions it permitted, was not because they regarded themselves as impaired decision-makers.  It was because the enacting Congress’s precommitment made sense to them.

B.     Grounded Hyperdepoliticization by Inaction: Federal Reserve Control of the Money Supply

A second example of grounded depoliticization, this time by subsequent approval of administrative action rather than prior design of such action, involves the control of the money supply by the Federal Reserve Board (the “Fed” or “Federal Reserve”).  The Fed was created in 1913 as an independent agency to serve this purpose and bring an end to the cycle of financial panics that bedeviled the American financial system throughout the nineteenth century.[151]  As thus conceived, it was an ordinary independent agency, free of direct executive control because the Board members could not be dismissed at will,[152] but subject to the same congressional control as other federal agencies.

The statutory mechanism that Congress provided for control of the money supply was the adjustment of bank reserve requirements.  Banks, as financial intermediaries, lend out most of the money they receive from deposits.  To maintain financial stability, however, they are legally required to retain, or reserve, a percentage of their deposits, perhaps ten percent, in liquid form.[153]  This can either be cash in the bank vaults or deposits with the Federal Reserve itself (which are, in effect, cash).  The Federal Reserve Act envisioned that the Fed would control the money supply by adjusting bank reserve requirements.[154]  Lowering the required reserve allows the bank to lend more money, and thus increases the amount of money in circulation; raising the reserve requirement has the opposite effect.[155]

This sounds plausible, but it is unwieldy; even minute changes in the reserve rate will generate rather massive changes in the money supply because the rate affects every bank.[156]  As a member of an earlier Congress, Fisher Ames, commented about a different legislative strategy: it is “a great clumsy machine . . . applied to the slightest and most delicate operations—a hoof of an elephant to the strokes of mezzotino.”[157]  During the 1920s, therefore, the Federal Reserve developed a different, more flexible mechanism for controlling the money supply.  Under the leadership of Benjamin Strong, Governor of the Federal Reserve Bank of New York and a recognized financial wizard, a committee of Fed administrators and other public officials began to control the money supply by buying and selling government securities on the open market.[158]  The original legislation gave each of the Fed’s twelve regional branches authority to buy and sell government securities for various purposes, but the idea that this authority could be centralized and used as the Fed’s primary monetary control device came from Strong and the Open Market Committee.[159]  Congress codified this technique in the Banking Act of 1933 (the Glass-Steagall Act),[160] revised by the Banking Act of 1935[161] and revised again in 1942.[162]

The Open Market Committee, in its current form, consists of twelve members: the seven members of the Fed, the president of the New York Reserve Bank, and four other Federal Reserve Bank presidents on a rotating basis.[163]  It meets regularly, currently about eight times a year.[164]  Its control of the money supply through open market operations depends on the fact that American money is nothing more than obligations issued by the Federal Reserve.  This enables the Fed to adjust the amount of money in circulation by buying and selling government securities.  When the Federal Reserve buys securities on the open market, it has the unique ability to pay for those securities by simply entering the payment price in the purchaser’s account.  By doing so, it has added that sum to the money supply.  When it sells securities, it receives money, but because an obligation to itself is no obligation at all, that sum of money simply disappears; thus, it has subtracted that amount from the money supply.[165]  As is apparent, the device is extremely flexible.  The Fed can buy or sell in any amount it chooses, down to the price of a single government bond; on the other hand, even its largest purchase will not disrupt the price structure of the government securities market, because the average daily trading volume in this market is about $400 billion.[166]

Congress can of course amend the Federal Reserve’s authorizing statute whenever it wishes, but it has left its post-hoc authorization of the Fed’s open market operations essentially intact for seventy-five years.  This authorization provides for congressional oversight; the Chair of the Fed is required to appear before Congress semiannually to report on “the efforts, activities, objectives and plans of the Board and the Federal Open Market Committee with respect to the conduct of monetary policy,” and the Fed is required to issue a concurrent report covering this and related topics.[167]  But Congress has not used this opportunity to interfere in the actual operation of the Open Market Committee and often treats the Chair of the Fed with a degree of deference that is afforded to few other administrators who testify before it.[168]

The hyperdepoliticization of the Federal Reserve’s monetary control function is further buttressed by the Fed’s freedom from congressional budget control.  This is due to a unique situation that, like the monetary control function, evolved without prior planning.  In the course of its open market operations, the Fed holds large quantities of government securities and receives the interest payments on these securities.  In 2011, these payments amounted to $83.6 billion.[169]  The Fed simply returns most of this money to the United States Treasury, but it retains the amount it needs to finance its own operations—$3.4 billion in 2011.[170]  As a result, the Fed does not need to obtain funding from Congress, and Congress has thereby relinquished its ability to control the Fed through reductions, or threatened reductions, of its annual budgetary allocation.  Like its control of the money supply by committee, and the deference it receives during the semi-annual oversight hearings, the Fed’s ability to fund itself could be readily reversed.  Instead, Congress has followed the course of action to which it committed itself when these practices developed.

The continuation and resilience of Congress’s commitment to insulate the Federal Reserve’s monetary control function from Congress itself, combined with the weakness of the stratagems that it has employed to achieve this result, indicates the role that principle plays in the self-commitment process.  That principle is the general view among economists that elected officials are likely to adjust the money supply to produce short-term electoral gains but long-term economic harm if they exercise control over it.[171]  Specifically, they will facilitate their reelection by inflating the currency before elections, giving a weak economy what is sometimes called a dead-cat bounce that deludes the voters.[172]  The precise advantage of independence (defined above as independence from legislative as well as executive control) is that it separates the fiscal control, which remains with the legislature, from monetary control, which is now in the hands of the insulated agency.[173]  If the legislature wants to increase employment prior to the election, it must make structural changes in the nation’s economy by raising taxes, lowering expenditures, or issuing debt, rather than simply inflating the currency.

But why is inflation so harmful?  Why shouldn’t elected, accountable authorities determine the inflation rate?  One explanation is that the responsiveness of political authorities, however justifiable it may seem at the time, is a problem in itself because it creates price instability, which has a negative impact on economic growth.[174]  Private actors must account for fluctuations by indexing salaries, altering investment strategy, adjusting debt burdens, and engaging in other suboptimal behaviors.  Placing monetary policy in a truly independent agency that is concerned about these inefficiencies and therefore committed to a consistently low inflation rate reduces the need for this suboptimal behavior.[175]  The effect is reinforced because the act of delegating monetary control represents a credible commitment by the government that its anti-inflationary policies will remain in place.[176]  This makes so much sense that most of the world’s leading industrial nations have acted, in the post-World War II era, to insulate their central banks from political control.[177]  This is a striking development because many of these nations have rather different political structures from that of the United States and no tradition of independent agencies.[178]

The policy and budgetary independence of the Federal Reserve, although originally the product of agency initiative rather than explicit congressional action, represents a grounded form of hyperdepoliticization.  Certainly, everyone on the relevant oversight committees is, and probably most members of Congress are, well aware that these institutionalized modes of nonintervention exist.  In fact, the House and Senate Banking Committees have considered legislation to reduce or eliminate the Federal Reserve’s control of the money supply on a fairly regular basis.[179]  John Woolley describes the situation in 1975, when a perfect storm of political and economic factors led both the Chair of the House Banking Committee, Henry Reuss, and the Chair of the Senate Banking Committee, William Proxmire, to introduce bills that would have taken effective monetary control away from the Fed.[180]  After extensive hearings, Congress settled for a much more modest House Concurrent Resolution that increased the level of congressional oversight but has turned out to be largely symbolic in effect.[181]

According to Woolley, the mouse that the mid-1970s mountain of concern produced resulted from Federal Reserve Chair Arthur Burns’s intimidation of the congressional committees,[182] lack of interest on the part of congressional members because the proposals had limited political payoffs, lack of ability of the members to understand monetary policy, and pressure from the banking industry.[183]  Like the public choice interpretation of the Base Closure Act, however, this is an interpretation driven by the a priori assumption that a democratic legislature can never act on public policy grounds.  The contrary interpretation is more plausible.  In 1975, Congress gave serious and sustained attention to revising or eliminating the policy of central bank independence that had then been in place for about forty years.[184]  Its ultimate decision to maintain that independence by means of hyperdepoliticization has won nearly universal approbation from economists.  The more plausible conclusion is that this was a case of responsible legislative decision making.

Once we are willing to accept the apparently heterodox notion that Congress can make decisions in the public interest, we can see the 1975 hearings as an example of grounded hyperdepoliticization.  Unlike the Base Closure Act, this policy originally resulted from agency initiative, not explicit congressional action.  But Congress explicitly approved the agency’s action and thus made a grounded precommitment.  This commitment was perceived as such when subsequent Congresses, the decision makers at t_2, were called on to decide whether that commitment should be continued or abandoned.  It was continued because the principle of central bank independence that supported it remained persuasive.

C.     Atmospheric Hyperdepoliticization: Deeply Embedded Agencies

Congress’s approach to base closure and monetary control are examples of grounded hyperdepoliticization.  In each case, Congress has adopted explicit policies—by statutory action in one case and by inaction where it has clear authority to act in another case—that restrict its ability to influence the relevant agency (the Defense Department and the Federal Reserve) at future times.  Grounded hyperdepoliticization is striking, but it is relatively rare.  However, a legislature can employ hyperdepoliticization in a less clear and dramatic way—not by explicitly committing itself to nonintervention but by establishing a mood or atmosphere to discourage future legislatures (that is, itself at t₂) from intervening.  Unlike grounded hyperdepoliticization, which by definition requires some action specifically directed to the legislature, atmospheric hyperdepoliticization can result from action that is primarily directed to another entity, in particular the executive.  In other words, atmospheric hyperdepoliticization can be a concomitant, or even a byproduct, of a statute that establishes ordinary depoliticization.

It would not make sense to assert that ordinary depoliticization always brings atmospheric hyperdepoliticization in its train.  To begin with, this assertion is probably not true; Congress has not shown any particular reluctance to engage in statutory amendment, budget control, oversight, or staff-level contacts involving independent agencies.  In addition, and perhaps even more importantly, depoliticization is so common that conflating it with hyperdepoliticization would undermine the coherence and usefulness of the latter category.  The task, therefore, is to identify particular statutes or other legislative actions that can be reasonably understood as addressing future legislatures, even if they do so by establishing a mood rather than adopting, or declining to adopt, an identifiable provision.

One example of atmospheric hyperdepoliticization is the creation of what may be called a deeply embedded agency.  This is an agency that is placed in a structural position that creates apparent difficulties, or awkwardness, for legislative budget-cutting, oversight, casework, or staff-level contacts.  While simple depoliticization, that is, the insulation of an agency from executive control, can operate definitively, as a matter of law, its hyperdepoliticized concomitants cannot possess this definitive quality.  A democratic legislature can amend the organic statute of any agency and can generally employ one of its less severe mechanisms of control as well, such as budget adjustment and oversight.  Rather, the embedding technique depends upon creating a situation where intervention seems less appropriate than usual, where the structure of the agency itself provides members of the legislature who do not want to intervene with an additional argument that they can use against those who favor intervention.  In the absence of a definitive legislative action, determining whether the technique actually succeeds in achieving hyperdepoliticization is ultimately a matter for empirical determination.

One example of a deeply embedded agency in the U.S. government may be the Public Company Accounting Oversight Board (“PCAOB”) created by the Sarbanes-Oxley Act.[185]  The purpose of the PCAOB is to “oversee the audit of public companies that are subject to the securities laws.”[186]  To this end, the statute provides that the PCAOB:

shall be a body corporate, operate as a nonprofit corporation, and have succession until dissolved by an Act of Congress. . . .  The Board shall not be an agency or establishment of the United States Government, and, except as otherwise provided in this Act, shall be subject to, and have all the powers conferred upon a nonprofit corporation by, the District of Columbia Nonprofit Corporation Act. No member or person employed by, or agent for, the Board shall be deemed to be an officer or employee of or agent for the Federal Government by reason of such service.[187]

The members of this independent Board are  appointed by the Securities and Exchange Commission (“SEC”) and removed by the SEC only for cause.[188]  Moreover, the statute states criteria for the selection of PCAOB members: two members (and only two) must be certified public accountants, and all must be “prominent individuals of integrity and reputation who have a demonstrated commitment to the interests of investors and the public, and an understanding of the responsibilities for and nature of the financial disclosures required of issuers under the securities laws and the obligations of accountants.”[189]  In addition, the PCAOB, like the Fed, funds itself, in this case from fees that it is authorized to charge to the accounting firms it audits.[190]  In fiscal year 2011, the PCAOB collected about $202 million in fees to fund its expenditures.[191]

These statutory provisions seem to be an elaborate effort to create an expert, independent body as far removed from the ordinary, and ordinarily political, operation of government as a statutorily created agency can be.  Their obvious goal is to insulate the PCAOB from executive control, so much so that the removal provision was declared unconstitutional by the Supreme Court as an excessive intrusion on presidential authority.[192]  This is a bit awkward for a discussion of congressional techniques, to be sure, but the Court carefully limited its decision to the “for cause” removal provision and upheld the remainder of the statute.[193]

For present purposes, however, the Court’s decision is not crucial, as it only addresses the insulation of the PCAOB from presidential control.  The question is whether all the separation and insulation that the statute provides—its status as a nonprofit corporation; the criteria for selection of its members; its ability to fund itself; and the fact that it is subject to, or embedded in, an independent agency—are also intended to discourage future Congresses from interfering with the PCAOB’s operations.  Is this statute, in other words, an effort by the enacting Congress to hyperdepoliticize the PCAOB, to precommit future Congresses to respect the PCAOB’s independence and insulate it from congressional as well as presidential politics?  At present, this would appear to be the case; that is, Sarbanes-Oxley seems to be an attempt to hyperdepoliticize the PCAOB that it creates.  Whether this attempt is successful, however, can only be known, as in the case of the Federal Reserve Board Open Market Committee, by observing the actions of Congress over time.

If the intent is hyperdepoliticization, it must necessarily be based on principle and will only be effective if that principle proves persuasive to subsequent Congresses.  The principle, of course, is disinterested judgment—a neutral, expert assessment of the auditing procedures used for large American corporations.  Congress enacted Sarbanes-Oxley in response to a series of corporate frauds involving large companies such as WorldCom, Enron, and Tyco International.[194]  A major reason for these frauds, in Congress’s view, was that the accountants who audited these firms were not sufficiently independent; they did not follow proper accounting industry standards because they were subject to various pressures from the companies that they were auditing.[195]  If an oversight agency is to be created to discipline accounting firms and enforce these standards, it seems natural to insulate that agency from the same types of influences that undermined the independence of the accounting firms and caused the problem in the first place.

A second deeply embedded agency is the newly created Consumer Finance Protection Bureau (“CFPB”).  It is one part of the massive Wall Street Reform and Consumer Protection Act (the “Dodd-Frank Act”), which was enacted in response to the 2008 financial crisis.[196]  The CFPB’s role is to “regulate the offering and provision of consumer financial products or services under the Federal consumer financial laws.”[197]  In what can be reasonably regarded as an effort to avoid constitutional problems, the provisions for appointment of the CFPB’s director are not exotic.  The CFPB is headed by a single director, to be appointed by the President with the advice and consent of the Senate.[198]  The only criterion for appointment is the Director must be a citizen of the United States.[199]  He or she serves for a five-year term and can be removed by the President for cause, which is defined as “inefficiency, neglect of duty, or malfeasance in office.”[200]  These are standard provisions for the creation of an ordinary independent agency.[201]

There are, however, several features of the CFPB that signal congressional intent for greater insulation.  To begin with, the CFPB is “established in the Federal Reserve System.”[202]  One implication of this somewhat Delphic phrase is that CFPB employees are appointed in accordance with Federal Reserve selection criteria, are compensated in accordance with terms of the Federal Reserve Act, and participate in the Federal Reserve’s separate pension system.[203]  Significantly, however, it does not mean that the Fed exercises supervisory authority over the CFPB.  The Act provides that the Fed may not “intervene in any matter or proceeding before the Director, including examinations or enforcement actions, unless otherwise specifically provided by law” or “appoint, direct, or remove any officer or employee of the CFPB.”[204]  The relationship between the Fed, or indeed the rest of the Federal Reserve System, and the CFPB is one of coordination, not hierarchical control.

Of far greater significance than the application of Federal Reserve personnel policies to the CFPB is the application of the Federal Reserve’s relationship to Congress.  The CFPB is subject to essentially the same reporting requirements as the Fed, namely, semiannual hearings with concurrent reports, and participates in the Fed’s self-funding system.[205]  The funding provision reads, in part, as follows:

Each year (or quarter of such year), beginning on the designated transfer date, and each quarter thereafter, the Board of Governors shall transfer to the Bureau from the combined earnings of the Federal Reserve System, the amount determined by the Director to be reasonably necessary to carry out the authorities of the Bureau under Federal consumer financial law.[206]

Given Congress’s extensive experience with the Federal Reserve itself, these provisions, and the general establishment of the CFPB “in the Federal Reserve System,” can readily be seen as an effort by Congress to insulate the CFPB from its own intervention in the future—that is, as an effort to hyperdepoliticize the CFPB.  In effect, the enacting Congress may well be signaling to future Congresses that they should adopt the same deferential approach to oversight, and the same relinquishment of budgetary authority, that the enacting Congress and its predecessors for many decades have adopted toward the Federal Reserve Monetary control function.  If so, however, the hyperdepoliticization of the Fed is atmospheric, rather than grounded in a specific provision.  The statute does not contain any explicit provision directed toward future Congresses.  Rather, its provisions create a mood that seems to emphasize the importance of insulating the Fed from politics, and thus discouraging intervention by any elected official.

Once again, it is an empirical question whether such hyperdepoliticization will be effectuated.  The recent controversy over the appointment of the CFCB’s director is certainly not a promising sign from this perspective.[207]  It seems plausible that this controversy betokens a current disagreement about the importance of the principle that underlies the effort to insulate the CFCB from political intervention.  The rationale for such insulation certainly seems less clear than it is for the monetary control function or the supervision of corporate auditing.  Consumer protection can be regarded as an adversary function, rather than a neutral or adjudicative one.  The idea is that the agency charged with this function should favor one side of a recognized dichotomy between merchants and consumers, that it should support what is perceived to be the weaker side in order to even the playing field and achieve just results.  That may be a perfectly good policy, but it is not necessarily the sort of policy that can support an effort by a current Congress to bind its successors.  Whether that ultimately undermines the hyperdepoliticization techniques employed in the statute, or whether the CFPB is able to define a role and stance for itself that supports the use of these techniques, is something that will be determined in the coming years.

Conclusion

Congress often attempts to insulate administrative agencies from the President’s political influence.  This is sometimes called depoliticization, and given the structure of American government, it reflects a fairly standard allocation of authority by a superior among its subordinates.  More rarely, Congress seeks to insulate agencies from its own political influence as well, a technique this Essay describes as hyperdepoliticization.  Because Congress has no superior in our system, hyperdepoliticization consists of a decision maker giving instructions to itself at a later time.  Whether such an instruction can have binding force is a conundrum, one that has been extensively explored in modern analytic philosophy.  Any solution based on the idea that the decision maker is more rational or more coherent at the earlier time dodges the real philosophical problem.  In any case, it cannot be analogized to a legislature, a collective body that functions, at any given time, as the people’s representative.

The explanation offered by this Essay is that a legislature’s effort to hyperdepoliticize an agency by instructing future legislatures to desist from deploying their standard means of supervision may be regarded as persuasive or binding if certain conditions are met.  The major one is that the effort must be based on a principle that makes sense to the subsequent legislature.  The Essay then distinguishes between grounded and atmospheric forms of hyperdepoliticization.  The former is an explicit effort to achieve this result.  Examples include the Base Closure Acts of 1988 and 1990 and the Federal Reserve System’s control of the money supply.  Hyperdepoliticization has been effectively achieved in these cases because it is strongly supported by the principle of economic efficiency.  Atmospheric hyperdepoliticization is implicit, a possible byproduct of a statute that goes beyond the usual means of insulating an agency from presidential control.  The examples given in the Essay are the Public Companies Accounting Oversight Board created by the Sarbanes-Oxley Act and the Consumer Finance Protection Bureau created by the Dodd-Frank Act.  A principle of disinterested judgment supports the possible hyperdepoliticization of the PCAOB; the principle that would support the use of this technique for the CFPB is not as apparent.

Underlying these examples, and the entire technique, is the recognition that democratic legislatures are, in fact, capable of acting on the basis of principle—that is, capable of at least attempting to legislate good public policy.  The examples thus serve as evidence that this is possible.  Once this possibility is recognized, hyperdepoliticization becomes a mechanism that a legislature can deploy in appropriate circumstances.  The goal of this Essay is to identify this mechanism and explain its basis so that public-oriented legislatures can think more regularly, and more explicitly, about when they want to use it.