Printer Friendly

Contra nemo iudex in sua causa: the limits of impartiality.

ESSAY CONTENTS

INTRODUCTION

  I. NEA40 IUDEX AND CLOSE RELATIVES
     A. Core and Periphery
     B. Reformulations?

 II. UNAVOIDABLE VIOLATIONS
     A. Federal Judicial Review of State Court Decisions
     B. Judicial Review in General
     C. Legislative Districting
     D. "Independent" Agencies and Courts
     E. Combination of Functions

III. COSTS AND TRADEOFFS
     A. Direct Costs
     B. Impartiality and Expertise
        1. Judicial Bias and Disqualification
        2. Legislative Districting (Redux)
        3. Administrative Combination of Functions (Redux)
     C. Impartiality and Independence
        1. Legislative Salaries
        2. Judicial Salaries and the Rule of Necessity
        3.Qualifications and Expulsion of Legislators
     D. Impartiality and Institutional "Energy"
        1. Presidential Self-Pardons
        2. Presidential Pardons for Treason
        3. Legislative Qualifications
        4. Legislation, Partiality, and the System of Checks and
           Balances

 IV. THE CALCULUS OF POLITICAL RISKS: RULES OF THUMB  A. Marginalism
     B. Optimizing
     C. Substitute Protections

 CONCLUSION


INTRODUCTION

The maxim nemo iudex in sua causa (1)--no man should be judge in his own case--is widely thought to capture a bedrock principle of natural justice and constitutionalism. The U.S. Supreme Court calls it "a mainstay of our system of government" (2) and regularly invokes it in diverse contexts (3)--most famously as a principle of natural law in Calder v. Bull (4) and, implicitly, to justify constitutional judicial review in Marbury v. Madison. (5) The maxim has venerable roots in the common law, the leading precedent being Bonham's Case. (6) And it rests upon a principle of impartial decisionmaking that has undoubted appeal, both on deontological and consequentialist grounds. (7) As we will see, the appeal of impartiality is so strong that legal and political actors and commentators have extended the scope of nemo iudex well beyond judging in the strict sense, to cover decisions by many types of officials in many institutions.

Despite all this, I will argue that the nemo iudex principle is an exaggerated and misleading half-truth. Sometimes rulemakers in public law do and should design institutions with a view to the nemo iudex principle. In other cases, however, they do not and should not. In many settings, public law makes officials or institutions the judges of their own prerogatives, power, or legal authority. Officials or institutions may determine their own membership, award their own compensation, rule on the limits of their own jurisdiction, or adjudicate and punish violations of rules they themselves have created. Some examples:

* In many jurisdictions, legislators determine the boundaries of the districts from which the legislators themselves are elected, or otherwise structure the system by which they are elected. (8)

* In many jurisdictions, legislatures have broad authority to determine the qualifications of their own members and to expel members. (9)

* In many jurisdictions, legislators set their own salaries and compensation. (10)

* The President arguably has the power to pardon himself and clearly has the power to pardon his friends, family, and advisers. (11)

* The Vice President arguably has the power to preside at his own impeachment trial and clearly has the power to preside at the impeachment trial of friends and political allies. (12)

* A ubiquitous feature of the administrative state is that agencies combine the functions of rulemaking, prosecution, and adjudication. (13)

* In America, federal judges and many state judges rule on the constitutionality of legislative acts setting judicial salaries. (14)

* In many jurisdictions, judges have the final say over the limits of judges' power, prerogatives, and jurisdiction. (15)

* A federal judge may rule on her own immunity from suit, (16) rule on motions asking her to recuse herself for bias, and decide whether to hold litigants in contempt for violations of her own commands. At the level of the Supreme Court, each Justice rules on motions asking that Justice to recuse himself, and the rulings are unreviewable. (17)

The list is heterogeneous on several dimensions, which I will explore at length below, but that very heterogeneity illustrates that the nemo iudex principle is qualified in many settings and on many different grounds.

In this Essay, I will offer two claims, one destructive and one constructive. The destructive claim is that the nemo iudex principle cannot even be understood as a presumption to which public law sometimes makes exceptions; rather, it amounts to little more than a banal counsel that impartiality is sometimes an important value in institutional design. Impartiality constantly trades off against and competes with other values. Far from being a mainstay of our system of government, the most that can be said is that the nemo iudex principle sometimes holds and sometimes does not.

As for the constructive claim, I will attempt to identify the main grounds on which, and the main conditions under which, rule designers depart from, override, or qualify the nemo iudex principle. As we will see, in some cases there is no impartial official or institution in the picture, so that wherever decisionmaking authority is lodged, someone or other will have to be the judge in his own case. In other situations, even where it would be feasible to respect the nemo iudex principle, the costs of doing so will exceed the benefits. In general, this will be the case when and because impartiality trades off against one of several competing considerations: the benefits of expertise, the value of institutional autonomy and independence, or the motivation and activity level of officials or institutions.

The upshot is that it is never sufficient to argue that a proposed institution, or a proposed interpretation of ambiguous constitutional rules or practices, would violate the nemo iudex principle or would "put the fox in charge of the henhouse." One must go on to ask whether the conflict is avoidable or unavoidable--and, if it is avoidable, whether it would be good or bad overall to avoid it. As we shall see, in a range of cases, violations of nemo iudex are either unavoidable or affirmatively desirable, on balance.

The largest aim is to illustrate a general point about how constitutional and institutional designers manage political risks. Among the various risks that rule designers in public law must consider is the risk of self-dealing or self-serving bias on the part of decisionmaking officials or institutions. Yet there are many countervailing risks to consider, which implies that rule designers will have to trade off those risks against one another: rather than selecting maximal precautions against official or institutional self-dealing, they will have to select optimal precautions. (18) In the most difficult cases, the nemo iudex principle appears on both sides of the balance, so that preventing self-dealing by one decisionmaker will increase the risks of self-dealing by another decisionmaker. In such cases, the principle is in conflict with itself, and an appeal to it will necessarily be question begging. A well-rounded analysis will consider risks of bias that may arise on all sides of the relevant institutional questions.

Part I examines several versions of the nemo iudex principle and addresses its scope and weight. Part II examines cases in which a violation of nemo iudex is unavoidable wherever decisionmaldng authority is lodged, so that the principle is in conflict with itself. Part III examines cases in which the value of impartial decisionmaking embodied in nemo iudex trades off against other values: the direct costs of appointing an impartial decisionmaker (Section III.A), expertise and information (Section III.B), institutional autonomy and independence (Section III.C), or institutional "energy" and activity levels (Section III.D). Part IV offers some rules of thumb for identifying conditions under which rule designers should or should not entrust officials or institutions with the authority to act as judges in their own cause. A brief Conclusion follows.

I. NEMO IUDEX AND CLOSE RELATIVES

Although I will speak throughout of the "nemo iudex principle," this is actually a simplification for ease of exposition. In fact there are multiple Latin tags whose content is more or less equivalent. These include nemo debet esse iudex in propria causa (no one ought to be judge in his own cause) and nemo potest esse simul actor et iudex (no one can be both litigant and judge at the same time); (19) the latter version clarifies that the principle forbids simultaneous judging and litigating in the same case.

A. Core and Periphery

If there is an inviolable core to these maxims--and I will later suggest that there is not--the core is that no person should be permitted to judge his own cause by sitting as a judge over litigation to which he is a named party. There are, however, a variety of ways of extending nemo iudex outward from this situation. For one thing, the principle is often said to apply when a judge sits to decide a case in which he has a more or less direct financial interest, even if the judge himself is not also one of the nominal litigants. (20) A slight further extension involves cases in which the judge's relatives or friends have a financial interest. Still further out are scenarios in which the judge's interest in the case might expand the power or perquisites of the judicial branch as a whole, in which the judge holds only a fractional share, (21) or of the government as a whole, in which the judge's fractional share is smaller still. Finally, the most aggressive extensions posit that not merely an interest in the litigation, but an appearance of an interest, such that the judge's impartiality "might reasonably be questioned," (22) suffices to trigger the principle.

Moreover, although the principle and its relatives speak of judging, they have long since been extended to embrace other types of decisionmaking that are implicitly analogized to judging. When James Madison invoked the principle in Federalist No. 10 to describe a majority legislative faction's voting for legal rules that would favor the interests of that faction, (23) the usage was at least doubly, and perhaps triply, metaphorical. Not only did the issue involve legislative rather than judicial voting, but the interests at issue were indirect rather than direct, and the cause that any given legislator was "judging" might be collective rather than individual, depending upon whether Madison saw the faction's interests as collective or distributed among its individual members. (24)

In addition, where executive officials apply settled law to facts--a class of decisions that is hard to distinguish from judging, except by reference to the institutional attributes of the decisionmaker--the Court has also invoked nemo iudex. In Gutierrez de Martinez v. Lamagno, (25) the question was whether a court could review a certification by the Attorney General that a federal employee sued for a tort was acting within the scope of federal employment. In the circumstances of the case, the certification would have had the effect of immunizing the employee from litigation, and the Court saw this as implicating nemo iudex, on the ground that the Attorney General would be acting as judge in her own cause. Not only was this an extension of the maxim to executive decisions rather than judicial ones, but the Court's description of the Attorney General's decision as judging in her own cause was highly strained. At most, as the dissent pointed out, the Attorney General would be favoring the interests of other federal employees. (26)

B. Reformulations?

One line of response to the many examples in which the constitutional system violates nemo iudex as broadly understood by the Court and by commentators is simply to reformulate the principle. Reformulations might address either (or both) of two margins: the weight of the principle or its scope. The former strategy would say that the principle does apply broadly, just with reduced force. The latter would attempt to preserve the strong force of the principle within some narrower domain.

Reformulation by reducing the weight of the principle is fully compatible with my account. Later in this Essay, I will attempt to show that the principle identifies impartiality as an institutional good under certain conditions, that impartiality is merely one good among the many goods that institutional designers must take into account, and that in a wide range of cases rulemakers will have to trade off these goods against one another. Thus, I have no quarrel in principle with this approach. To be clear, however, this would indeed be a reformulation. The traditional conception holds that nemo iudex is an "inflexible" (27) principle that is fundamental to the legal system, rather than a mere reminder that impartial decisionmaking is sometimes a good thing, in the right circumstances.

A different strategy would constrict the scope of the principle in order to preserve its force within the new, narrower scope. The reformulation would say that nemo iudex as such applies only to cases in which judges are also named litigants and presumably also to cases in which judges have a direct pecuniary interest in the outcome. As a result, other institutional structures that create risks of partial or biased decisionmaking might then be labeled (mere) "conflicts of interest" and deemed subject to institutional tradeoffs.

For several reasons, however, this approach fares poorly. For one thing, the historical genesis of the principle is not at all consistent with it. Consider Bonham's Case, frequently cited for the sacrosanct status of nemo iudex, and according to some, the foundation for something like constitutional judicial review. (28) The facts of the case involved an institutional conflict of interest in which a chartered college of physicians claimed legal authority to fine or imprison physicians practicing without a license. In his invocation of nemo iudex, Chief Justice Coke objected to this arrangement because it made the censors of the college into "judges, ministers, and parties" simultaneously, (29) but we will see that similar institutional conflicts of interest on the part of public and quasi-public agencies are a routine feature of the modern American administrative state. If the reformulated principle condemns this sort of arrangement, it is too broad; if it does not, then it does not even encompass Bonham's Case--the leading nemo iudex precedent at common law.

Furthermore, current constitutional law simply cannot be squared with this reformulation. On the one hand, the Court invokes nemo iudex in situations that lie outside the scope of the narrowed reformulation, as in the "attenuated" circumstances of Gutierrez de Martinez. (30) On the other hand, constitutional law licenses violations of nemo iudex even in situations that should lie squarely within the core. When federal judges sit to decide cases concerning judicial salaries--cases brought by plaintiffs-judges to determine the salaries of the whole group of sitting judges, including the judges who will decide the case itself--there is not an attenuated conflict of interest, but rather a direct one. If the plaintiff-judges represent a class that includes the judge-judges, then the latter are litigants in name and the violation of the nemo iudex maxim is quite literal. One cannot salvage nemo iudex by constricting its scope, trying to hive off the institutional violations while preserving a core domain for the principle, if at least some of the violations themselves lie within the core. Whatever the normative justifications for the reformulation--and I will argue that the normative arguments are complex and highly contingent--it fails the test of fit with current law.

All told, the nemo iudex principle is invoked promiscuously, in a range of different settings with different decisionmakers and different types of decisions. My suggestion, however, is that the principle is not consistently honored in any of these settings, whether inside or outside the situation of self-judging that lies at the putative core of nemo iudex. It is useful here to compare nemo iudex with any one of the many settled principles that indisputably pervade our morality and law--say, the principle that it is a legal and moral wrong to kill another person intentionally. (31) That principle is something like a presumption subject to exceptions, such as self-defense and defense of third parties. But those exceptions are well-defined and are plausibly deemed exceptional relative to the large domain within which the principle unquestionably applies. By contrast, the list I gave earlier suggests that central structural features of the constitutional system are inconsistent with nemo iudex, not only in its broadest formulations but even at its narrower putative core. And as I will suggest later, constitutional rules and structures that are plausibly based on nemo iudex are for the most part afterthoughts--second-decimal adjustments or minor side constraints.

The upshot is that there are two ways of characterizing the actual scope and weight of nemo iudex in our constitutional and legal order. One might say that nemo iudex is a sham principle because it is more exception than rule, like a house that is made mostly of doors. Alternatively, one might say that nemo iudex is a genuine principle, just one that is both weak within the domain in which it applies, and that is also riddled with a large set of ill-defined exceptions--something like the moral principle against lying, which is frequently defeasible and actually defeated, and which is likewise riddled with a large set of ill-defined exceptions. Which of these two characterizations is best will turn, in part, upon what counts as a "principle," a jurisprudential issue that I will leave to philosophers of law. For the purposes of constitutional theory, however, I need not choose between the two characterizations because both are inconsistent with the traditional legal claim that nemo iudex is a mainstay or axiom of the American constitutional order. Whether described as a sham principle or as a weak and porous principle, nemo iudex amounts to far less than the received wisdom suggests.

Overall, in light of the numerous tradeoffs that the following Parts will detail, nemo iudex is best understood merely as one competing consideration among many. Nemo iudex points to the value of impartial decisionmaking, but there are many institutional goods besides impartiality, and in many settings rule designers decide that those other goods are more important under the circumstances, or that none of the feasible institutional arrangements can produce impartiality. Or so I will proceed to argue.

II. UNAVOIDABLE VIOLATIONS

One major class of rejoinders to the nemo iudex principle arises when the alternative decisionmaker is said to be equally partial or biased. In the strongest versions of this argument, the claim is that for structural reasons there can he no impartial decisionmaker in the relevant domain, so that any allocation of decisionmaking authority must necessarily violate nemo iudex. I will offer a series of examples.

A. Federal Judicial Review of State Court Decisions

Joseph Story provides an early example of this response in his role as a Supreme Court Justice. In Martin v. Hunter's Lessee, (32) the main issue was whether the Court could review and, if necessary, overturn state court decisions on matters of federal law. An underlying concern for state officials and Anti-Federalists was that federal courts would be biased in favor of extending the scope of federal laws and federal constitutional power, including the power of the federal judges themselves: federal judges deciding federal questions, they worried, would be acting as the arbiters of their own power. In response to this concern, Justice Story argued, among other things, that there was no impartial decisionmaker anywhere in the picture:
   From the very nature of things, the absolute right of decision, in
   the last resort, must rest somewhere--wherever it may be vested it
   is susceptible of abuse. In all questions of jurisdiction the
   inferior, or appellate court, must pronounce the final judgment;
   and common sense, as well as legal reasoning, has conferred it upon
   the latter. (33)


Although Justice Story was too much of a political veteran to be explicit here, given the sensitivity of state-federal relations in his day, his point was that if the federal courts are not the final judges of their own jurisdiction, the state courts will be the final judges of their own jurisdiction. As the latter regime would also violate nemo iudex, the choice between the two regimes must be made on other grounds.

B. Judicial Review in General

Later commentators have transposed Justice Story's argument into an objection against judicial review of individual rights claims and indeed generalized from it an objection to judicial review generally. These commentators are responding to a long and disreputable tradition in which nemo iudex is invoked, in the most facile manner imaginable, to justify judicial review. After all, the reasoning runs, if there is no independent body charged with deciding whether legislatures have stayed within the constitutional boundaries of their powers, then the only judge of that question is the legislature itself, and this would violate nemo iudex or "appoint the fox as guardian of the henhouse." (34) As against this sort of claim, Jeremy Waldron argues quite rightly that no institutional arrangement can avoid the nemo iudex problem:
   Those who invoke the maxim nemo iudex in sua causa in this context
   say that it requires that a final decision about rights should not
   be left in the hands of the people. Rather, it should be passed on
   to an independent and impartial institution such as a court.

   It is hard to see the force of this argument. Almost any
   conceivable decision rule will eventually involve someone deciding
   in his own case. Unless we envisage a literally endless chain of
   appeals, there will always be some person or institution whose
   decision is final. And of that person or institution, we can always
   say that because it has the last word, its members are ipso facto
   ruling on the acceptability of their own view. (35)


The issue is somewhat more complicated than Waldron allows, but his basic point is correct: judicial review allows the Court to rule on the limits of its own power, and thus puts the fox in charge of the henhouse just as much as does a system with no judicial review at all. The complication is that there is a third way between legislative and judicial supremacy in constitutional matters: a departmentalist system in which both Congress and the Court can decide constitutional questions as they see fit, with neither having the authority to bind the judgment of the other. But in that sort of system, both Congress and the Court still do rule--so far as their rulings go--on the limits of their own power. The departmentalist system, in other words, puts several foxes in charge of several henhouses. That may or may not be a better system, all things considered, than either the system of legislative supremacy or the system of judicial supremacy, depending on whether the foxes somehow check one another. But it is not a system that avoids violating the nemo iudex principle; on the contrary, it multiplies the violations.

C. Legislative Districting

The unavoidability arguments offered by Justice Story and Waldron rebut nemo iudex on conceptual grounds: in the relevant settings, there is no possible allocation of decisionmaking authority that avoids making some official or institution judge in its own cause. In other settings, however, the unavoidability argument is not conceptual, but institutional. Although it may be possible, in the abstract, to allocate decisionmaking authority to an impartial institution, the incentives of the relevant political actors ensure either that no such institution will actually come into being or that impartiality will prove unsustainable in the long run.

An example involves legislative districting. The current system--in which legislators often determine the shape and composition of their own districts--represents a large-scale violation of nemo iudex. Accordingly, a popular position holds that districting should be entrusted to independent and impartial expert commissions or bodies that will supposedly enjoy the freedom to draw districts based on objective, welfare-maximizing criteria. Against this argument, proponents of legislative districting offer many rebuttals, one of which is that independent and impartial districting is infeasible because it is not incentive compatible. As the districting commission must itself be created by the political system it is set up to regulate, the political actors who establish it will predictably rig its powers, procedures, and composition to achieve the same ends that they pursue in a system of legislative districting. On this view,
   it is almost impossible to design institutions to be authentically
   nonpartisan and politically disinterested.... Whoever draws the
   lines must get authority from somewhere--the person will either be
   appointed or elected. Elected officials ... are almost certainly
   conflicted. And appointed officials will be beholden to those
   appointing them or at least selected because their intentions are
   well-known. (36)


Alternatively, political actors may simply undo the independence of the districting commission if it produces politically unacceptable results. In Arizona in 2011, for example, the Republican governor and state senate fired the chair of an independent districting commission, with the governor claiming that the chair displayed a bias in favor of Democrats. (37) This was a rare case of an observable confrontation, but the threat of such action will implicitly constrain "independent" commissions by virtue of the law of anticipated reactions. Sometimes the underlying incentive-compatibility problem can be overcome by law or norms, but it is an open question how often, and for how long, true independence can be sustained.

D. "Independent" Agencies and Courts

Similar political mechanisms underpin two other examples, both of which are sufficiently familiar in the public law literature that I will mention them only briefly. The first involves the "independent" agencies, (38) such as the National Labor Relations Board, whose members can be discharged only for cause, and who typically serve staggered terms. Although such structures theoretically insulate the independent agencies from presidential control, evidence suggests that by the end of their first term, presidents typically control policymaking at "independent" agencies, in part by appointing members whose political preferences are predictable. (39)

Second, and more famously, a large body of political science evidence shows that the Supreme Court itself follows the election returns, albeit with a lag, largely because appointments to the Court are made by the same political system that the Constitution is supposed to regulate. (40) Although genuinely countermajoritarian rulings are possible in the short run, (41) in the long run, constitutional law is highly elastic, and changes as public opinion affects the selection of Justices by the President and the Senate. At least in this long-run sense, a genuinely independent and politically impartial Court is an unlikely arrangement given the incentives in place.

E. Combination of Functions

But perhaps the largest compromise of the nemo iudex principle to be found in public law is the combination of functions in administrative agencies--a routine feature of the massive administrative state. Although the variety of administrative institutions and procedures is bewildering, many agencies in some way or another combine the powers of rulemaking, investigation, prosecution, and adjudication. Such agencies, in other words, may decide cases that they themselves have investigated and decided to bring, under rules that they themselves have made. From a traditional perspective, this combination of functions makes the agencies both parties and judges--a violation of the supposed core of the nemo iudex principle. Libertarian and originalist critics of the administrative state thus routinely complain that agencies act as judges in their own cause. (42)

In the face of similar objections, James Landis's argument for the combination of prosecutorial and adjudicative functions in administrative agencies relied in part on the claim that separating functions by lodging adjudicative power in courts alone would merely risk a different form of bias. (43) On this view, while the combination of administrative functions risks biased decisionmaking by agencies that decide the cases they themselves have initiated, judicial determination of agency prosecutions risks running aground on the ideological biases of the judges, whose opinions will on average have been formed in an earlier era. As Landis put it in his discussion of the Federal Trade Commission's creation,
   Judicial interpretation of the statutory standards laid down by the
   Congress plainly gave the judges power to mold the statute to their
   own conceptions; and that molding had too frequently set at naught
   the public and political effort which had so hopefully expended
   itself in the passage of the statute. Judicial interpretation
   suffered not only from inexpertness but more from the slowness of
   that process to attune itself to the demands of the day. (44)


Although the political appointment mechanism ensures that in the long run independent judges will follow the election returns, it does not follow that they are legally impartial at any given time. Rather, the risk is that, given the lag time inherent in the appointment mechanism, the judges will impose the views of "the day before yesterday" (45) on the administrative agencies. Landis argues in effect that the insulation of the judiciary from current politics itself creates room for the operation of ideological biases, as opposed to impartial legality; hence there is no impartial institution in the picture, and the combination of functions in agencies cannot be rejected simply on the ground that it creates a risk of biased decisionmaking. Rather, one must compare the biases of all relevant institutional alternatives. (46)

III. COSTS AND TRADEOFFS

I now turn to cases in which the benefits of an impartial decisionmaker trade off against, and may be outweighed by, competing considerations: the direct costs of appointing an impartial decisionmaker, expertise, institutional independence or autonomy, and institutional energy. These tradeoffs are hardly amenable to precise analysis. In such cases, rule designers engage in an impressionistic balancing, with ill-specified weights, under conditions of grave uncertainty. But the rule designers cannot simply throw up their hands in the face of conceptual and empirical difficulties. A rule must be chosen, and as we will see, rule designers will sometimes have substantial reasons to conclude that nemo iudex should give way.

A. Direct Costs

I begin with the nearly trivial point that the nemo iudex principle's application is routinely constrained by cost. Where two parties are in conflict, appointing a third, impartial arbiter can be an expensive exercise. If the state subsidizes the appointment, as when the state maintains a system of courts to which all parties may have nearly free access, those costs are subsidized and spread across all taxpayers. In some settings, however, the out-of-pocket costs of obtaining an impartial judgment may be decisive.

An analogy is in order. In local tennis tournaments, a common arrangement is that referees are present only for the final, decisive rounds. In earlier rounds, the players themselves judge whether balls hit toward their side are in or out, in effect becoming judges in their own cause. The principal reason for this regime is that it is simply too expensive to staff local tournaments with a large complement of referees for all matches in the early rounds.

In theory, this structure also illustrates another idea: offsetting violations of nemo iudex can "mimic ... impartiality." (47) In this example, because each player judges the shots hit to her own side, each is judge in her own cause, and thus both players are in a kind of repeat-play relationship for the duration of the match. The players thus have a tit-for-tat incentive to make fair calls. This mechanism is highly fragile, however. Self-serving bias is pervasive even if the players are seeking to cooperate, and crucial calls late in the match will no longer take place under an indefinite horizon of future cooperation. The result is defection and blatantly self-serving calls that increase in number as the match progresses.

Of course, the constitutional system can draw upon a cadre of state-subsidized impartial decisionmakers--judges--so the cost constraint is somewhat relaxed. However, the constraint still bites. Statutes require federal judges to recuse themselves from a case when their impartiality "might reasonably be questioned," (48) and if recusal occurs, a second judge is called in to preside. However, the first judge rules on the disqualification motion, and thus assesses her own bias, subject only to deferential review. (49) One of the main justifications for this practice is the simple "[in]efficiency" (50) or costliness of calling in a second judge to decide a preliminary motion of this sort; the costs include delay, which harms both the litigants and the legal system. More generally, one of the reasons that officials and citizens cannot instantly obtain an impartial judicial opinion on the legal validity of their actions is that resource constraints create a queue for the use of the courts. The delay that is part and parcel of litigation arises, in part, simply because it is not feasible to create a cadre of judges of the size that would be necessary to process all cases swiftly.

B. Impartiality and Expertise

In the theory of institutional design, a standard tradeoff involves a conflict between the values of impartiality and expertise. In some settings, it is possible to decrease the bias of decisionmakers solely by reducing the information they hold or their incentives to invest in acquiring new information. For example, the rules for selecting petit juries exclude jurors with antecedent knowledge of the case in order to minimize bias, in contrast to the medieval English rules, under which jurors were selected from the locality. It has been argued that the impartiality produced by adversarial rules of litigation, which prohibit ex parte presentations and restrict judges' ability to gather relevant information outside of court, have the effect of reducing judges' information. (51) "The tradeoff is between a political process that integrates far more information but with a more significant risk of bias and an adjudicative process that suppresses information but decreases distortions in its presentation." (52)

Conversely, selecting the most informed decisionmakers tends to produce an increased risk of bias. In the administrative state, agencies are often staffed by actors with close ties to regulated groups. Although such actors are more likely to have an agenda, where the regulated domain is highly complex, actors of that sort will typically have indispensable specialized knowledge, (53) Here and elsewhere, actors with a stake in the decision are precisely the actors with the greatest incentive to acquire relevant information, which yields a chronic tradeoff between well-informed decisionmaking and impartial decisionmaking.

I. Judicial Bias and Disqualification

Where the nemo iudex principle is qualified or violated, the actual or stated justification is sometimes that enforcing the principle would produce unacceptable costs by eliminating expertise or weeding out the best-informed decisionmakers. In the judicial setting, judges rule on motions for disqualification, in effect ruling on their own bias rather than calling in a second judge. Although relevant statutes allow the inferior judges' rulings on their own bias to be reviewed by higher courts, the review is deferential (and decisions about recusal by Supreme Court Justices themselves are not reviewable at all).

One of the standard justifications for this set of rules is that the second judge lacks the first's information about the background circumstances of the case. The Court has held that the rules governing judicial bias and recusal contain an implied exception: bias in the pejorative sense arises only if the judge's preconceptions derive from an "extrajudicial source." (54) Preconceptions that the judge has formed during preliminary stages of the proceeding, or related earlier proceedings, do not generally count as "bias" of the sort that the law will find invidious. The justification for the exception is based in part on the value of the information held by the presiding judge, which would be lost if previous determinations in the case could create grounds for disqualification. As the Court put it in Liteky v. United States,
   The judge who presides at a trial may, upon completion of the
   evidence, be exceedingly ill disposed towards the defendant.... But
   the judge is not thereby recusable for bias or prejudice, since his
   knowledge and the opinion it produced were properly and necessarily
   acquired in the course of the proceedings, and are indeed sometimes
   (as in a bench trial) necessary to completion of the judge's task.
   As Judge Jerome Frank pithily put it: "Impartiality is not
   gullibility. Disinterestedness does not mean child-like innocence.
   "(55)


By contrast, the rules for selecting petit jurors strictly maximize impartiality and thus in effect aspire to child-like innocence.

2. Legislative Districting (Redux)

A similar tradeoff between impartiality and information can also be observed in legislative settings. As discussed earlier, in many jurisdictions legislatures determine the composition of the districts from which the legislators themselves are selected-an institutional conflict of interest that may produce incumbent-favoring gerrymanders, partisan gerrymanders, or collusive, bipartisan gerrymanders. Many commentators thus call for independent and nonpartisan districting by expert commissions or courts. (56)

One of the main defenses of the current regime, however, is that legislative redistricting is desirable because legislators possess crucial information about relevant constituencies and their distinctive problems-information that is perhaps held in a largely tacit or experiential form, and thus cannot be easily transmitted to nonlegislative redistricting bodies. (57) On this view, the basic tradeoff is that on the one hand, "incumbents' knowledge of their districts gives them almost unparalleled expertise as to the effect of a given set of lines and insiders are more likely to be sensitive to community concerns," while on the other hand, "bipartisan or partisan gerrymanders sometimes intentionally disrupt or divide communities and often ignore policy goals en route to creating safe seats." (58) However this tradeoff cashes out, and whatever the optimal design of redistricting institutions in a given political environment, it is fatally simplistic to condemn legislative redistricting on the grounds that it puts the fox in charge of the henhouse. (59)

3. Administrative Combination of Functions (Redux)

As discussed earlier, originalist and libertarian critics of the administrative state object to the administrative combination of functions on nemo iudex grounds and call for strict separation of functions within or across agencies. Both Congress and the Supreme Court have, however, consistently rejected such claims. The Administrative Procedure Act requires that separate personnel perform the functions of investigation and prosecution, on the one hand, and adjudication, on the other, at lower levels of the administrative agency. (60) Despite this, the statute exempts the agency or its members--that is, the top-level agency heads or commissioners-from this separation requirement. (61) Many agency heads thus hear and decide cases that they themselves have directed subordinates to investigate and prosecute, and in which the agency itself is one of the named parties. The Supreme Court has consistently upheld this arrangement against due process challenges complaining that the combination of prosecutorial and adjudicative functions compromises the impartiality of agencies. (62) And the main ground for the Court's decisions has been that the combination of functions is necessary to secure expert administrative decisionmaking in a complex society.

In Withrow v. Larkin, (63) for example, the Court framed the issue as a tradeoff among competing political risks. The case involved a due process challenge to a state administrative board charged with policing medical misconduct (akin to Bonham's Case). The board had the power to investigate licensed physicians, cause charges to be brought against them, and then adjudicate the charges after a hearing; the principal sanction the board could impose was suspension of the physician's license. The Court rejected "[t]he contention that the combination of investigative and adjudicative functions necessarily creates an unconstitutional risk of bias in administrative adjudication." (64) The tradeoff was that a constitutional rule barring combination of administrative functions "would bring down too many procedures designed, and working well, for a governmental structure of great and growing complexity." (65) Here the Court echoed one of the original New Deal defenses of the combination of functions, which held that the development of administrative expertise would be hamstrung if agencies could not shape their own regulatory agendas by bringing cases that the agency itself would adjudicate. (66) Given basic tradeoffs between impartiality and informed expertise in the administrative state, consistent implementation of nemo iudex would promote impartiality at too great a price. (67)

C. Impartiality and Independence

In other settings, rule designers qualify or violate the nemo iudex principle in order to ensure the independence or autonomy of institutions. In these settings, a given institution is made judge in its own cause because allocating decisionmaking authority to a different institution creates a risk that the second institution will leverage its authority to control the first in ways that will be undesirable from a larger systemic perspective. Where this is so, giving the first institution the authority to act as judge in its own cause may create a risk of self-dealing, but the rule designer may believe this to be an acceptable risk-an unavoidable precondition for, or byproduct of, a system of institutional independence that is desirable overall.

I. Legislative Salaries

The network of constitutional rules governing legislative salaries illustrates these tradeoffs. The enacted Constitution authorized the Congress, acting by law, to pay its members a salary out of the national treasury. (68) This outcome emerged only after extensive debates among the delegates at the Philadelphia Convention. The discussion centered on three related questions. First, should the legislators be paid at all? Second, if they were to be paid, should they be paid by the states from which they were selected--as was the practice under the Articles of Confederation--or instead by legislative appropriation from the national treasury? Third, should the amount of compensation be fixed by the Convention or left to the discretion of future legislators themselves?

For present purposes, the second and third questions are the most critical. Madison argued at the Convention that legislators should not be allowed to compensate themselves because of the risk of self-dealing: "Mr. Madison thought the members of the [legislature] too much interested to ascertain their own compensation. It [would] be indecent to put their hands into the public purse for the sake of their own pockets." (69) On the other hand, Madison, like many other delegates, also argued that "it would be improper to leave the members of the [national] legislature to be provided for by the State [legislatures]: because it would create an improper dependence." (70) As Alexander Hamilton put it in one of his rare interventions, "Those who pay are the masters of those who are paid." (71) Hamilton "pressed the distinction between State [Governments and] the people. The former [would] be the rivals of the [General Government]. The State legislatures ought not therefore to be the pay masters of the latter." (72)

In the face of this dilemma, Madison argued for a constitutionally fixed standard of compensation, perhaps indexed to the price of wheat or some other commodity to allow increase over time. Most of the other delegates, however, thought that a fixed compensation was undesirable or unworkable, either because its visibility would instigate opposition to the proposed Constitution (73) or because any fixed standard would be insufficiently sensitive to changing circumstances. (74) As between the solutions that allowed for variable compensation--payment by the states or self-payment by federal legislators-the Convention opted for the latter solution, with several members arguing that the risk of self-dealing was low, perhaps because the great visibility of legislative salaries would trigger popular oversight. Indeed, Roger Sherman "was not afraid that the Legislature would make their own wages too high; but too low." (75) Throughout the debate, delegates took account of Elbridge Gerry's observation that "there are difficulties on both sides" (76) and offered judgments about how best to balance the competing risks--a far cry from simple invocation of nemo iudex.

A different solution would be to let a different branch of the national government, such as the President, set the salaries for federal legislators. Toward the close of the debate, the Convention agreed to a motion to provide that payment from the national treasury should be "ascertained by law"--that is, by statute as opposed to by an internal legislative rule. (77) Although the shape of the executive branch was still fluid and contested when this amendment was carried, the eventual consequence was that the President would have the power to veto a bill setting legislative salaries. Here too, however, legislators would still play a role in setting their own compensation. No one raised the possibility of a pure regime in which, for example, the President acting alone would set the legislature's salary. Despite its apparent compatibility with nemo iudex, the obvious concern about such a regime would be that the President could leverage his power over compensation to compromise legislative autonomy. In light of the notorious practices by which British monarchs of the eighteenth century had corrupted members of Parliament, the Convention adopted several provisions intended to check executive vote buying, (78) and presidential control over legislators' compensation would have been inconsistent with those efforts.

2. Judicial Salaries and the Rule of Necessity

The value of institutional independence is also said to underpin the Supreme Court's invocation of the "rule of necessity" in cases where federal judges bring lawsuits complaining that their salaries have been unconstitutionally diminished. Article III's Compensation Clause provides that the judges "shall, at stated times, receive for their Services a Compensation which shall not be diminished during their Continuance in Office." (79) In some cases, judges bring suits that will determine the compensation of all federal judges as such, so that every federal judge has a direct financial interest in the outcome. (80) The implication is that any judge who decides the case is sitting in her own cause, a core violation of nemo iudex. The Court, however, allows such suits to proceed under the common law rule of necessity, which holds that if all judges would be disqualified, none is. (81)

It is hardly obvious that the rule of necessity is really necessary. If the case cannot proceed without a nemo iudex violation, perhaps it should not proceed at all. The Court sometimes cites an "absolute duty of judges to hear and decide cases within their jurisdiction," (82) a phantom duty that is routinely violated by federal judges under myriad prudential doctrines that license abstention from the exercise of jurisdiction. (83) Alternatively, the Court sometimes points to the need for judge-litigants to obtain a forum for litigating their individual constitutional rights, (84) but this rationale is no more successful. If the rule of necessity even applies, then what is at stake is really an attempt by judge-litigants to obtain a forum that is necessarily biased in their favor, and it is unclear why it is better that they should enjoy such a forum than that their constitutional claims should be remitted to whatever protection the political process will afford.

If there is a plausible rationale for the rule of necessity, it is that suits for compensation by judge-litigants produce a positive externality by helping to protect the independence of the judges from political retaliation--one of the main aims of the Article III rules. As the Court puts it, "[T]he Compensation Clause is designed to benefit, not the judges as individuals, but the public interest in a competent and independent judiciary." (85) The judges say, in other words, that the risk of biased adjudication in their own favor is an unavoidable byproduct of a regime that minimizes the risks of political interference with judicial independence. On this happy view, the judges' interests align with those of the overall system; what is good for judicial salaries is good for the nation. (86) Whatever the ultimate merits of that claim, the tradeoff between impartiality and independence is clear. (87)

3. Qualifications and Expulsion of Legislators

A final example involves the authority of each chamber of Congress, under Article I, to "judge" the "Elections, Returns and Qualifications" of its members and to expel members. (88) Although there is a norm that legislators do not vote on their own personal qualifications or expulsion, this regime nonetheless violates the extended sense of nemo iudex that Madison invoked in Federalist No. 10. In the aggregate, each chamber as a group determines its own composition. The standard defense of this arrangement, offered by Justice Story in his Commentaries on the Constitution, holds that "[t]he only possible question on such a subject is, as to the body, in which such a power shall be lodged. If lodged in any other, than the legislative body itself, its independence, its purity, and even its existence and action may be destroyed, or put into imminent danger." (89) Justice Story's argument refers in explicit terms to the political risk that committing the power to judge qualifications elsewhere would compromise the independence of the chamber. Implicitly, Justice Story trades this off against the risk of institutional self-dealing that is created by cameral self-judging of qualifications and finds the former risk more serious. As we will see in Part IV, however, the Philadelphia Convention also adopted substitute protections against the latter risk.

D. Impartiality and Institutional "Energy"

At least since Niccolo Machiavelli's analysis of glory seeking as a spur to executive action, (90) constitutional actors and analysts have discussed the effects of constitutional rules on institutional "energy." (91) Although the nature of institutional energy is obscure, a straightforward interpretation is that institutions and officials that at least partly control their own agendas may choose varying levels of activity or outputs, and that constitutional rules may shape and constrain such choices. The more institutional activity is desirable, the more costly are rules that give institutional actors an incentive to do nothing rather than something.

For present purposes, the important point is that in certain environments, impartiality may trade off against institutional activity levels. Where constitutional rules assign decisionmaking authority to self-interested or biased actors, the benefit may be that those actors will have greater motivation or incentive to act, for the very reason that they have a stake in doing so. A fallacy lurking behind the stock arguments for the nemo iudex principle is that biased action (somehow defined) is necessarily socially undesirable action. On the contrary, self-interested or biased motivation may be the spur to undertake action that happens to be socially desirable from some external perspective. Where this occurs, constitutional actors are led, as if by an invisible hand, to promote social welfare as a byproduct of pursuing their own interests.

I. Presidential Self-Pardons

An example involves the debate over presidential self-pardons. To date, this debate is strictly hypothetical; no President has ever issued a pardon for himself. Yet the structure of the debate helps to illuminate the scope and limits of nemo iudex. For critics of the possibility of presidential self-pardons, (92) nemo iudex is a deep principle of the constitutional structure, which should be read into specific provisions where possible. The critics accordingly claim that the facially unqualified language of Article II's Pardon Clause--"The President ... shall have power to grant reprieves and pardons for offenses against the United States, except in cases of impeachment" (93)--should be read as containing an implied prohibition against presidential self-pardons. On this view, the Constitution's specific provisions evidence a general "structural distaste for self-dealing" and "self-judging." (94) The asserted examples of this structural distaste include the constitutional prohibitions on legislators accepting offices whose emoluments they have voted on, (95) the Twenty-Seventh Amendment's ban on salary raises without an intervening election, (96) and a putative prohibition on the Vice President's presiding over the Senate during his own impeachment trial--although the last example is an entirely implicit prohibition that rests on nothing more solid than the same "structural distaste" for self-judging, the same assertion of nemo iudex that underlies the main argument against presidential self-pardons. (97)

One of the main responses to this critique invokes the countervailing risk that a structural prohibition on presidential self-pardons may rule out self-pardons that would be desirable from a social point of view. On this view, the Philadelphia Convention rejected Edmund Randolph's "view that [presidential] self-dealing may be a problem" and instead opted for "a stronger presidency with the risk of occasional abuse." (98) The problem with the simplistic invocation of nemo iudex in this setting is that
   even a presidential self-pardon may not be solely an act of
   self-dealing ....Richard Nixon could have thereby spared Gerald
   Ford the odious task of issuing a pardon for his predecessor. While
   self-dealing may be a factor, the country may actually be a
   significant beneficiary of the self-pardon concept. (99)


This is, in effect, an invisible hand possibility theorem applied to presidential self-pardoning: the self-regarding incentives of presidents may align with the public good. A structural constraint on presidential self-pardons, on the other hand, might produce too few pardons, sacrificing social welfare on the altar of impartiality.

In this example, the critique of presidential self-pardons illustrates a methodological mistake about constitutional law. Some of the Constitution's provisions may rest on the logic of nemo iudex, as in the examples given above (although it is striking that these examples are mostly marginal or second-decimal rules). With respect to the issue of legislative salaries, for example, the main feature of the constitutional design is that legislators vote on their own salaries, whereas the asserted examples of nemo iudex safeguards, involving emoluments and the Twenty-Seventh Amendment's delay rule, just tweak the rules around the edges. Even insofar as the examples are valid, however, they cannot be generalized into a constitutional principle against, or a "structural distaste" for, self-dealing or self-judging. In some contexts, constitutional rulemakers follow nemo iudex. In others they do not, depending upon context-specific judgments about the direction and magnitude of competing political risks, and about relevant tradeoffs. Just as it is treacherous in the extreme to generalize principles of "federalism" from the Constitution's many, highly specific, and calibrated provisions for assigning powers and duties between states and the union, (100) so too any generalization of a structural nemo iudex principle would deliberately have to overlook central features of the constitutional design.

2. Presidential Pardons for Treason

At Philadelphia, the main thread of debate over the pardon power addressed whether the power should cover the offense of treason. Randolph moved that it should not, holding that "the prerogative of pardon in these cases was too great a trust" either because "[t]he President himself may be guilty"-the self-pardoning issue--or because "[t]he Traytors may be his own instruments." (101) Madison noted that lodging the power to pardon treason in the legislature would create problems of its own, but said that the power was "peculiarly improper for the President," (102) presumably because of Randolph's concerns over direct self-dealing or indirect cronyism. The Convention, however, voted down Randolph's motion and thereby included treason within the scope of the President's pardon power. Later commentators such as Hamilton and Justice Story justified that decision by appealing to institutional energy and activity levels. The "principal argument[]," in Hamilton's view, was that
   [i]n seasons of insurrection or rebellion, there are often critical
   moments when a well-timed offer of pardon to the insurgents or
   rebels may restore the tranquility of the commonwealth; and which,
   if suffered to pass unimproved, it may never be possible afterwards
   to recall. The dilatory process of convening the legislature, or
   one of its branches, for the purpose of obtaining its sanction to
   the measure, would frequently be the occasion of letting slip the
   golden opportunity. The loss of a week, a day, an hour, may
   sometimes be fatal. (103)


The implicit structure of this argument differs from the argument for presidential self-pardons. In the latter case, presidential self-interest is itself the spur that motivates a socially desirable level of activity, whereas here the presidency is assumed, for exogenous reasons, to be more active and decisive than the legislature. Yet the two arguments have a common feature: even if presidential power to pardon for treason creates risks of direct self-dealing or indirect cronyism, those risks are an unavoidable and acceptable byproduct of the institutional allocation of pardoning power that produces optimal levels of activity.

3. Legislative Qualifications

In some cases, the tradeoff between impartiality on the one hand and energy or motivation on the other is closely tied to the issue of institutional autonomy or independence. Justice Story's argument that each chamber of the bicameral Congress should have the authority to judge the qualifications of its members tied together independence and motivation by observing, in effect, that only institutional self-interest would provide the necessary motivation for institutional self-defense, and that institutional self-defense is necessary to the maintenance of an ongoing system of independent institutions that check and balance one another. In Justice Story's words,
   No other body, but itself, can have the same motives to preserve
   and perpetuate these attributes; no other body can be so
   perpetually watchful to guard its own rights and privileges from
   infringement, to purify and vindicate its own character, and to
   preserve the rights, and sustain the free choice of its
   constituents. (104)


4. Legislation, Partiality, and the System of Checks and Balances

Similar reasoning implies that Madison's famous nemo iudex argument in Federalist No. 10 must be understood in conjunction with his argument for checks and balances in Federalist No. 51. In the former, Madison argued that "[n]o man is allowed to be judge in his own cause, because his interest would certainly bias his judgment, and, not improbably, corrupt his integrity. With equal, nay with greater reason, a body of men are unfit to be both judges and parties at the same time.... " (105) Accordingly, Madison argued, legislative majorities would make biased and partisan decisions.

Madison's solution was that the effects of legislative partiality should be offset by structural devices of checks and balances. In the language of Madison's Federalist No. 51, such devices would make "[a]mbition ... counteract ambition," (106) harnessing the motivational energy of partial, biased, or self-interested officials to promote the power and independence of the institutions in which those officials served. Federalist No. 51 thus offers an invisible hand argument for competition among self-serving individuals and institutions, an argument ultimately derived from the Scottish Enlightenment theorists David Hume, Adam Ferguson, and Adam Smith, who influenced Madison. (107) Read in tandem, the point of Federalist No. 10 and Federalist No. 51 is that the cure for biased decisionmaking on the part of legislatures is not legislative impartiality, but rather equally biased decisionmaking by competing institutions. And the key argument for such a system of offsetting biases, according to Federalist No. 51, is that only the motivational energy supplied by "ambition" will suffice to protect institutions from mutual encroachment, thus ensuring the ongoing maintenance of a system in which separated powers check and balance one another.

Elsewhere, I have claimed that the Madisonian argument for checks and balances fails. (108) It specifies no mechanism for ensuring that institutions will pursue "their" interests, as opposed to the interests of the individuals or political parties who happen to staff them at any given time, (109) and equally fails to guarantee that competition among institutions will produce impartial lawmaking, liberty, efficiency, a stable system of interactions among autonomous institutions, or any other social or political good. Whatever the substantive merits of Madison's argument, however, its structure is conceptually important. It illustrates that multiplying, rather than attempting to reduce, violations of nemo iudex may be necessary to produce systemically valuable institutional energy and motivation-here the spurs of ambition, self interest, and institutional interest, which Madison assumed would work in harness to maintain the system of mutually checking institutions.

IV. THE CALCULUS OF POLITICAL RISKS: RULES OF THUMB

If nemo iudex is a misleading half-truth, what follows? At a minimum, "[f]acile invocation[]" (110) of nemo iudex should, one hopes, come to be widely seen as an embarrassing slip--the intellectual equivalent of burping at a dinner party. Chanting "nemo iudex" is never a sufficient argument; one must always go on to consider whether impartial decisionmaking is feasible and desirable in the relevant domain. Nemo iudex problems may afflict all possible arrangements, so that some fox or other must necessarily be placed in charge of some henhouse or other, and extrinsic considerations will have to be called into service to determine institutional arrangements. Alternatively, impartial decisionmaking may be overbalanced by other competing considerations, such as the direct costs of appointing an impartial decisionmaker, expertise, institutional independence, and institutional energy.

Clearing the intellectual ground in this way is at least a step forward. The constructive claim, however, is more difficult. In general, the theory of public law should attempt to make progress toward specifying the conditions under which rule designers should entrust officials or institutions with authority to act as judges in their own cause. The difficulty is that, although the competing political risks and the main tradeoffs can be stated in the abstract, it is not possible to say in the abstract how the balance should be struck. In particular contexts, with respect to different institutional problems, the calculus of political risks will come out differently; the questions are inherently local and are afflicted by severe empirical uncertainty. Nonetheless, it may be possible to suggest some pragmatic rules of thumb for coping with the tradeoffs and identifying, in a rough-and-ready way, conditions under which nemo iudex should be either discarded or honored, and how, if at all, it should be implemented. I will discuss three such rules of thumb: marginalism, optimizing, and the availability of substitute protections against self-dealing by officials or institutions.

A. Marginalism

Rule designers will rarely face an all-out choice between impartiality, as such, and some competing value, as such. Rather they will have to adopt rules and institutions in a political context in which some other rules and institutions are already settled. This is true even at constitutional conventions. Framers do not usually consider all institutions as up for grabs all at once, but instead accept some extant institutions for political reasons, and also consider institutional choices piecemeal, reaching agreement on some issues--for example, the composition of the legislature--before considering, for example, the powers of the courts.

Moreover, the extant institutions will often build in some protections against official self-dealing or biased decisionmaking. Where that is so, the problem facing rule designers is marginalist: the question is not whether impartiality as such is desirable; instead, the question is whether, given some extant set of institutional checks or precautions against official self-dealing, it is desirable to add further precautions. If each additional precaution produces diminishing marginal benefits and increasing costs to the other values discussed in Part III, precautions should be added just up to the point at which the marginal benefit equals the marginal cost. Needless to say, in the rough and tumble of rule design for real polities, the location of that point will usually be unclear, and rule designers will have to proceed by informed guesswork.

The Philadelphia debate over congressional salaries turned in part upon this theme. Given the structure of the new federal legislature, a decisive fraction of the participants believed that political checks would sufficiently deter self-dealing by federal legislators empowered to set their own salaries. Indeed, as we have seen, Sherman worried that the federal legislators would be subject to so much public pressure that they would set their salaries too low, rather than too high (111)--bending over backwards to demonstrate their lack of self-interest. In modern terms, given the robust political checks already in place, the Convention seems to have believed that extra precautions would have slight marginal benefits, and indeed might even have net costs. Whether or not that assessment was correct, the structure of the analysis illustrates that a marginalist approach can cut daunting problems clown to size and make design choices more tractable, even under conditions of grave uncertainty.

B. Optimizing

As a corollary of the marginalist approach, rule designers will ordinarily do best by optimizing across all relevant values in a given domain, rather than by maximizing any one of them. Where impartiality trades off against information, institutional autonomy, or institutional activity, it will rarely be the best strategy simply to choose one value wholesale and pursue it to the maximum possible extent, giving no weight to the competing values. Rather, on the assumption that pursuit of any one value has diminishing marginal benefits (or conversely that violation of any one value produces increasing marginal costs as the violation becomes increasingly severe (112)), it will usually be best to have some of each, rather than all of one and none of the other(s). This is not a conceptual claim, but a seat-of-the-pants empirical judgment about the shape of the marginal cost and benefit curves that institutional designers usually face.

Many of the institutions of the U.S. constitutional order are structured as optimizing compromises of this kind. We have seen, for example, that the structure of congressional authority over salaries pursues the "some of both" strategy. The core feature that legislators set their own salaries tends to promote institutional autonomy, but rules like the Twenty-Seventh Amendment and the Emoluments Clause temper that feature with collateral protections for impartiality. Such outcomes partly reflect conscious strategies of constitutional design, yet also result from the unpredictable interplay of political forces combined with a large dose of path dependence. The main point is not historical, but prescriptive: in many settings, modern constitutional designers could do far worse than to aim for roughly optimizing compromises of this sort.

C. Substitute Protections

In the example of congressional salaries, political checks against self-dealing already provided substitute protections. In other examples, however, rule designers can promote values that trade off against nemo iudex while simultaneously adjusting some other margin of institutional design to safeguard against the risks of self-dealing. Where there is a choice of alternative means by which to promote impartiality or suppress self-dealing, the nemo iudex approach--barring the decisionmaker from making decisions in which he has an interest-need not be the best solution, all things considered. As Landis argued in defense of the combination of functions in agencies, "[T]he fact that there is this fusion of prosecution and adjudication in a single administrative agency does not imply the absence of all checks. It implies simply the absence of the traditional check." (113)

Putting aside Landis's analysis of substitute protections in the administrative state as a subject worthy of separate treatment, I will illustrate the theme of substitute protections with two examples from the original Constitution. In the debate at the Philadelphia Convention over presidential pardons for treason, Wilson dismissed concerns about presidential self-dealing by arguing that "[p]ardon is necessary for cases of treason, and is best placed in the hands of the Executive. If he be himself a party to the guilt he can be impeached and prosecuted." (114) Tracking this argument in part, the Convention specifically excluded cases of impeachment from the scope of the pardon power. (115) In modern terms, the expected harm of presidential self-dealing through abuse of the pardon power is a function both of the risk that self-dealing will occur and of the magnitude of the harm if it does occur. Wilson's argument points out that the magnitude of the harm is limited; while the President might use the pardon power to immunize himself from criminal punishment, no amount of self-pardoning will keep him in office if a sufficient majority desires to remove him.

Another illustration involves the authority of the houses of Congress to "judge" the qualifications of their members and to expel members. As we have seen, Justice Story's argument for entrusting that authority to the respective chambers was that placing the authority anywhere else would create an unacceptable threat to institutional autonomy. The Philadelphia Convention, however, was concerned about the risk that cameral authority over qualifications would be abused by majority legislative factions in order to expel or otherwise oppress members of minority legislative factions--the same type of group self-dealing by legislative majorities that Madison classed as a violation of nemo iudex in Federalist No. 10. At the Convention, "Mr. Madison observed that the right of expulsion ... was too important to be exercised by a bare majority of a quorum: and in emergencies of faction might be dangerously abused. He moved that [expulsion would require a two-thirds vote of the chamber]." (116) Morris objected that a supermajority requirement "may produce abuses on the side of the minority. A few men from factious motives may keep in a member who ought to be expelled." (117) But the Convention unanimously approved Madison's amendment with one state divided. (118)

In both of these examples, the risk of self-dealing was addressed not by shifting decisionmaking authority to another official or institution, but by adjusting a different margin of institutional design--either the availability of impeachment for corrupt presidential pardons (an adjustment to the sanction for a biased decision) or the voting rule for legislative expulsions (an adjustment to the threshold necessary for a biased decision). The point generalizes. Nemo iudex arguments often assume, implicitly, that identifying a problem (the risk of self-dealing) necessarily implies a particular solution (transferring decisionmaking authority away from the biased decisionmaker). But in some institutional settings, other types of precautions or remedies are also available, and the assumption no longer holds.

Where other precautions or remedies are indeed available, the rule designer faces a choice and should adopt the solution that produces the greatest net benefits, not necessarily the solution that strictly minimizes the risk of self-dealing. As compared to a regime in which another institution decides on expulsion, Madison's supermajority solution may not be the one that strictly minimizes the risk of factional self-dealing in the legislative chamber. Given the benefits of institutional autonomy identified by Justice Story, however, the combination of cameral expulsion authority and a supermajority voting rule has a plausible claim to represent an optimal set of precautions--or at least a better set of precautions than any feasible competitor.

CONCLUSION

Billed as one of the law's great maxims, nemo iudex is in fact a shifty half-truth, useful at some times and in some settings, useless or positively misleading in others. Perhaps many grand legal principles have this characteristic when closely inspected. But it requires no such inspection to see that nemo iudex is contradicted by central structural features of our constitutional order. In our constitutional system, legislators shape their own elections and determine their own compensation, judges decide cases about their own salaries, and administrators rule on the validity of cases they themselves have brought under rules that they have written. In a system like that, the real puzzle is how nemo iudex maintains its grip on the legal mind. A well-rounded analysis should see the impartiality of decisionmakers as one institutional good among others, to be pursued, or not, as a larger calculus of institutional optimization suggests.

(1) The principle is classically articulated in the Justinian Codex, which includes a provision entitled "Ne quis in sua causajudicet vel sibi jus dicat" ("No one shall be judge in his own cause"). FRED H. BLUME, ANNOTATED JUSTINIAN CODE 3.5.1 (Timothy Kearley ed., 2d ed. 2008), http://uwacadweb.uwyo.edu/blume&Justinian/Book%203PDF/Book%203-5.pdf. As Kearley notes, the older Codex Theodosianus seemingly derives this principle from the rule that no man should be allowed to testify for himself. Id.; see 1 THE THEODOSIAN CODE AND NOVELS AND THE SIRMONDIAN CONSTITUTIONS 2.2.1, at 39-40 (Clyde Farr ed. & trans., 1952). For the history of the principle in medieval law, see D.E.C. Yale, Iudex in Propria Causa: An Historical Excursus, 33 CAMBRIDGE L.J. 80 (1974).

(2.) Gutierrez de Martinez v. Lamagno, 515 U.S. 417, 428 (1995).

(3.) See, e.g., Nev. Comm'n on Ethics v. Carrigan, 131 S. Ct. 2343, 2348 (20ll) (noting the principle that legislators should recuse themselves from voting on questions in which they have a personal interest because "the fundamental principles of the social compact [forbid] ... any man to be a judge in his own case" (quoting THOMAS JEFFERSON, A MANUAL OF PARLIAMENTARY PRACTICE FOR THE USE OF THE SENATE OF THE UNITED STATES 31 (1801))); Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 886 (2009) ("[N]o man is allowed to be a judge in his own cause...."); Pounders v. Watson, 521 U.S. 982, 992 (1997) (Stevens, J., dissenting) (quoting Sacher v. United States, 343 U.S. 1, 36-37 (1952), which refers to "established judicial practice, which makes it unfitting for a judge who is personally involved to sit in his own case"); Aetna Life Ins. Co. v. Lavoie, 475 U.S. 813 (1986) (finding that a state supreme court justice's participation in a case violated the defendant's due process rights because the justice had created a precedent that would affect the outcome of a similar suit that he himself had filed); Tumey v. Ohio, 273 U.S. 510, 523 (1927) (holding that an individual's constitutional rights were violated when his case was heard before a judge who had a pecuniary interest in finding that individual guilty); Spencer v. Lapsley, 61 U.S. (20 How.) 264, 266 (1857) ("The act of Congress [at issue] proceeds upon an acknowledgment of the maxim, 'that a man should not be a judge in his own cause[]'....").

(4.) 3 U.S. (3 Dall.) 386, 388 (1798) (identifying "a law that makes a man a Judge in his own cause" as an example of an act "contrary to the great first principles of the social compact" that "cannot be considered a rightful exercise of legislative authority").

(5.) 5 U.S. (1 Cranch) 137, 176 (1803) ("To what purpose are powers limited, and to what purpose is that limitation committed to writing, if these limits may, at any time, be passed by those intended to be restrained?").

(6.) Dr. Bonham's Case, (1610) 77 Eng. Rep. 638 (C.P.) 652; 8 Co. Rep. 107a, 118a (Coke, C.J.) (holding that a college of physicians given statutory powers to punish unlicensed medical practice could not act as "judges, ministers, and parties" simultaneously). Justices sometimes cite the decision for some version of the nemo iudex maxim. See, e.g., Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 329 n.* (2007) (Scalia, J., concurring in the judgment) (citing Bonham's Case for the "ancient maxim 'aliquis non debet esse Judex in propria causa'--no man ought to be a judge of his own cause"); Arnett v. Kennedy, 416 U.S. 134, 197 (1974) (White, J., concurring in part and dissenting in part) (citing Bonham's Case for the "first principle: '[N]o man shall be a judge in his own cause'").

(7.) For both accuracy arguments and fairness arguments in support of the principle, see, for example, THOMAS HOBBES, LEVIATHAN 102 (Michael Oakeshott ed., Macmillan 1946) (1651):
      [S]eeing every man is presumed to do all things in order to his
   own benefit, no man is a fit arbitrator in his own cause; and if he
   were never so fit; yet equity allowing to each party equal benefit,
   if one be admitted to be judge, the other is to be admitted also.

      For the same reason no man in any cause ought to be received as
   arbitrator, to whom greater profit, or honour, or pleasure
   apparently ariseth out of the victory of one party, than of the
   other: for he hath taken, though an unavoidable bribe, yet a bribe;
   and no man can be obliged to trust him.


(8.) See, e.g., Justin Levitt & Michael P. McDonald, Taking the "Re" Out of Redistricting: State Constitutional Provisions on Redistricting Timing, 95 GEO. L.J. 1247, 1255 n.38 (2007) (referring to a number of states, including Indiana, Michigan, Wisconsin, and California, whose constitutions vest the duty of apportionment in those states' legislatures).

(9.) See H.W. DODDS, PROCEDURE IN STATE LEGISLATURES 3 (1918) ("The right to judge of the elections and qualifications of its own members is expressly conferred upon each house by the constitutions of forty-six states.").

(10.) See, e.g., Ronald E. Weber, Presidential Address, The Quality of State Legislative Representation: A Critical Assessment, 61 J. POL. 609, 610 (1999) ("[M]ost state legislatures set their own salaries and benefits....").

(11.) See infra notes 92-100 and accompanying text.

(12.) See infra note 97.

(13.) See infra Section II.E.

(14.) See infra notes 79-87 and accompanying text.

(15.) See, e.g., Paul D. Carrington, Judicial Independence and Democratic Accountability in Highest State Courts, 61 LAW & CONTEMP. PROBS., Summer 1998, at 79, 97 (noting that the "highest state courts in several states have a constitutional power to enact procedural rules for use in inferior courts," that some state courts have held that "procedural matters are beyond the kin of legislatures and are [those courts' own] exclusive responsibility," and that "[i]n some states, the power to enforce standards of judicial conduct is vested within the judicial branch").

(16.) See, e.g., Pierson v. Ray, 386 U.S. 547 (1967) (explaining that judges enjoy absolute immunity from suit for actions taken in the exercise of their judicial functions).

(17.) See infra Subsection III.B.1.

(18.) See Adrian Vermeule, Precautionary Principles in Constitutional Law, 4 J. LEGAL ANALYSIS 181, 217 (2012).

(19.) For discussion of these versions of the maxim, see GEORGE FREDERICK WHARTON, LEGAL MAXIMS WITH OBSERVATIONS AND CASES: IN TWO PARTS 101-02 (London, Law Times Office, 2d ed. 1892).

(20.) See, e.g., Tumey v. Ohio, 273 U.S. 510, 523 (1927).

(21.) See infra notes 80-81 and accompanying text.

(22.) 28 U.S.C. [section] 455(a) (2006) ("Any justice, judge, or magistrate judge of the United States shall disqualify himself in any proceeding in which his impartiality might reasonably be questioned."); see Liteky v. United States, 510 U.S. 540, 548 (1994) ("Quite simply and quite universally, recusal [is] required whenever 'impartiality might reasonably be questioned.'" (quoting 28 U.S.C. [section] 455(a) (1974))).

(23.) THE FEDERALIST NO. 10, at 47 (James Madison) (Clinton Rossiter ed., 1999) ("No man is allowed to be a judge in his own cause, because his interest would certainly bias his judgment, and, not improbably, corrupt his integrity.").

(24.) Id. ("With equal, nay with greater reason, a body of men are unfit to be both judges and parties at the same time; yet what are many of the most important acts of legislation but so many judicial determinations, not indeed concerning the rights of single persons, but concerning the rights of large bodies of citizens? And what are the different classes of legislators but advocates and parties to the causes which they determine?").

(25.) 515 U.S. 417 (1995).

(26.) Id. at 448-49 (Souter, J., dissenting).

(27.) WHARTON, supra note 19, at 119 ("The rule [of nemo iudex] is inflexible, and as well the king as the commoner is subjected to it....").

(28.) There is a running debate among legal historians on the question of whether Bonham's Case counts as an example of judicial review of parliamentary legislation. For recent analyses, see PHILIP HAMBURGER, LAW AND JUDICIAL DUTY 622-30 (2008) ; and R.H. Helmholz, Bonham's Case, Judicial Review, and the Law of Nature, 1 J. LEGAL ANALYSIS 325 (2009).

(29.) Dr. Bonham's Case, (1610) 77 Eng. Rep. 638 (C.P.) 652; 8 Co. Pep. 107a, 118a. As to the college's power to levy fines, it appears that the college would receive a part of the fines and in this sense had a pecuniary interest. But that was a corporate pecuniary interest, not a direct individual one; presumably the fines would go to the college's treasury, not into the pockets of the members who judged the case. (When federal judges levy fines, of course, the money likewise goes into the treasury of the organization that pays their salaries.) And in any event, the point would not apply to the college's power to imprison, which for Chief Justice Coke and the plaintiff was at least as objectionable as the power to fine.

(30.) 515 U.S. at 449 (Souter, J., dissenting).

(31.) Thanks to John Goldberg for suggesting this comparison.

(32.) 14 U.S. (1 Wheat.) 304 (1816).

(33.) Id. at 345 (emphasis added).

(34.) David P. Currie, The Constitution in the Supreme Court: The Powers of the Federal Courts, 1801-1835, 49 U. CHI. L. REV. 646, 657 (1982).

(35.) Jeremy Waldron, The Core of the Case Against Judicial Review, 115 YALE L.J. 1346, 1400-01 (2006).

(36.) Nathaniel Persily, Reply, In Defense of Foxes Guarding Henhouses: The Case for Judicial Acquiescence to Incumbent-Protecting Gerrymanders, 116 HARV. L. Rev. 649, 674 (2002).

(37.) See Marc Lacey, Arizona Governor and Senate Oust Redistricting Leader, N.Y. TIMES, Nov. 2, 2011, http://www.nytimes.com/2011/11/02/us/chairwoman-of-arizona-redistricting-commission -ousted.html.

(38.) For an excellent overview, see Marshall J. Breger & Gary J. Edles, Established by Practice: The Theory and Operation of Independent Federal Agencies, 52 ADMIN. L. REV. 1111 (2000).

(39.) Neal Devins & David E. Lewis, Not-So Independent Agencies: Party Polarization and the Limits of Institutional Design, 88 B.U.L. REV. 459, 469-71, 477-87 (2008).

(40.) The classic account is Robert A. Dahl, Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker, 6 J. PUB. L. 279 (1957). For citations to the enormous follow-on literature, see, for example, Micheal W. Giles, Bethany Blackstone & Richard L. Vining, Jr., The Supreme Court in American Democracy: Unraveling the Linkages Between Public Opinion and Judicial Decision Making, 70 J. POL. 293 (2008).

(41.) Richard H. Pildes, Is the Supreme Court a "Majoritarian" Institution?, 2010 SUP. CT. REV. 103, 114-16.

(42.) See, e.g., Gary Lawson, The Rise and Rise of the Administrative State, lO7 HARV. L. REV. 1231, 1248-49 (1994) (originalist critique); Ilya Shapiro & Caitlyn W. McCarthy, Are Federal Agencies the Sole Judges of Their Own Authority?, REG., Summer 2011, at 4 (libertarian critique).

(43.) JAMES M. LANDIS, THE ADMINISTRATIVE PROCESS 91-113 (1938).

(44.) Id. at 96.

(45.) Id. at 97 (quoting ALBERT VENN DICEY, LECTURES ON THE RELATION BETWEEN LAW & PUBLIC OPINION IN ENGLAND DURING THE NINETEENTH CENTURY 369 (2d ed. 1914)).

(46.) See NEIL K. KOMESAR, IMPERFECT ALTERNATIVES: CHOOSING INSTITUTIONS IN LAW, ECONOMICS, AND PUBLIC POLICY 141 (1994) (explaining that judges are insulated from a large amount of information about the needs and desires of the public, and that the "distant judge" or the "more distanced administrative agency" may be less biased).

(47.) Jon Elster, Mimicking Impartiality, in JUSTICE AND DEMOCRACY: ESSAYS FOR BRIAN BARRY 112 (Keith Dowding, Robert E. Goodin & Carole Pateman eds., 2004).

(48.) 28 U.S.C. [section] 455(a) (2006).

(49.) See Susan B. Hoekema, Comment, Questioning the Impartiality of Judges: Disqualifying Federal District Court Judges Under 28 U.S.C. [section] 455(a), 60 TEMP. L.Q. 697, 698 (1987).

(50.) Id. at 713.

(51.) See KOMESAR, supra note 46, at 141.

(52.) Id.

(53.) See Saul Levmore, Efficiency and Conspiracy: Conflicts of Interest, Anti-Nepotism Rules, and Separation Strategies, 66 FORDHAM L. Rev. 2099, 2103 (1998); Dorit Rubinstein Reiss, The Benefits of Capture (Aug. 2, 2011) (unpublished manuscript), http://ssrn.com/abstract =1904023.

(54.) Liteky v. United States, 510 U.S. 540, 550 (1994).

(55.) Id. at 550-51 (emphasis added) (quoting In re J.P. Linahan, Inc., 138 F.2d 650, 654 (2d Cir. 1943)).

(56.) See, e.g., J. Gerald Hebert & Marina K. Jenkins, The Need for State Redistricting Reform To Rein in Partisan Gerrymandering, 29 YALE L. & POL'Y REV. 543, 558 (2011) ("State-level redistricting reform, particularly in the form of independent redistricting commissions, is absolutely necessary in order to fulfill the promise of government for the people, by the people.").

(57.) Persily, supra note 36, at 671-72.

(58.) Id. at 678-79.

(59.) Id.

(60.) 5 U.S.C. [section] 554(d)(2) (2006).

(61.) Id. [section] 554(d)(2)(C).

(62.) See Withrow v. Larkin, 421 U.S. 35 (1975); Marcello v. Bonds, 349 U.S. 302 (1955); FTC v. Cement Inst., 333 U.S. 683 0948).

(63.) 421 U.S. 35

(64.) Id. at 47.

(65.) Id. at 49-50 (quoting Richardson v. Perales, 402 U.S. 389, 410 (1971)).

(66.) LANDIS, supra note 43, at 91-113.

(67.) Id. at 101 (arguing that rulemakers must "take into consideration the assets as well as the liabilities" of the combination of functions in administrative agencies).

(68.) U.S. CONST. art. I, [section] 6. Over two hundred years later, the Twenty-Seventh Amendment created a requirement that salary increases could take effect only after an intervening election. See id. amend. XXVII. Given the high rate of reelection for federal legislators, it is an open question how much restraint the Amendment imposes. And agenda-setting power still lies with the legislators themselves, rather than other institutions; the legislators may thus time salary increases strategically, waiting until other political circumstances align to make reelection a safe bet.

(69.) 1 MAX FARRAND, THE RECORDS OF THE FEDERAL CONVENTION OF 1787, at 373-74 (1911).

(70.) Id. at 2.15-16.

(71.) Id. at 373.

(72.) Id. at 374

(73.) Nathaniel Ghorum argued that the fixed compensation could not be made "as liberal as it ought to he without exciting an enmity [against] the whole plan." Id. at 372. Benjamin Franklin objected on the ground that the Convention itself could be accused of self-dealing: "He wished the Convention to stand fair with the people. There were in it a number of young men who would probably he of the Senate. If lucrative appointments should be recommended we might be chargeable with having carved out places for ourselves." Id. at 427.

(74.) James Wilson "was [against] fixing the compensation, as circumstances would change and call for a change of the amount." Id. at 373.

(75.) Id. at 2-91.

(76.) Id.

(77.) See U.S. CONST. art. I, [section] 6 (providing that federal legislative salaries shall be "ascertained by Law, and paid out of the Treasury of the United States").

(78.) The Ineligibility and Emoluments Clauses state:
   No Senator or Representative shall, during the Time for which he
   was elected, be appointed to any civil Office under the Authority
   of the United States, which shall have been created, or the
   Emoluments whereof shall have been encreased during such time; and
   no Person holding any Office under the United States, shall be a
   Member of either House during his Continuance in Office.


Id. art. I, [section] 6, cl. 2.

(79.) Id. art. III, [section] 1.

(80.) See, e.g., United States v. Hatter, 532 U.S. 557 (2001); United States v. Will, 449 U.S. 200 (1980).

(81.) For a discussion of the rule, see Charles Gardner Geyh, Roscoe Pound and the Future of the Good Government Movement, 48 S. Tex. L. Rev. 871, 885 (2007).

(82.) Will, 449 U.S. at 225.

(83.) RICHARD H. FALLON, JR. ET AL., HART AND WECHSLER'S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 1049-82 (6th ed. 2009).

(84.) See Will, 449 U.S. at 217.

(85.) Id. (citing Evans v. Gore, 253 U.S. 245, 253 (1920)).

(86.) Expanding the lens beyond the U.S. federal court system to include state courts and the courts of other nations, some judiciaries have asserted powers to determine their own budgets or their own membership. Courts in several U.S. states have ordered legislatures to appropriate judicially specified amounts to fund the court system, asserting that the power to do so is inherent in or necessary for judicial independence. The leading case is Commonwealth ex rel. Carroll v. Tare, 274 A.2d 193 (Pa. 1971); for further discussion and citations, see Adrian Vermeule, The Judicial Power in the State (and Federal) Courts, 2000 Sup. CT. REV. 357, 382-87. In India, the Supreme Court has held that judicial independence requires the government to appoint the most senior judge of the court to the position of chief justice, despite the absence of any textual provision to that effect in the constitution. See In re Special Reference No. 1 of 1998, (1998) 7 S.C.C. 739, 760 (India).

(87.) Similarly, it has been argued that tighter congressional regulation of recusal practice by Supreme Court Justices--who under current norms have unreviewable discretion to rule on their own disqualification-would violate the separation of powers, in part because it would threaten to undermine the "integrity and professionalism" of the Court. Louis J. Virelli III, The (Un)constitutionality of Supreme Court Recusal Standards, 20ll WIS. L. REV 1181, 122-9.

(88.) U.S. CONST. art. 1, [section] 5, cl. 1.

(89.) 2 JOSEPH STORY, COMMENTARIES ON THE CONSTITUTION OF THE UNITED STATES [section] 831, at 295 (Boston, Hilliard, Gray & Co. 1833).

(90.) Adrian Vermeule, The Glorious Commander-in-Chief in THE LIMITS OF CONSTITUTIONAL DEMOCRACY 157 (Jeffrey K. Tulis & Stephen Macedo eds., 2010).

(91.) THE FEDERALIST NO. 70, supra note 23, at 391 (Alexander Hamilton). For later discussion, see, for example, TERRY EASTLAND, ENERGY IN THE EXECUTIVE : THE CASE FOR THE STRONG PRESIDENCY (1992).

(92.) For such a critique, see, e.g., Brian C. Kalt, Note, Pardon Me?: The Constitutional Case Against Presidential Self-Pardons, 106 YALE L.J. 779 (1996).

(93.) U.S. CONST. art. II, [section] 2, cl. 1.

(94.) Kalt, supra note 92, at 795.

(95.) U.S. CONST. art. I, [section] 6, cl. 2.

(96.) Id. amend. XXVII.

(97.) For the debate over the question of whether the Vice President may preside at his own impeachment, compare Michael Stokes Paulsen, Someone Should Have Told Spiro Agnew, 14 CONST. COMMENT. 245 (1997), arguing that he can, with Joel K. Goldstein, Can the Vice President Preside at His Own Impeachment Trial? A Critique of Bare Textualism, 44 ST. Louis U. L.J. 849 (2000), claiming that he cannot.

(98.) Robert Nida & Rebecca L. Spiro, The President as His Own Judge and Jury: A Legal Analysis of the Presidential Self-Pardon Power, 52 OKLA. L. REV. 197, 218 (1999).

(99.) Id. at 219.

(100.) John F. Manning, Federalism and the Generality Problem in Constitutional Interpretation, 122 HARM. L. REV. 2003, 2004 (2009) (arguing that "the specific means chosen to implement our form of concurrent sovereignty in fact define the concept of federalism and that, contrary to the Court's recent cases, there is no freestanding federalism").

(101.) 2 FARRAND, supra note 69, at 626.

(102.) Id. at 627.

(103.) THE FEDERALIST NO. 74, supra note 23, at 417 (Alexander Hamilton). Justice Story echoes Hamilton. See 3 STORY, supra note 89, [section] 2494, at 351.

(104.) 2 STORY, supra note 89, [section] 831, at 295.

(105.) THE FEDERALIST NO. 10, supra note 23, at 47 (James Madison).

(106.) THE FEDERALIST NO. 51, supra note 23, at 290 (James Madison).

(107.) Roy Branson, James Madison and the Scottish Enlightenment, 40 J. HIST. IDEAS 235 (1979).

(108.) ADRIAN VERMEULE, THE SYSTEM OF THE CONSTITUTION 38-43 (2011).

(109.) See Daryl J. Levinson, Empire-Building Government in Constitutional Law, 118 HARV L. KEY. 915, 927 (2005); Daryl J. Levinson & Richard H. Pildes, Separation of Parties, Not Powers, 119 HARV. L. REV. 2311, 2324 (2006).

(110.) Waldron, supra note 35, at 140l.

(111.) 2 FARRAND, supra note 69, at 291.

(112.) See Yair Listoldn, Taxation and Marriage: A Reappraisal (Yale Law Sch., Pub. Law Working Paper No. 247, 2012), http://ssrn.com/abstract=2070171.

(113.) LANDIS, supra note 43, at 98.

(114.) 2 FARRAND, supra note 69, at 626 (emphasis added).

(115.) See U.S. CONST. art. II, [section] 2, cl. 1.

(116.) JAMES MADISON, NOTES OF DEBATES IN THE FEDERAL CONVENTION OF 1787, at 431 (bicentennial ed. 1987) (emphasis added).

(117.) Id.

(118.) Id.

AUTHOR. John H. Watson, Jr. Professor of Law, Harvard Law School. For helpful comments, thanks to Richard Fallon, John Goldberg, John Manning, Martha Minow, Jonathan Rose, Peter Schuck, Steve Shavell, David Strauss, Mark Tushnet, and participants in the Harvard Public Law Workshop and the Harvard Law School Conference on Political Risk and Public Law. Thanks to Samantha Goldstein for excellent research assistance.
COPYRIGHT 2012 Yale University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Watson, John H., Jr.
Publication:Yale Law Journal
Date:Nov 1, 2012
Words:15183
Previous Article:Welfare and rights before the movement: rights as a language of the state.
Next Article:Judicial capacity and the substance of constitutional law.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters