Dual Office Holding and the Constitution: A View From Hayburn's Case

Mark Tushnet

The Supreme Court's January 1989 decision in Mistretta v. United States rejected a constitutional challenge to the manner in which the United States Sentencing Commission was composed.[1] The constitutional challenge, premised on the principle of separation of powers, had several elements. As Justice Blackmun's opinion for the Court noted, one of these elements recalled the arrangements brought into question early in the history of the federal judiciary, for the Sentencing Commission, described as an independent agency located in the judicial branch, used the talents of sitting federal judges by making three of them members of the Commission. Because the Commission was not a court but was instead an agency of the United States, the federal judges who serve on the Commission hold two positions under the United States, one as judges and the other as Commissioners. This dual office holding was challenged as a violation of the Constitution. Justice Blackmun's opinion noted that a similar issue had arisen in connection with Hayburn’s Case, 2 U.S. 409 (1792), in which Congress asked judges of the Circuit Courts to serve as commissioners for the determination of certain questions regarding entitlements to pensions for service during the Revolutionary War. Although the Justices of the Supreme Court, sitting as circuit judges, held that the underlying statute was unconstitutional, most of them agreed that Congress could require them to serve as commissioners, not as judges.

This paper examines the constitutional terrain in which the Justices located the problem in Hayburn's Case, in an attempt to understand the distinction they drew between their constitutionally limited duties as judges and the more expansive possibilities for action in their individual capacities. At the outset, though, it should be noted that what we are dealing with may perhaps best be described as a 500-piece jigsaw puzzle, of which we have before us only a handful of pieces from which we are to determine what the overall picture is like. Under the circumstances, the best I can hope to do is identify certain aspects of the conceptual universe in which the federal judiciary was located, which shed some light on the problem of dual office holding and therefore some light on the conception of judging embedded in the Constitution.

A. The Constitutional Issues Addressed

The story behind Hayburn's Case is well-known.[2] The Invalid Pensions Act of 1792 was a public assistance program[3]designed to help the families of soldiers injured in the Revolutionary War adjust to the dislocations caused both by their injuries and by the economic disruption that occurred in the war's aftermath.[4] The Act suspended a previous statute of limitations on claims by soldiers' widows and orphans for two years, and allowed disabled soldiers and seamen to receive a pension. The applicant had to present the circuit court of his residence with a certificate or affidavits attesting to his disability. The circuit court, which was required to sit for at least five days to receive pension applications, would, after receiving the required documents, certify the degree of disability to the Secretary of War, along with a determination of the appropriate pension. The Secretary could then determine whether there had been "imposition or mistake" and withhold the pension recommended by the circuit court. Finally, the Secretary of War would report the list of applicants he found ineligible to Congress, which then would appropriate money for the pensions of the eligible applicants. Congress apparently chose this system for administering the pension scheme for sensible reasons. In modern times the duties given to the circuit courts would be assigned to some bureaucracy, either already existing or created for the purpose. In the early republic, though, the administrative apparatus of the new national government was rudimentary to say the least.[5] The circuit courts had the advantage of being already in place throughout the nation, even though they had been created for other purposes. In addition, one of the fears expressed during the debates over the ratification of the Constitution was that the new government would become a powerful source of patronage, and would therefore come to displace the states as the primary locations of citizen identification. In Federalist 45 Madison had responded to this fear by saying that "the number of individuals employed under the Constitution... will be much smaller than the number employed under the particular States."[6]

One way to minimize the number of national executive officials, of course, was to give multiple duties to the ones that were created.[7] Further, on one obvious interpretation of the Act the division of authority between the circuit courts, located throughout the country, and the Secretary of War, located in the nation's capital, made a great deal of sense. The circuit courts would determine the degree of disability by examining the applicant personally or by evaluating affidavits that could be produced readily in the district of the applicant's residence; the Secretary of War, in turn, would examine the records of the military forces to determine whether the applicant had in fact served during the Revolutionary War, "imposition or mistake" thus being defined as fraud or mistake with respect to service rather than with respect to disability. Finally, in proposing a pension system to Congress, Secretary of War Henry Knox was concerned that applicants for pensions had to be examined skeptically because it would be "easy, from the influence of humanity, to obtain plausible certificates, even from men of good character."[8] Federal judges, certainly men of good character, might be able to resist the pull of humanity better than any other possible examiner of pension applicants.

The judges of the circuit courts responded to the Invalid Pensions Act by holding it unconstitutional. The Justices of the Supreme Court, who were required by statute to serve as judges in the circuit courts, were already unhappy with the burdens that circuit riding placed on them,[9]and did not find the new duties under the Act at all attractive. The reasons for finding the Act unconstitutional varied slightly among the circuit courts, though they all relied on concepts of separation of powers. The circuit court for New York, consisting of Chief Justice John Jay, Associate Justice William Cushing, and Judge James Duane, ruled on April 5, barely two weeks after the Act had been adopted. The court said that the duties assigned to it under the Act were not judicial, as was shown by the fact that

On April 18 the judges of the circuit court of Pennsylvania--Associate Justices James Wilson and John Blair, and Judge Richard Peters--wrote President Washington of their decision not to "proceed" under the Act, because "the business directed by this act is not of a judicial nature" and because the judgments of the court

Travel being what it was, it took longer for the circuit court for North Carolina to register its objections.[11] On June 8, Associate Justice James Iredell and Judge John Sitgreaves wrote Washington that the Act may have conferred a "power not in its nature judicial" on the circuit courts but in any event that the possibility of revision of the decisions of the courts by the Secretary of War "subjects the decision of the court to a mode of revision which we consider to be unwarranted by the constitution," for all forms of appellate review of judicial decisions had to be done by judges with the guarantees of tenure provided in Article III.[12]

The circuit courts rested their objections, then, on two grounds. Since 1792 it has become clear that neither ground is entirely well-founded. To the objection that the duties under the Act were not 'judicial in nature," one could respond that the courts were being asked to make simple factual determinations, of the degree of disability and the amount of an appropriate pension, and that these determinations are indistinguishable in principle from a wide range of decisions made by judges exercising the judicial power of the United States.[13] Further, even though they did not mention this difficulty, to the extent that the judges were concerned about the fact that the proceedings under the Act were non adversarial, one could respond first that at this early stage in the development of constitutional law they need not have defined the judicial power to require the presence of full-fledged adversariness in all instances and second that the power of judges to issue certificates of naturalization has historically occurred in nonadversarial proceedings without being held to violate Article III.[14]

The circuit courts' second objection was that their decisions were subject to revision by the Secretary of War and by Congress. As to the former, the answer seems clear. If the Secretary's power to refuse to place an applicant on the list sent to Congress because of "imposition or mistake" were interpreted to mean that the Secretary could so act only in cases where the applicant had not served in the armed forces during the Revolutionary War--an issue not determined by the circuit courts--there would be no executive revision whatsoever.

The answer to the objection based on Congress's power to refuse to appropriate money for some or all of the pensions is more complicated. Roger Taney, who served as Chief Justice from 1836 to 1864, drafted an opinion holding that the Supreme Court could not take jurisdiction over appeals from the Court of Claims; one of his reasons for the constitutional difficulty was the fact that Congress had the unreviewable power to decide whether to appropriate money to pay judgments of the Court of Claims.[15] The Court itself, in a later case, dismissed an appeal from the Court of Claims after Congress had repealed the statute on which the respondent had received a judgment and directed that no such judgments be paid.[16] The Court subsequently avoided deciding the constitutional questions posed by a provision that Congress had to appropriate separately funds to pay judgments over $100,000.[17] With respect to that provision, Justice Harlan wrote that an historical record showing that Congress had refused to appropriate money to pay judgments only fifteen times in seventy years established the justiciability of decisions by the Court of Claims.[18] Of course the judges in 1792 could not rely on that sort of historical record, but one would think that the political pressures that led Congress to establish the pension scheme would operate effectively to ensure that the pensions, once determined by the courts and the Secretary of War, would be paid. The Supreme Court itself never ruled directly on these constitutional objections to the Act of 1792. The Act was amended within a year to remove the objectionable features,[19] although only by authorizing the district judge to appoint commissioners to do the work, thereby beginning to create the kind of national bureaucracy that had been discussed during the ratification debates.[20] When the full court was presented with the constitutional question, the judges backed away from their previous decisions on circuit to find the Act unconstitutional if interpreted to give them the claims-inspection tasks as commissioners. The Justices of the Supreme Court, sitting as circuit judges, though, did more than express their views on the constitutionality of the Act. As Justice Blackmun put it in his 1989 Mistretta opinion, Jay and Cushing "believed that individual judges acting not in their judicial capacities but as individual commissioners could exercise the duties conferred upon them by the statute."[21] Analyzing the distinction between judicial capacity and individual role is the topic of the remainder of this paper.

B. The Statutory Issue Addressed

The belief that the judges could sit as commissioners rested on two propositions, a question of statutory interpretation that the judges discussed and an additional constitutional question regarding the permissibility of dual office holding that they did not discuss.[22] The statutory difficulty is that the Act imposed its duties on the "circuit courts." Had it imposed the duties on the judges of the circuit courts, or even more cleanly on the present judges of the circuit courts, the statute could be read to designate those people as what the judges ended up calling "commissioners," that is, bureaucrats for the purpose of administering the Act. In New York the judges understood the statutory difficulty but finessed it. As they saw it, the Act appointed commissioners by official instead of personal descriptions." Having been so designated, the individual judges believed themselves to be commissioners and "therefore" to be "at liberty to accept or to decline that office." Because

The judges in North Carolina were more circumspect, in part perhaps because at the time they wrote they had not yet been asked by any applicant to perform any duties either as judges or as commissioners. They too praised Congress's "benevolence" and spoke of their own "feelings as men for persons whose situation requires the earliest as well as the most effectual relief," but had "great doubts" that they "could be justified in acting under this act personally in the character of commissioners during the session of a court." Their concern was that the Act "appears" to give power "to the court only, and not to the judges of it." Implicitly countering one canon of statutory construction, that statutes of this sort should be construed to accomplish their benevolent purposes, with another one, the judges noted that the effect of acting as commissioners would be to "draw [J money out of the public treasury," and that their designation as commissioners should therefore be authorized by clearer language than that used in the Act. They did say that, once an application to them had been made, they might give more consideration to the view that they could act as commissioners.[24]

Finally, two of the judges in the district of Connecticut acted as commissioners "notwithstanding some objections." A newspaper in York, Pennsylvania, praised these judges' "candor and indulgence in proceeding to the laborious task of examining the claims of the numerous applicants for pensions, a task which, in their opinion, their duty does not require them to undertake." This, it said, did "great honor to their humanity and compassion."[25]

One of the judges in Connecticut was Associate Justice James Iredell, who had expressed misgivings when in the North Carolina circuit court. He elaborated on his reasons for acting as a commissioner, which he did only after "mature reflection" had overcome "considerable hesitation." His concern was whether the statute could fairly be interpreted to allow the judges to exercise "the authority individually... out of Court." He began by stressing the use in the statute of the term "Circuit Court."

Justice Iredell then launched into an extremely ingenious bit of statutory interpretation, which, but for the benevolent purposes to which it was put, might seem more than a little hypertechnical. Justice Iredell found other language in the Act that led "to a very probable supposition that Congress may have contemplated it as a personal rather than a judicial exercise of power."

At one point the Act mentioned the District Judge, which

Of course Congress could not have "meant otherwise" when dealing with the circuit courts. Justice Iredell had now found the statute "equivocal," and thought it appropriate to adopt the construction of the Act that would support it. Similarly, in directing that applications be received for at least five days, the statute provided that "it shall be the duty of the judges... to remain at the places where the said courts shall be holden, five days, at the least." To Justice Iredell, phrasing this requirement as imposing a duty on the judges, rather than directing the Court to sit for five days, further brought out the ambiguity in the Act. Consider, he said, a court that concluded its business in three days. The Act required them to remain for two more days. If they did so, they would be fulfilling a "personal trust" imposed on them as individuals, for their duties as judges would have been concluded. Next Justice Iredell mentioned the sloppiness of legislative drafting "where a Legislature are employed in transacting in a very short time business of the most intricate and important nature." It would have been better to designate the judges in their individual capacities, but, given the fact that circuit riding meant that Congress could not know in advance which Justice of the Supreme Court would be attending which circuit court, it would have taken a careful draftsman to figure out how to identify the judges individually.[27]

In 1794 the Supreme Court decided the case of United States v. Yale Todd.[28] Sitting as commissioners, Chief Justice Jay, Justice Cushing, and Judge William Law advised that Todd should receive a pension of $100 a year. Todd received his pension from the Treasury, and then was sued by the Attorney General for return of the payment on the ground that it was not lawfully authorized. The Supreme Court agreed with the position of the United States that, as the pleadings put it, "the.., judges sitting as commissioners and not as a circuit court had [no] power and authority...so to order and adjudge of and Concerning the premises." The Supreme Court did not issue an opinion in Yale Todd, and we therefore cannot know whether the judges lacked power because the statute did not purport to authorize them to act as commissioners or because, even if it did, it was unconstitutional, although the phrasing of the pleadings suggests that the statutory ground played the major role.

Several provisions in the Constitution bear on the question of dual office holding. Article I, section 6, clause 2, the so called "incompatibility clause," addresses the issue directly:

This provision, which bars members of Congress from serving in the executive branch and bars members of the executive branch from serving in Congress, might be taken to approve, or at least not invalidate, service by judges in Congress or the executive branch, according to the principle expression unis est exclusio alterius.[29] In addition, the provisions regarding the ability of Congress to alter the salaries of the President and of judges suggest that the Framers thought that judges were less susceptible to certain types of corruption than were ordinary politicians. Yet, I believe, the arguments available from the framing ultimately have a gap that can be filled only by making assumptions about judges that are not easily reconciled with the basic presuppositions of the Framers' political thought.

A. The Incompatibility Clause

The primary purpose of the incompatibility clause is obvious. It is designed to avoid that bane of political life in civic republican theory, "corruption." Corruption, in this context as in others, took two forms. First, there is dependency. Members of Congress who also served in the executive branch might find themselves torn between their desire to advance their executive branch careers, which would make them dependent upon the President, and their desire to retain their electoral office. They could. reconcile these desires by using their executive positions to enhance their political power as legislators, for example, by dispensing patronage to their constituents, but, given the President's role in the executive hierarchy, only if their use of patronage were approved by the President. By disbursing some portion of their executive assets to their constituents, thereby corrupting the constituency as well, members of Congress could enhance their long-term executive and legislative assets. This version of the concern for corruption as dependency, though, has some anomalies. It is, of course, quite symmetrical. I have presented the concern in the form of distortion of executive action in the service of legislative goals. But, corruption as dependency could also occur through the distortion of legislative action in the service of executive goals. To advance their executive branch careers, members of Congress could enact legislation that promoted the narrow goals of the executive branch, buying off opposition by enacting legislation that conferred benefits, unrelated to the executive's narrow goals, on the opponents. The symmetry of the arguments about corruption as dependency suggests that, without an additional theory to account for which force would be more powerful, dual office holding might be self-limiting, with those members of Congress dependent on the executive branch cancelled entirely by the members of the executive branch dependent on Congress.[30]

In addition, corruption as dependency accounts for only part of the incompatibility clause, the outright prohibition on dual office holding. In the debates over the Constitution, this prohibition was completely uncontroversial. What concerned the Framers was the first part of the clause, barring members of Congress from positions created or enhanced during their terms of office. Here the concern was for a second type of corruption, the use of public offices to enhance the personal wealth of office holders. An ambitious man might seek election to Congress in order to create a position for himself to occupy after his brief period of service in the legislature.[31] Or, members of the House of Representatives might pay off their allies in the Senate by creating positions for them.[32]

The incompatibility clause guarded against this form of corruption, but only imperfectly. After all, as opponents of the Constitution noted, nothing in the clause barred a member of Congress from occupying a position already in existence during his term of office. Members could' manipulate the occupants of existing positions to create vacancies to which they would then be appointed.[33] The risk of this sort of corruption might be lessened by certain structural constraints. For example, if "term of office" meant the entire period of service of a member, the longer the term--either by constitutional design, as with the Senate's six-year term, or through the practice of reelection--the less attractive this strategy would be, for more positions would have been created during the member's term. In addition, the ban on enhanced emoluments meant that the executive branch position would have to be one that was more attractive at the moment of entry into Congress, which would further limit the strategy. And, to the extent that a member might hope that his former colleagues would reward his service in Congress by increasing the pay after he left Congress, the implicit bargain could not be enforced, thereby introducing some risk into the strategy of seeking election in order to occupy an executive branch position.

These constraints on corruption as office seeking seemed inadequate to some at the Convention, for the initial version of the incompatibility clause barred members of Congress from taking an executive branch position for one year after they left Congress, as well as during their term of service in Congress. Hamilton, who recognized the danger of dual office holding, opposed this broader exclusion. As he saw it, "take mankind in general, they are vicious." People are motivated by a combination of ambition and interest, and the prospect of an executive branch position was one of the motivations ambitious and self-interested people--all there were, after all--would have for serving in Congress.[34] Madison proposed the emoluments clause as "a middle ground" that would encourage legislative service without running the danger of a proliferation of unnecessary or unnecessarily expensive offices.[35] The incompatibility clause in its final version, then, combined the civic republican concern for avoiding corruption with the liberal recognition that people were moved primarily by self-inter-. est (of which ambition was a subdivision). In this it mirrored the structure of the Constitution as a whole.

Having seen the two forms of corruption that the incompatibility clause guarded against, we can examine the possible grounds for exempting federal judges from a similar prohibition.[36] One ground might be that the anticipated length of service of federal judges was so great that even self-interested and ambitious people would not rationally calculate that their long-term goals could be met first by serving in the judiciary and then moving on to an executive or legislative branch position.[37] This might reduce the risk of corruption as office seeking, but it is not responsive to the problem of corruption as dependency. Consider, though, the federal judge who simultaneously serves in an executive branch position. In what sense is that judge dependent on the President? As the Court in Mistretta pointed out, the constitutional guarantees of tenure and salary mean that a judge who displeases a President might lose the executive branch position but otherwise can suffer no retaliation.[38]

Corruption, though, is not avoided simply by establishing a structure that makes it possible for someone to be independent. Dual office holding poses the risk that the office holder will shade his or her judgments in the service of Congress or the President. Here, finally, we come to what seems to me the only substantial difference between legislators and executive officials on the one hand and judges on the other. As Hamilton put it in one of the most celebrated passages in Federalist 78, the judiciary has "neither force nor will but merely judgment."[39] It would appear, then, that as a matter of definition judges are not susceptible to the corruption of dependency. A similar definitional move occurs at the end of the same paper, in Hamilton's defense of life tenure for federal judges. "To avoid an arbitrary discretion in the courts," Hamilton wrote, judges must be bound by "strict rules and precedents," which "must unavoidably swell to a very considerable bulk and must demand long and laborious study to acquire a competent knowledge of them." Few people would have the necessary skill, "and making the proper deductions for the ordinary depravity of human nature, the number must be still small of those who unite the requisite integrity with the requisite knowledge."[40] Federal judges will simply be people of sufficient integrity to guarantee that they need not be barred from dual office holding in order to avoid the risk of corruption as dependency.

At this point, though, we have come up against one of the most fundamental difficulties in the Constitution's attempt to reconcile civic republicanism and liberalism. Structures get us a long way toward a virtuous government made up of vicious people, but at crucial points we apparently must simply assume that people of integrity will occupy at least some positions in the government. I will return to this difficulty after considering another structural contrast between federal judges and other officials of the national government.

B. The Salary Guarantee

Federal judges are protected in Article III against a reduction in their salaries. There is another salary provision in the Constitution. Article II, section 1, paragraph seven provides that

Originally Article III would have similarly barred increases as well as decreases in judicial salaries, but the ban on increases was struck by a vote of six states to two.[41]

On the face of it, eliminating the bar to salary increases is entirely sensible. The President was to serve a four-year term, while the judges had lifetime appointments. The longer the term was, the more vulnerable the occupant of the position was to fluctuations in the value of money. A ban on salary increases for federal judges would make it impossible for Congress to respond to changing economic circumstances.[42]

Yet, giving Congress that power did create some risks. In discussing the salary provision for the President, Hamilton said that it meant that Congress could neither "reduce him by famine" by reducing the President's salary nor "tempt him by largesses" by increasing it. "They can neither weaken his fortitude by operating upon his necessities; nor corrupt his integrity by appealing to his avarice."[43] The ban on salary reduction for judges did indeed avoid "famine." Governor Morris, moving to strike the ban on salary increases for judges, contended that "this would not create any improper dependence in the Judges." Madison, though, responded that a situation in which there would be "some dependence" was troublesome.[44]

And "some dependence" there would surely be. Consider, for example, the problems faced by federal judges during periods of relatively rapid inflation. Congress will be concerned with a range of public issues, some related to inflation and others unrelated to it. Somehow the federal judges have to get the attention of a Congress with many other things to do. One attention-getting device is to act visibly in ways showing that the judges are basically on Congress's side. Even more dramatically, consider a Congress desirous of getting the courts to rule in a particular way. Just as it could tempt the President "by largesses," so it could tempt the judges, offering them substantial increases in salaries as part of an implicit deal regarding what the judges would then do. The judges might resist the temptation, or they might renege on the implicit deal, but then, so could a President, whom the Constitution had to hedge around with a ban on salary increases.

It is possible, of course, that the risk of this sort of behavior by judges was low enough to be acceptable, particularly in light of the difficulty of devising a salary provision that took account of the judges' lifetime terms.[45] We might wonder, though, why the identical risk was too great in the case of the President, again putting aside the greater ability to guard against the risk in that instance. As with the absence of a ban on dual office holding for federal judges, it seems likely that the Framers assumed that judges were somehow different from ordinary politicians. Their technical training, and the fact that their power lay in exercising judgment rather than force or will, meant that they were simply less susceptible to corruption than ordinary politicians. I will consider this assumption, and its implications for our understanding of the structure of the Constitution, in the final section of this paper.

Recent scholarship has directed our attention to the civic republican assumptions that underlie the Constitution, assumptions that, we are told, were held by the Framers' generation and are, in any event, normatively attractive.[46] As we have seen, though, the more traditional account of the Framers as liberal individualists is also accurate. The Framers' generation, that is, was both liberal and republican.[47] Because the fundamental assumptions of civic republicanism and liberalism are incompatible, the structure of the Constitution, which attempts to incorporate both sets of assumptions, is bound to be awkward. This awkwardness is apparent in The Federalist Papers, whose discussion of certain issues unrelated to dual office holding illuminates that problem as well.

Consider first the basic problem of dual office holding when judges are involved. Federal judges might have been used as commissioners in pension cases for a number of reasons, including efficiency and avoiding patronage. Among those reasons might also be the republican one of utilizing the federal judiciary to demonstrate visibly the virtue of the national government in dispensing justice, both in deciding cases and in the broader domain of public policy represented by the pension statutes.[48] Yet, we should recall that the judges were employed in the first place in part to avoid the susceptibility of ordinary "men of good character" to the "influence of humanity," and then compare that to the judges' willingness, out of concern for the "exceedingly benevolent" purposes of the statute, to act as commissioners. This contrast suggests that, at least in situations where the judges' distinctive technical abilities were not directly implicated,[49] they were not all that different from ordinary men of good character.[50]

In Ralph Lerner's analysis of the judiciary as expositors of republican virtue, technical ability plays a crucial role.

As an exposition of the Framers' assumptions, Lerner's is persuasive. Yet, he offers little reason to explain why technical ability as a limitation to the power of the courts solves the problems of corruption, or susceptibility to corruption, that concerned the Framers.[52] What we might call today the socialization of judges into the professional culture is assumed to constrain them from corruption in contrast to ordinary politicians, who are socialized into either a purely political culture or into the cultures of diverse non-technical occupations and professions. What this overlooks, though, are two possibilities. First, as the judges' benevolent instincts in administering the pension act suggest, the technical and professional culture may not be strong enough to support the kind of fortitude that Lerner's analysis requires. Second, the technical culture of lawyers itself contains the potential for corruption, at the least in the service of the profession itself and perhaps more generally in the service of the strata of society from which lawyers are likely to be drawn.

Lerner offers a reading of The Federalist in which virtue prevails in the judiciary because of ungrounded assumptions about the impact of technical training on judges. The same kinds of difficulties pervade The Federalist and a brief examination of some other important aspects of its argument will shed further light on the problem of dual office holding. Madison's classic discussion of the virtues of an extended representative democracy in Federalist 10 begins by saying that a representative democracy

If we could be sure that the representatives would be people of that description, the case for representative democracy would be easy. But, as Madison immediately notes, there are no such guarantees.

Madison then argues that such people would have a harder time of it in an extensive republic. Organizing a vicious faction is more difficult in a larger republic, both because the individual districts will be larger so that to "obtain the suffrages" by corruption and the like will be more difficult and because combining the interests of numerous districts is likely to be difficult for technical reasons.

Already we can see some difficulties in Madison's account: If members of the electorate whose votes are not obtained by corruption are "more free" and are therefore "more likely to center on men who possess the most attractive merit and the most diffusive and established characters,"[55] one wonders why representative rather than direct democracy is appropriate. Madison's later discussion of federalism compounds the difficulty. In arguing that those who are attached to their state governments need not fear encroachments by the new national government, Madison says that the people who are selected for office in the national government are likely to have extensive contacts with the states, so much so that "a local spirit will infallibly prevail much more in the members of Congress than a national spirit will prevail in the legislatures of the particular States."[56] Here too Madison appears to argue that the people, left to themselves, will choose as members of Congress people just like themselves, in which case it is unclear how the people's views will be "refined and enlarged" by means of their representation in Congress. These difficulties in Madison's account result, I believe, from the impossibility of combining civic republican assumptions with liberal ones. The people are simultaneously self-interested and factious, susceptible to corruption in the service of self-interest, and yet able to discern and choose those with the more diffusive characters. Similarly, their representatives are going to be just like them in their attachment to the states, and yet somehow able to select national policies that overcome those local attachments.

A second form of the general problem is presented in The Federalist's account of judicial review. As we have seen, the judiciary is said to have only judgment, and not force or will. Yet, the general account of the separation of powers offered earlier in The Federalist insists that, for the system to work, each branch must have a will of its own, such that "the interest of the man must be connected with the constitutional rights of the place" through the mechanism of "ambition...counteract[ing] ambition."[57] If the judges do not have will--ambition in the relevant sense--they cannot resist the encroachments of the other branches, and yet if they are willful they lose what makes the judicial branch distinctive. I have argued elsewhere that The Federalist is able to resolve this tension only by adopting a normative theory of constitutional interpretation that is ungrounded in the assumptions about human motivation that it adopts.[58] The normative theory is ungrounded, once again, because the assumptions are internally incompatible.

The problem of dual office holding displays the same difficulties. Nothing in the structure of the Constitution guarantees that judges will differ from ordinary politicians in their motivations. Technical training, life tenure, and salary protection all contribute to a set of motives that is different from the set of motives held by ordinary politicians, but there are elements common to each set, and it is those common elements that cause the difficulties of corruption by dependence. It turns out, then, that the differences in structural arrangements for legislators, members of the executive branch, and judges are the product of the fundamental tension in the Constitution between civic republican and liberal assumptions.

The problem posed by Hayburn 's Case suggests one method for resolving that tension. The tripartite structure of the government created by the Constitution suggests that the Framers had some distinctions in mind among policy, law, and administration. Yet, the problem posed by Hayburn's Case shows that it is too facile to identify policy with the legislature, law with the judiciary, and administration with the executive branch. Rather, the allocation of those functions to different branches was worked out in the early decades of the Republic. Judges sitting as commissioners were law propounders who, in another capacity, could administer the law. Eventually administration and law were more sharply separated. Law and policy were more cleanly separated from the beginning, with the rejection of the proposal that the Justices of the Supreme Court sit as a Council of Revision whose charter would allow them to disapprove legislation on policy as well as on legal grounds. Yet, as George Haskins and Herbert Johnson argued, the final steps in the separation of law and policy were not taken until, as part of a general political strategy on the part of the Marshall Court, the Supreme Court in the early years of the nineteenth century provided a firm grounding for the distinction.[59]

In concluding, it may be helpful to examine two recent cases in which the Supreme Court examined the problem of dual office holding. In both the Court's conceptualization of the problem seems rather different from the concern for dependency and corruption that the Framers expressed. Schlesinger v. Reservists Committee to End the War was a challenge to a system in which members of Congress were allowed to hold commissions in the reserve forces of the United States.[60] The challengers argued that commissions in the reserves were "offices of the United States" which members of Congress could not, under the Incompatibility Clause, occupy. The Supreme Court refused to address the merits of the challenge, holding that the plaintiffs lacked standing. Justice Douglas dissented from the denial of standing, arguing that the essence of plaintiffs claim was that the Incompatibility Clause was designed to protect against the appearance of a conflict of interest arising because of dual office holding.[61] As we have seen, the concepts of corruption and dependency are related to, or at least can be rephrased in modern terms as, the idea of avoiding a conflict of interest. Yet, in the Framer's era, the problems of corruption and dependency were much more intimately connected to ideas of governing a democratic republic than the relatively bland phrase "conflict of interest" suggests.

As in Reservists, so too in Mistretta is the sense that there is something problematic about dual office holding expressed only in dissent. The majority opinion in Mistretta conveys no sense that there is some tension between dual office holding and the basic premises of our constitutional system, although it expresses some misgivings about the overall design of the Sentencing Commission. Justice Scalia's dissenting opinion focuses on the delegation of lawmaking authority to the Commission, and objects to the Commission in part because it violates norms of democratic responsibility and in part because even if it does not do so directly, it threatens to lead Congress down the slippery slope to real incursions on democratic responsibility.[62] Here we can see some indication of the connections among dual office holding, corruption, dependency, and the design of a democratic republic. Even so, the threat that concerns Justice Scalia comes from Congress; he does not acknowledge what the Framers knew but could not fully deal with, that judges perhaps only slightly less than ordinary politicians combine self-interest, ambition and civic virtue in ways that pose threats to the development of sound public policy. Justice Blackmun's opinion for the Court is more straightforward. Transforming Hamilton's idea that judges have special technical training in the law into a general defense of bureaucratic expertise, Justice Blackmun found the Sentencing Commission justified because of the judges' "experience and expertise."[63]

I have argued that the Constitution did not resolve, because the Framers' political theories made it impossible for them to resolve, the tension between civic republicanism and liberalism that the Constitution's treatment of dual office holding exemplifies. Those theories, though, are quite rich. Somehow Mistretta seems distressingly thin in contrast. Yet, may not that result, too, from the impossibility of carrying out the internally inconsistent program of the Constitution?

Acknowledgements: I would like to thank Vicki Jackson, Susan Low Bloch, Maeva Marcus, amid Gerry Spann for their comments on a draft of this essay.

1. Mistretta v. United States, 109 S. Ct. 647 (1989).
2. The most recent presentation is Marcus & Teir, Hayburn's Case: A Misinterpretation of Precedent," 1988 Wis. L. Rev. 527, which focuses on an aspect of the case not dealt with in this paper.
3. For reasons that will appear, I use this anachronistic term deliberately.
4. Act reasons that will appear, I use this anachronistic term deliberately.
5. Act of Mar. 23, 1792, 1 Stat. 243 (1792).
6. Rakove, "the First Phases of American Federalism," in Comparative Constitutional Federalism (M. Tushnet Ed. 1990).
7. The Federalist Papers 45, p. 291 (C. Rossiter ed. 1961).
8. For an additional reason for conferring the duties under the Invalid Pensions Act on the circuit courts, see text accompanying note 47 infra.
9. For a recent discussion, see Holt, "'The Federal Courts Have Enemies in All Who Fear Their Influence on State Objects': The Failure to Abolish Supreme Court Circuit-Riding in the Judiciary Acts of 1792 and 1793," 36 Buff. L. Rev. 301 (1987).
10. 1 American State Papers (Misc.), 50 (New York), 51 (Pennsylvania).
11. The delay may also have resulted from the strength of Justice Iredell's misgivings about his position, and from his desire to wait until he was on his home ground to express them.
12. Id. at 52-53. In October 1792, Justice Thomas Johnson joined his colleagues by refusing to process applications by veterans, saying only that the court could not "constitutionally take Cognizance of and determine [the] said Petitions." See Bloch, "The Office of Attorney General on Its Bicentennial; From Edmund Randolph and Hayburn's Case to Edwin Meese and Special Prosecutors--Lessons From the Past," Duke L.J. (forthcoming).
13. Russell Wheeler suggests that the determination of the pension level was what made the task nonjudicial, because the standard that the rate must be comparable to the applicant's degree of disability was so vague that it would inevitably elicit nonjudicial policy judgments. Wheeler, "Extrajudicial Activities of the Early Supreme Court," 1973 Sup. Ct. Rev. 123, 137. In modern terms this would be an objection that the delegation of authority to the judges failed to provide them sufficient guidance: again in modern terms, that objection would almost certainly fall, as the Court's analysis of the nondelegation issue in Mistretta demonstrates.
14. See Tutun v. United States, 270 U.S. 568 (1926); Hart and Weschler's The Federal Courts and the Federal System 99-101 (3d ed. 1988); Wheeler, supra note 12, at 134-35.
15. Gordon v. United States, 117 U.S. 697, 702-03 (1866) (appendix).
16. District of Columbia v. Eslin, 183 U.S. 62 (1901).
17. See Hart & Weschler, supra note 14, at 104-05.
18. Glidden Co. v. Adanok, 370 U.S. 530, 570 (1962) (opinion of Harlan, J., announcing the judgment of the Court).
19. Act of Feb. 28, 1793, 1 Stat. 324 (1793).
20. What the Supreme Court actually did do is insightfully analyzed in Marcus & Tier, supra note 2.
21. 109 S. Ct. at 670.
22. The judges in Pennsylvania did not address this question at all.
23. 1 American State Papers (Misc.) 50 Note, incidentally, that the praise of Congress's benevolence undercuts the argument that the possibility that Congress might not appropriate money for the pensions makes the judges' decisions subject to legislative revision.
24. Id. at 53.
25. Pennsylvania Herald and York General Advertiser, York, Pa. Oct. 10, 1792. (I am grateful to Maeva Marcus and Susan Low Bloch for making this and other material used in this essay available to me.)
26. Quoted in Marcus and Teir, supra note 2, at 39-40.
27. Wythe Holt has pointed out to me that at the time the Act was passed, the Justices were fixed in their circuits not by statute but by their own internal rule. Before the Act was passed on March 23, Justice Iredell had apparently been assured by his brother-in-law that the Judiciary Act would be amended to require rotation of the Justices among the circuits, which it was on April 13. See Holt, "'The Federal Courts Have Enemies in All Who Fear Their Influence on State Objects': The Failure to Abolish Supreme Court Circuit-Riding in the Judiciary Acts of 1792 and 1793," 36 Buff. L. Rev. 301, 329-30 (1987).
28. For a recounting of the decision, see United States v. Ferriera, 54 U.S. (13 How.) 40, 52-53 (appendix).
29. See Mistretta, at 668.
30. The additional account, of course, was the assumption common to the Framers that Congress was the branch most likely to end up controlling the government. See, e.g., The Federalist No 51, p. 322 (C. Rossiter ed. 1961) ("In republican government, the legislative authority necessarily predominates").
31. See, e.g., 2 The Framers' Constitution 346 (P. Kurland & R. Lerner eds. 1987) (Butler) 347 (Mason), 350 (Gerry).
32. See, Id., at 352-53 (Wilson at Pennsylvania ratifying convention).
33. In fact, this possibility materialized in 1801. Three Federalist Senators and one Federalist Representative from the outgoing Congress that had created the so-called "midnight judges" in February 1801, each of whom had been defeated for reelection, were appointed by President Adams to the federal district court to replace district judges elevated to the new circuit courts. Although three of these nominees never sat (two because the judges who they were named to replace decided not to accept the elevation, and one because his commission as (improperly filled out), one Elijah Paine was a district judge for 41 years. (I thank Wythe Holt for this information.) S. Turner, "The Midnight Judges," 109 U. Pa. L. Rev. 494 (1961).
34. Id. at 347.
35. Id. at 347-48.
36. It may be worth pointing out that a provision like the emoluments clause standing alone might have been sufficient to reduce the risk of corruption to an acceptable level, because the length of service of federal judges, with their life appointments, might reduce the number of positions for which they are eligible quite substantially.
37. It may be worth noting that this prediction, if it was part of the calculation, may not be borne out by recent experience. For a list of people who resign federal judgeships for executive or legislative branch positions, see Tushnet, Kovner, & Schneider, "Judicial Review and Congressional Tenure: An Observation," 66 Tex L. Rev. 967, 980 (1988). (The list would now include Kenneth Starr.)
38. Mistretta at 674-75. The Court did note, at 675 n. 34, that a judge fired by a President might suffer "some embarrassment or even damage to reputation," but considered that a judge who held an executive branch position would have assumed the risk of that harm. (It should be noted that Congress provided that the Sentencing Commission would be "located in the Judicial Branch." As far as I can tell, however, this statement has no analytical consequences either for the Court's discussion in Mistretta or for general constitutional concerns.)
39. The Federalist 78, p. 465 (C. Rossiter ed. 1961).
40. Id. at 471.
41. 4 The Framers' Constitution, supra note 31, at 1333, 137.
42. See Id. at 136-37 (Franklin: money may "become plentier"), 140 (Thomas McKean at Pennsylvania ratifying convention); The Federalist 79, p. 473 (C. Rossiter ed. 1961) (mentioning "fluctuations in the value of money," and the difference between salary provision for judges and the President). It should be noted that the "fluctuations" of concern all must be in the direction of inflation; in cases of substantial deflation, the ban on salary reductions would convert an acceptable level of pay into an extravagant one.
43. The Federalist 73, pp. 441-42 (C. Rossiter ed. 1961).
44. 4 The Framer's Constitution, supra note 30, at 136-37.
45. Madison proposed fixing salaries with reference to wheat or "some other thing of permanent value," Id. at wheat or "some other thing of permanent value," Id. at 137, which suggests the difficulty.
46. See "Symposium: The Republican Civic Tradition," 97 Yale L. J. 1493-723 (1988).
47. See M. Tushnet, Red, White and Blue: A Critical Analysis of Constitutional Law 4-16 (1988), criticized on this point among others, in Gerhardt, "Critical Legal Studies and Constitutional Law," 67 Tex. L. Rev. 393, 404-10 (1988).
48. For the basic insight, see R. Lerner, "The Supreme Court as Republican Schoolmaster," in The Thinking Revolutionary (1987; this essay was originally published in 1967). Lerner focuses on the judges' explicit instruction in republican principles, as conveyed through their charges to grand juries, but his insight seems valid across a broader range of the judges' activities.
49. But see note 13 supra.
50. According to Wheeler, note 13 supra, at 138, the elimination of the judges from the administration of the pension act in 1793 was not primarily the result of the judges' constitutional objections to their role, but rather resulted from the generosity of the judges, sitting as commissioners, in awarding persons. Given that many of the judges were in fact administering the pension statute, Wheeler's conclusion seems sound; the constitutional objection had been overcome by statutory interpretation, and there was therefore no longer any need to remove the judges from the administration of the act in order to avoid a constitutional problem.
51. Lerner, supra note 48, at 124. See also Id., at 130 ("at still another level--transcending its other functions, and implied in the technical knowledge needed by this branch of government alone--the judiciary stands as special guardian of the principles of the Constitution"), 133 ("The judges--and the judges alone, of all government officials--needed to have special training and character in order to do their jobs at all").
52. This is particularly evident because Lerner goes to some length to establish that the courts will not be "weak" or "unnoticed." Id. at 127.
53. The Federalist 10, p. 82 (C. Rossiter ed. 1961).
54. Ibid.
55. Id. at 83.
56. The Federalist 46, p. 296 (C. Rossiter ed. 1961).
57. The Federalist 51, p. 322 (C. Rossiter ed. 1961).
58. Tushnet, "Constitutional Interpretation and Judicial Selection: A View from The Federalist Papers," 61 S. Cal. L. Rev. 1669 (1988).
59. G. Haskins & H. Johnson, The Oliver Wendell Holmes Devise History of the Supreme Court of the United States, Volume II: Foundations of Power: John Marshall 1801-1815 (1981).
60. 418 U.S. 208 (1974).
61. Id. at 232-234 (Douglas, J., dissenting).
62. Mistretta, at 680 (Scalia, J., dissenting).
63. Id. at 673.