Whether a Former President May Be Indicted and Tried for the Same Offenses for Which He Was Impeached by the House and Acquitted by the Senate

     The Constitution permits a former President to be indicted and tried for the same offenses for which he was impeached by the House of Representatives and acquitted by the Senate.

MEMORANDUM FOR THE ATTORNEY GENERAL

     We have been asked to consider whether a former President may be indicted and tried for the same offenses for which he was impeached by the House and acquitted by the Senate.¹ In 1973, in a district court filing addressing a related question in the criminal tax evasion investigation of Vice President Agnew, the Department took the position that acquittal by the Senate creates no bar to criminal prosecution. A 1973 Office of Legal Counsel ("OLC") memorandum discussing the same question adopted the same position. As far as we are aware, no court has ever ruled on this precise issue. During the impeachment of Judge Alcee Hastings in the late 1980s, though, a district court and both the House and Senate passed on the related question whether an acquittal in a criminal prosecution should bar an impeachment trial for the same offenses. Each of those bodies concluded that the Constitution permits an official to be tried by the Senate for offenses of which he has been acquitted in the courts. Although we recognize that there are reasonable arguments for the opposing view, on balance, and largely for some of the same structural reasons identified in the United States's filing in the Agnew case and the 1973 OLC memorandum, we think the better view is that a former President may be prosecuted for crimes of which he was acquitted by the Senate. Our conclusion concerning the constitutional permissibility of indictment and trial following a Senate acquittal is of course distinct from the question whether an indictment should be brought in any particular case.

     This memorandum has three parts. First, we review the reasoning of the United States's filing in the Agnew case and of the 1973 OLC memorandum. Second, we consider in greater depth the arguments for and against the constitutional permissibility of criminal prosecution of officials for the same offenses of which they have been acquitted by the Senate. Third, we summarize and consider the significance of the Hastings impeachment process and of the Senate trials of two other federal judges who were impeached and convicted during the 1980s following criminal prosecution.

     I. The 1973 Justice Department Documents

     A. The United States's Brief in the Grand Jury Investigation of Vice President Agnew

     In 1972, the United States Attorney for the District of Maryland empaneled a grand jury to investigate criminal charges against Vice President Spiro Agnew. The Vice President filed a motion with the district court supervising the grand jury seeking to enjoin the grand jury from investigating or indicting him, claiming that his office gave him immunity from indictment and criminal trial. The United States filed a brief, signed by Solicitor General Robert Bork, opposing the Vice President's motion. The brief's central contention was that "all civil officers of the United States other than the President are amenable to the federal criminal process either before or after the conclusion of impeachment proceedings." Memorandum for the United States Concerning the Vice President's Claim of Constitutional Immunity, In Re Proceedings of the Grand Jury Impaneled December 5, 1972: Application of Spiro T. Agnew, Vice President of the United States, Civ. No. 73-965 (D. Md. filed Oct. 5, 1973) at 3 ("Agnew Brief").

     One of the arguments the brief addresses is the contention that the Impeachment Judgment Clause, Article I, Section 3, Clause 7 of the Constitution dictates that impeachment must precede indictment. That clause provides:

     Judgment in Cases of Impeachment shall not extend further than to removal from Office, and disqualification to hold and enjoy any Office of honor, Trust, or Profit under the United States: but the Party convicted shall nevertheless be liable and subject to Indictment, Trial, Judgment, and Punishment, according to Law.

In response to the argument that impeachment must precede prosecution, the brief first states, "As it applies to civil officers other than the President, the principal operative effect of Article I, Section 3, Clause 7, is solely the preclusion of pleas of double jeopardy in criminal prosecutions following convictions upon impeachments." Agnew Brief at 7. It goes on, however, to contend that the clause allows criminal prosecution upon acquittal by the Senate as well. See id. at 8.

     It then provides, though in very summary form, five arguments for that conclusion. First, impeachment and trial by the Senate, on the one hand, and prosecution in the courts, on the other, "serve different ends." Ibid. Although the brief does not actually spell out those different ends, they seem to be protection of our institutions of government from corrupt or incompetent officials, on the one hand, and punishment of those individuals, on the other. The only illustration the brief offers is that "a civil officer found not guilty by reason of insanity in a criminal trial could certainly be impeached nonetheless." Id. at 9. In a related vein, the brief argues that trial on impeachment is a civil proceeding akin to deportation rather than a criminal proceeding. Id. at 10 n.**. Second, the brief points out that impeachment trials "may sometimes be influenced by political passions and interests that would be rigorously excluded from a criminal trial." Id. at 9. Third, an acquittal by the Senate will often rest on a determination by at least a third of the Senate that the conduct alleged, though proven, does not amount to a high crime or misdemeanor. Such a judgment in no way reflects a determination that the conduct is not criminal in the ordinary sense. Id. Fourth, if the scope of the Impeachment Judgment Clause were restricted to convicted parties, "the failure of the House to vote an impeachment, or the failure of the impeachment in the Senate, would confer upon the civil officer accused complete and - were the statute of limitations permitted to run - permanent immunity from criminal prosecution however plain his guilt." Id. at 9-10.² Fifth, such a view would give Congress an indirect power of pardon - via impeachment and acquittal - even though the Constitution vests the President alone with the power to pardon. Id. at 10.

     B. The 1973 OLC Memorandum

     In 1973, this office prepared a memorandum on the amenability of the President, the Vice President, and other civil officers to federal criminal prosecution while in office. The memorandum's central conclusion was that all federal officers and the Vice President, but not the President, are amenable to federal prosecution while in office. The memorandum did not discuss at any length the question whether a former President who has been acquitted by the Senate may be indicted and criminally tried. It did spend considerable time, however, refuting the notion that the Impeachment Judgment Clause required officers to be impeached by the House and tried by the Senate before they may be criminally prosecuted. Instead, the memorandum stated, "[t]he purpose of this clause . . . is to permit criminal prosecution in spite of the prior adjudication by the Senate, i.e., to forestall a double jeopardy argument." Memorandum From Robert G. Dixon, Jr., Assistant Attorney General, Office of Legal Counsel, Re: Amenability of the President, Vice President and other Civil Officers to Federal Criminal Prosecution While in Office (Sept. 24, 1973) at 3 ("1973 OLC Memo"). In support of that claim, the memorandum cited a passage from the argument made by Luther Martin in his role as defense counsel in the impeachment trial of Justice Chase in 1805³ and quoted a passage from Justice Joseph Story's 1833 Commentaries on the Constitution.⁴ Story, the memorandum suggested, took the position that neither conviction nor acquittal by the Senate would bar a criminal prosecution. Id. at 2 n.2. The reasoning supporting our embrace of the position we attributed to Story was contained in a single sentence in a footnote: "The conclusion that acquittal by the Senate does not bar criminal prosecution follows from the consideration that such an acquittal may be based . . . on jurisdictional grounds, e.g., that the defendant is not an officer of the United States in the constitutional sense, or on discretionary grounds, e.g., that the defendant no longer is an officer of the United States and unlikely to be reappointed or reelected, or on grounds which are partly jurisdictional and partly substantive, e.g., that the offense was not of an impeachable nature." Id. The memorandum thus rested its conclusion on a somewhat elaborated version of the third argument made in the United States's brief in the Agnew case.

     II. The Arguments Considered in Greater Depth

     There appear to be two possible bases in the Constitution for the claim that a former President who was acquitted by the Senate while he was in office may not be criminally prosecuted for the same offenses: the Impeachment Judgment Clause and the Double Jeopardy Clause. We will consider each in turn.

     A. The Impeachment Judgment Clause

     1. The Argument That Senate Acquittal Bars Subsequent Prosecution

     The Constitution itself expressly authorizes indictment and trial of officials who have been impeached and convicted. As noted above, Article I, Section 3, Clause 7 of the Constitution states:

     Judgment in Cases of Impeachment shall not extend further than to removal from Office, and disqualification to hold and enjoy any Office of honor, Trust, or Profit under the United States: but the Party convicted shall nevertheless be liable and subject to Indictment, Trial, Judgment, and Punishment, according to Law.

     The clause is ambiguous when it comes to officials who have been impeached and not convicted. Some commentators have argued that the reference to "the Party convicted" implies that the exception to the double jeopardy principle created by the clause does not extend to parties who are impeached but not convicted.⁵ Judge Alcee Hastings made the same argument in challenging the Senate's jurisdiction to try him on impeachment after he had been tried and acquitted in a federal criminal prosecution.⁶

     This argument rests on the well-known canon of statutory construction, expressio unius est exclusio alterius, "the expression of one is the exclusion of others." United States v. Wells Fargo Bank, 485 U.S. 351, 357 (1988). The Impeachment Judgment Clause says "the party convicted," not "the party, whether convicted or acquitted." Its failure to mention parties acquitted by the Senate implies that they, unlike convicted parties, are not subject to regular criminal prosecution.

     This argument has some force. The Court has regularly relied on the expressio unius canon. See, e.g., Custis v. United States, 511 U.S. 485, 491-492 (1994); Leatherman v. Tarrant County Narcotics Intelligence & Coordination Unit, 507 U.S. 163, 168 (1993); National R.R. Passenger Corp. v. National Ass'n of R.R. Passengers, 414 U.S. 453, 457 (1974). Although the canon has most often been applied to statutes, rules, and contracts, the Court has used it as well in analyzing constitutional provisions. See, e.g., U.S. Terms Limit, Inc. v. Thornton, 514 U.S. 779, 793 n.9 (1995) (qualifications for Representatives specified in the Qualifications Clause are exclusive). Indeed, one might argue that the canon has particular strength when applied to constitutional provisions because, as the Court has noted, those provisions are likely to be drawn with particular care. See, e.g., Township of Pine Grove v. Talcott, 86 U.S. (19 Wall.) 666, 674-75 (1873) ("[t]he case as to the [Michigan] constitution is a proper one for the application of the maxim, 'Expressio unius . . .'. The instrument is drawn with ability, care, and fulness of details"). In addition, if the Impeachment Judgment Clause is understood as creating an exception to the general background rule of a prohibition on successive prosecutions, the expressio unius canon is particularly apt since it has often been wielded to support the conclusion that when a statute identifies specific exceptions to a general rule it by implication prohibits other exceptions. See, e.g., Leatherman, 507 U.S. at 168; TVA v. Hill, 437 U.S. 153, 188 (1978); City of Walla Walla v. Walla Walla Water Co., 172 U.S. 1, 22 (1898); Arthur v. Cumming, 91 U.S. 362, 363 (1875); Sturges v. Collector, 79 U.S. (12 Wall.) 19, 27 (1870).

     The expressio unius argument gains plausibility from a comparison of the federal Impeachment Judgment Clause with the equivalent clauses in state constitutions. Of the forty-five state constitutions that authorize impeachment and limit the punishment upon conviction, all forty-five provide for further prosecution in the courts. In doing so, however, only fifteen follow the federal wording of "the party convicted"⁷; thirty, by contrast, expressly provide that the party impeached is liable to criminal proceedings regardless of the outcome of the legislative trial.⁸ Moreover, express provisions concerning those acquitted in impeachment trials are not a recent innovation. The first state constitution to include a reference making clear that an impeachment acquittal created no bar to criminal prosecution was the Pennsylvania charter of 1790.⁹ That state's constitution, like many others, says that "the party, whether convicted or acquitted" is liable to prosecution in the courts.¹⁰ Perhaps most telling is the New York constitution, the original 1777 version of which contained language strikingly similar to that later included in the U.S. constitution and which may well have been the source of the wording for the federal clause.¹¹ In the mid-nineteenth century, the New York charter was amended to refer to "the party impeached" rather than "the party convicted" precisely because of a concern that the latter phrase might be understood to give immunity from criminal prosecution to those who had been impeached and acquitted.¹²

     Finally, the expressio unius argument rests on more than the wording of the Impeachment Judgment Clause. The framers might well have had a principled basis for treating acquittals and convictions by the Senate distinctly. The American rule of double jeopardy derives from the common law pleas of auterfois acquit, formerly acquitted, and auterfois convict, formerly convicted.¹³ As Blackstone explained, both pleas are grounded in the "universal maxim of the common law of England, that no man is to be brought into jeopardy of his life, more than once, for the same offence,"¹⁴ and the Double Jeopardy Clause, in giving that maxim constitutional stature, embraces the protections both against re-prosecution following acquittal and against re-prosecution following conviction.¹⁵ But, as the Supreme Court has explained, the rationales for the two components of the double jeopardy rule are somewhat different. "The primary purpose of foreclosing a second prosecution after conviction . . . is to prevent a defendant from being subjected to multiple punishments for the same offense." Justices of Boston Municipal Court v. Lydon, 466 U.S. 294, 307 (1984). By contrast, the "primary goal of barring reprosecution after acquittal is to prevent the State from mounting successive prosecutions and thereby wearing down the defendant." Id. "The underlying idea," the Court has repeatedly affirmed,

one that is deeply ingrained in at least the Anglo-American system of jurisprudence, is that the State with all its resources and power should not be allowed to make repeated attempts to convict an individual for an alleged offense, thereby subjecting him to embarrassment, expense and ordeal and compelling him to live in a continuing state of anxiety and insecurity, as well as enhancing the possibility that even though innocent he may be found guilty.

Green v. United States, 355 U.S. 184, 187-188 (1957).

     The central innovation of the Impeachment Judgment Clause, as explained more fully below, was the restriction on the types of sanctions the Senate could impose when it convicted someone upon impeachment. Breaking with English practice, in which the House of Lords could impose regular criminal punishments up to death, the framers provided that the Senate could do no more than remove an offender from office and disqualify him from future federal officeholding. The framers might reasonably have concluded that their innovative restriction of impeachment sanctions justified a relaxation of the normal ban on multiple punishments - and thus a relaxation of the former jeopardy principle in the case of Senate convictions - in order to ensure that federal officials did not escape the punishments suffered by offenders against the criminal law who held no federal office. No similar relaxation, they might have reasoned, was warranted in the case of successive trials following acquittals. The central rationales of the ban on successive trials - the unfairness of the government's repeatedly subjecting an individual to the ordeal and expense of prosecution and the unfairness of giving the government a chance to hone its case and thus to secure the conviction of an innocent party - arguably still applied. Thus the use of the phrase "the party convicted" in a restrictive sense might well have had a perfectly reasonable basis in the underlying concerns of the double jeopardy rule.¹⁶

     Moreover, if the Impeachment Judgment Clause is seen not as addressing double jeopardy concerns per se, but rather as providing protections for officers accused of wrongdoing, its silence about parties acquitted by the Senate makes sense and suggests the framers thought acquittal by the Senate would bar criminal prosecution. The Impeachment Judgment Clause provides protection most directly by depriving the Senate of the ability to impose regular criminal punishments, but it also ensures that even those convicted by the Senate will get a regular trial, with a jury and other guarantees, rather than having additional punishments imposed in some more summary proceeding. As Hamilton put it in Federalist 65, the guarantee of trial in the courts following trial in the Senate provides "the double security, intended them by a double trial."¹⁷ Once the defendant-protecting function of the Impeachment Judgment Clause is recognized, its silence about acquitted parties is most reasonably understood as reflecting the assumption that such parties, like those acquitted in the courts, would not be subject to further prosecution.

     Even apart from the special functions of the Impeachment Judgment Clause, the framers might have considered protection of the finality of acquittals more fundamental than protection of the finality of convictions.¹⁸ The one state constitution in the revolutionary period that contained a double jeopardy clause only barred re-trials when there had been an acquittal,¹⁹ as did one of the two state proposals for a federal double jeopardy clause.²⁰ In the case law that has grown up under the federal Double Jeopardy Clause, the Supreme Court has recognized that "[a]n acquittal is accorded special weight." United States v. DiFrancesco, 449 U.S. 117, 129 (1980); see Tibbs v. Florida, 457 U.S. 31, 41 (1982) ("the Double Jeopardy Clause attaches special weight to judgments of acquittal"). The special place of acquittals helps explain several asymmetries in double jeopardy law, notably that the Constitution places no restrictions on defendants' ability to appeal convictions but prevents government appeals of acquittals that would lead to re-trial. See United States v. Wilson, 420 U.S. 332, 345, 352 (1975).

     2. The Impeachment Judgment Clause Permits Prosecution Following Acquittal: Textual and Historical Considerations

     Despite its initial plausibility, we find this interpretation of the Impeachment Judgment Clause ultimately unconvincing for several reasons.

     a. Expressio Unius Is Only an Aid to Construction

     The expressio unius canon is only an aid to interpretation, an aid that cannot trump larger considerations of context and purpose. Although the Court has regularly endorsed expressio unius arguments, it has also regularly rejected them. See, e.g., Freightliner Corp. v. Myrick, 514 U.S. 280, 288-289 (1995) (statutory preemption); Pauley v. Bethenergy Mines, Inc., 501 U.S. 680, 703 (1991) (methods of rebuttal in regulations; citing Sunstein, 90 Columbia L. Rev., at 2190, n.182 for the proposition that "the principle of expressio unius est exclusio alterius 'is a questionable one in light of the dubious reliability of inferring specific intent from silence'"); Sullivan v. Hudson, 490 U.S. 877, 892 (1989); Herman & MacLean v. Huddleston, 459 U.S. 375, 387 n.23 (1983); Bingler v. Johnson, 394 U.S. 741, 749-750 (1969). Again and again, the Court has cautioned that the maxim "is an aid to construction, not a rule of law," Neuberger v. Commissioner, 311 U.S. 83, 88 (1940), and that "[h]owever well these rules may serve at times to aid in deciphering legislative intent, they long have been subordinated to the doctrine that courts construe the details of an act in conformity with its dominating general purpose [and] will read text in the light of context," SEC v. C.M. Joiner Leasing Corp., 320 U.S. 344, 350-351 (1943); see Herman & MacLean, 459 U.S. at 387 n.23.²¹

     b. Origins of the Impeachment Judgment Clause and Early Understandings

     We are unaware of any evidence suggesting that the framers and ratifiers of the Constitution chose the phrase "the party convicted" with a negative implication in mind. In its most recent decision approving an expressio unius argument concerning the meaning of a constitutional provision, the Court noted that it found the argument compelling in significant part because such direct evidence of the framers' intent was available. See U.S. Terms Limit, Inc., 514 U.S. at 793 n.9. Here, by contrast, the record offers no similar signs of awareness that "the party convicted" would be read to exclude acquitted parties from the effect of the Impeachment Judgment Clause's final sentence. Indeed, while a number of participants in the ratification debates and several early commentators simply repeated the words of the Impeachment Judgment Clause in describing it, at least two influential participants in the debate, one Member of Congress in the early republic, and at least one of our most distinguished early constitutional commentators understood the clause to allow prosecution of parties who had been acquitted by the Senate as well as of those who had been convicted.

     In 1787, impeachment already had a long history in Britain, but in Britain conviction on impeachment might result in a wide array of criminal penalties, including fines, imprisonment, and even execution.²² Restriction of the punishments attendant on conviction by the legislature to removal and disqualification was an American innovation developed over the course of the 17^th and 18^th centuries.²³ Five of the state constitutions from the revolutionary period expressly addressed the types of punishments that conviction on impeachment could bring,²⁴ and three of the five contained language that the drafters of the federal clause may well have borrowed. New York's charter of 1777 created a court for the trial of impeachments consisting of the members of the senate, the chancellor, and the judges of the supreme court, and provided that "no judgment of the said court . . . shall . . . extend farther than to removal from office, and disqualification to hold or enjoy any place of honor, trust, or profit under this State. But the party so convicted shall be, nevertheless, liable and subject to indictment, trial, judgment, and punishment, according to the laws."²⁵ The Massachusetts constitution of 1780 and the New Hampshire constitution of 1784 (largely patterned on its Massachusetts predecessor) made their senates the court for the trial of impeachments and then stated that "[t]heir judgment, however, shall not extend further than to removal from office, and disqualification to hold or enjoy any place of honor, trust, or profit, under this Commonwealth: But the party so convicted, shall be, nevertheless, liable to indictment, trial, judgment, and punishment, according to the laws of the land."²⁶

     At the federal Constitutional Convention, most of the debate over impeachment concerned three subjects: the wisdom of allowing impeachment of the President, the tribunal in which impeachments should be tried, and the nature of the offenses that should impeachable.²⁷ The limitation on the types of punishments available on conviction and the provision for criminal prosecution despite conviction on impeachment were proposed by the Committee of Detail, to which the Convention on July 23 gave the assignment of crafting a draft constitution based on the convention's deliberations so far. That committee made its report on August 6.²⁸ Their report made the Supreme Court the tribunal for trying impeachments, and the Impeachment Judgment Clause appeared in the final section of the their proposed judiciary article.²⁹ The convention approved it, apparently without division.³⁰ The Impeachment Judgment Clause remained unchanged throughout the debate over the proper tribunal for trying impeachments and the eventual giving of that responsibility to the Senate.³¹ When the Committee of Style and Arrangement near the end of the convention reported the clause in its present terms,³² it occasioned no debate except a proposal, rejected by the convention, to add a provision that a party impeached be suspended from office until tried and acquitted.³³

     To sum up, then, the Impeachment Judgment Clause was written as part of a draft constitution that made the Supreme Court, not the Senate, the tribunal for trying impeachments. The records of the Convention do not show any discussion of whether the change in the impeachment court had any effect on the meaning of the clause. More broadly, the records do not reflect any substantive discussion of the clause's meaning.

     As in the Convention, so during the ratification debates most of the discussion of impeachment concerned the proper tribunal for trying impeachments and the range of impeachable offenses. Critics of the Constitution questioned the Senate's role as the court for impeachments, and several state ratifying conventions proposed alternative bodies, at least for the trial of senators.³⁴ References to the Impeachment Judgment Clause were rare.

     Some commentators, in describing the Clause, simply repeated its own terms or mentioned only the particular circumstance it explicitly sanctioned: liability to prosecution following conviction by the Senate. Hamilton devoted Federalist No. 65, for example, to a defense of the selection of the Senate as the tribunal for trying impeachments. One of his claims for the Senate's superiority over the Supreme Court was that, if impeachments were tried before the Supreme Court, the same body would improperly have final review over each of the two trials to which an impeached official might be subjected. For "[t]he punishment, which may be the consequence of conviction upon impeachment," he noted, "is not to terminate the chastisement of the offender. After having been sentenced to a perpetual ostracism from the esteem and confidence, and honors and emoluments of his country; he will still be liable to prosecution and punishment in the ordinary course of law."³⁵ Others invoked the Clause in order to defend the Senate's judicial role by stressing the limited nature of its judgments. Tench Coxe, a leading advocate of the Constitution in Pennsylvania, in an essay assessing the roles assigned to the newly designed Congress, parried the contention that the Senate had unwisely been given judicial functions, by pointing out that the Senate "can only, by conviction on impeachment, remove and incapacitate a dangerous officer, but the punishment of him as a criminal remains within the province of the courts of law to be conducted under all the ordinary forms and precautions, which exceedingly diminishes the importance of their judicial powers."³⁶ Still other commentators held up the Impeachment Judgment Clause as evidence that the newly created federal executive would not be able to abuse his power without facing severe punishment. A Virginia supporter of the Constitution argued that should the President "at any time be impelled by ambition or blinded by passion, and boldly attempt to pass the bounds prescribed to his power, he is liable to be impeached and removed from office; and afterwards he is subject to indictment, trial, judgment, and punishment according to law."³⁷

     These remarks on the Impeachment Judgment Clause reflect the two concerns motivating it. Because impeachment was designed to serve above all as a legislative check on executive power,³⁸ the Impeachment Judgment Clause was intended to make sure both that the special legislative court for the largely political offenses justifying impeachment would be able to impose only political, not ordinary criminal, punishments and that offenders who also violated regular criminal laws would not stand above the law because they had been officeholders when they committed their misdeeds. Presumably, these commentators did not address the consequences of acquittal by the Senate because that was not a subject the Impeachment Judgment Clause addressed. Indeed, if the Impeachment Judgment Clause were intended to imply that acquittal by the Senate would block criminal prosecution for the same offenses, one would expect that at least one participant in the process of framing and ratifying the Constitution would have pointed out this negative implication. We are aware of none.

     Two well-informed participants did, however, understand the Impeachment Judgment Clause to imply that an acquittal, like a conviction, would not bar criminal prosecution for the same offences. James Wilson, a leading figure at the Constitutional Convention (and member of the Committee of Detail, which drafted the Impeachment Judgment Clause), and at the Pennsylvania ratifying convention, and later an Associate Justice of the Supreme Court, revealed such an understanding in remarks during the Pennsylvania ratifying convention. Assuming, as many did during the ratification debates, that Senators as well as executive and judicial officers would be liable to impeachment, Wilson responded to the charge that the Senate could not serve as an effective impeachment court for its own members. Noting that one third of the Senate faced re-election every two years, Wilson suggested that voters would throw out those who behaved improperly and that enough new Senators would regularly be added so that personal connections or collective involvement in the impeachable acts would not prevent fair trials. Moreover, he argued, "Though they may not be convicted on impeachment before the Senate, they may be tried by their country; and if their criminality is established, the law will punish."³⁹ Edmund Pendleton, the President of the Virginia Supreme Court and of the Virginia Ratifying Convention, apparently interpreted the Impeachment Judgment Clause in this way as well. Shortly after the completion of the Constitutional Convention, Madison sent Pendleton a copy of the Constitution for his consideration. In his generally favorable response, Pendleton confessed his leeriness of impeachments because of their susceptibility to partisan misuse, but noted that the impeachment power "is in the hands of the House of Representatives, who will not use it in the case Supposed, or if they do, and meet the obstruction, may yet resort to the courts of Justice, as an Acquital would not bar that remedy."⁴⁰

     At least some participants in the first federal impeachment trial, that of Senator William Blount of Tennessee in 1798, shared Wilson's and Pendleton's understanding of the Impeachment Judgment Clause. In a debate over whether an impeachment trial was a criminal proceeding and thus whether the House should instruct the managers to request that the Senate compel the defendant's appearance, Samuel Dana, a Representative from Connecticut, observed that "[w]ere the offence to be considered as a crime, merely, the judgment of the court should involve the whole punishment; whereas, it has no connexion with punishment or crime, as, whether a person tried under an impeachment be found guilty or acquitted, he is still liable to a prosecution at common law."⁴¹

     Two of our earliest and most eminent commentators on the Constitution also addressed the implications of the Impeachment Judgment Clause for Senate acquittals. St. George Tucker, a distinguished jurist and editor of an edition of Blackstone's Commentaries that gained widespread use in the early nineteenth-century United States, included the first extended commentary on the new federal constitution since the ratification debates as an appendix to his edition of Blackstone. In a section questioning the wisdom of making the Senate the tribunal for trying impeachments, Tucker acknowledged that "a person convicted upon an impeachment, shall nevertheless be liable and subject to indictment, trial, judgment, and punishment, according to law." In a footnote he then added: "And as a conviction upon an impeachment, is no bar to a prosecution upon an indictment, so perhaps, an acquittal may not be a bar."⁴² If Tucker thought the implication of the Impeachment Judgment Clause that Senate acquittals would be no bar to criminal prosecution was only possible, Justice Story seemed to take the point for granted in his 1833 Commentaries on the Constitution. Story observed that if the Senate had been given the authority to mete out regular criminal punishments, "then, in case of an acquittal, there cannot be another trial of the party for the same offence in the common tribunals of justice" because the common law double jeopardy principle would forbid it.⁴³ Without the Impeachment Judgment Clause, Story contended, "it might be a matter of extreme doubt" whether, in light of the double jeopardy rule, "a second trial for the same offence could be had, either after an acquittal, or a conviction in the court of impeachments."⁴⁴ In Story's view, the Impeachment Judgment Clause removed any doubt about a double jeopardy bar in the case of Senate acquittals no less than in the case of Senate convictions.

     c. Reading the Impeachment Judgment Clause as a Whole

     That two participants in the ratification process and a number of other early readers of the Constitution did not understand "the party convicted" as containing a negative implication concerning parties acquitted by the Senate fits our understanding of the role of the Impeachment Judgment Clause as a whole. The clause as a whole serves to make clear how the methods for punishing misconduct by high officials in the new American national government would differ from those in the English system. Indeed, the clause might well be called the Impeachment Conviction or Impeachment Punishments Clause.⁴⁵ Again, in England, the House of Lords could not only remove officials from office and disqualify them from holding office, but also impose a full range of criminal punishments on impeachment defendants, including, for example, banishment, forfeiture of estate, imprisonment, and death. In the new American national government, the first sentence of the Impeachment Judgment Clause establishes that the Senate would be limited to the first two sanctions: removal and disqualification. That restriction would raise the question whether the other punishments the founding generation was accustomed to seeing imposed by the House of Lords could be imposed at all under the new American government. If the Senate could not impose such sanctions, perhaps no body could. In support of that view, the phrase "Judgment in cases of impeachment" might have been read to mean the entire group of sanctions imposed by any tribunal considering a case arising from facts that led to an impeachment.⁴⁶ The Impeachment Judgment Clause's second part makes clear that the restriction on sanctions in the first part was not a prohibition on further punishments; rather, those punishments would still be available but simply not to the legislature. The courts would be the bodies entrusted with imposing those punishments even on high officials. The clause's final sentence ensured that high officials would be fully punished for their misdeeds. Thus, because the clause addressed a problem concerning the nature of punishments and the institutions entrusted with imposing them -- a problem created by the American break from longstanding English practice -- it simply had no need to address the effect of acquittal by the Senate.

     d. Impeachment and Jeopardy: Early Understandings

     We recognize that the final sentence of the Impeachment Judgment Clause might be read instead as a partial response to a perceived double jeopardy problem raised by that very American innovation. Indeed, the expressio unius argument sketched earlier in this memorandum rests on the assumption that the founders understood an impeachment trial as an instance of jeopardy within the meaning of the double jeopardy rule and consciously chose to override that rule in the case of Senate convictions but not acquittals. We find that assumption hard to square with the little evidence we have concerning the framers' and ratifiers' understanding of the possible applicability of the double jeopardy rule to the novel impeachment proceeding created by the Constitution in which the only sanctions upon conviction were removal and disqualification.

     The principle of double jeopardy, though not called by that name, was well known at the time of the founding. And some participants in the process of drafting and ratifying the Constitution may well have thought that the restriction of impeachment sanctions to removal and disqualification did not remove impeachment trials from the principle's operation. The citizens of Sutton, Massachusetts, for example, responding in 1778 to a draft state constitution that included an impeachment judgment clause very similar to what was later included in the federal constitution, expressed their conviction that a provision for "two trials and two punishments for one crime" was "shocking to humane Nature!"⁴⁷

     We think it unlikely, though, that most of the framers or ratifiers had such a clear view that the double jeopardy rule applied to the new species of impeachment trial they had created. Indeed, the formulations of the rule in the sources upon which the framers and ratifiers most heavily relied restricted its reach to cases where the defendant's life was at stake. Blackstone, for example, stated the governing maxim as "no man is to be brought into jeopardy of his life, more than once, for the same offence."⁴⁸ Other leading writers on criminal law expressed the principle in similar terms.⁴⁹ When, just two years after the drafting of the Constitution, the First Congress proposed a double jeopardy clause as part of the Bill of Rights amendments, it too restricted the principle's reach, using the phrase "life or limb." Even if "life" and "life or limb" in this context were understood to encompass all felonies,⁵⁰ and thus some statutory offenses for which the penalties were significant terms of imprisonment, those expressions still limited the reach of the double jeopardy principle to cases where at least the defendant's liberty was at stake.⁵¹ On that understanding, a proceeding in which conviction could bring no more than removal and disqualification simply did not amount to an instance of jeopardy.

     A number of comments by participants in the framing and ratification of the Constitution support this view of the relationship between the double jeopardy rule and the new American impeachment process. Those comments interpret the restriction of impeachment sanctions to removal and disqualification as a decisive break with the English practice of criminal punishments in impeachments and thus view those limited sanctions as distinct from the normal criminal punishments that were necessary to place someone in jeopardy.

     At the Constitutional Convention, Gouverneur Morris explained his shift from opposition to, to support of, presidential impeachment in part based on the limited nature of the punishments the court of impeachment should be empowered to impose. "Our Executive," Morris explained, "was not like a magistrate having a life interest, much less like one having an hereditary interest in his office. He may be bribed by a greater interest to betray his trust; and no one would say that we ought to expose ourselves to the danger of seeing the first Magistrate in foreign pay without being able to guard agst it by displacing him. . . . The Executive ought therefore to be impeachable for treachery; Corrupting his electors, and incapacity were other causes of impeachment. For the latter he should be punished not as a man, but as an officer, and punished only by degradation from office."⁵² Morris thus clearly distinguished between mere removal from office, a sanction aimed at protecting the public from corrupt or otherwise dangerous officials, and regular criminal punishments, aimed at preventing crime by invading the offender's liberty or property.

     Participants in the ratification debates similarly pointed out that the punishments imposable by the Senate were political, not criminal, sanctions, aimed more at protecting the integrity of the government than at penalizing the offender. Tench Coxe, in one of his American Citizen essays, stressed that the Senate "can only, by conviction on impeachment, remove and incapacitate a dangerous officer, but the punishment of him as a criminal remains within the province of the courts of law . . . ."⁵³ In another essay, Coxe made the same point more fully. The Senate, as the impeachment court, "can produce no punishment in person or property, even on conviction. Their whole judicial power lies within a narrow compass. They can take no cognizance of a private citizen and can only declare any dangerous public officer no longer worthy to serve his country. To punish him for his crimes, in body or estate, is not within their constitutional powers."⁵⁴ In the first North Carolina ratifying convention, William Lenoir made the same point more concisely. The punishment for conviction on impeachment, he noted, was "[o]nly removal from office and future disqualification. It does not touch life or property."⁵⁵ Thus, if they thought about a double jeopardy problem at all, many among the framers and ratifiers probably thought the restriction on impeachment sanctions in the first part of the Impeachment Judgment Clause took care of the problem. Whether for that reason or because they thought the Impeachment Judgment Clause simply did not address the issue, James Wilson and Edmund Pendleton concluded (as did Representative Dana, Justice Story, and perhaps St. George Tucker) that the Impeachment Judgment Clause allowed prosecution following acquittal by the Senate.

     The expressio unius reading of the Impeachment Judgment Clause assumes that the founding generation understood an impeachment trial to be an instance of jeopardy within the meaning of the double jeopardy rule. The evidence on point is sparse, but much of it supports the opposite conclusion, namely, that the framers and ratifiers believed that an impeachment trial where only removal and disqualification were at stake did not constitute an instance of jeopardy.

     B. Structural Considerations

     Our examination of the Impeachment Judgment Clause's text and history reveals little support for reading into it an implied prohibition on the criminal prosecution of those acquitted by the Senate. At the same time, while there is some support in the history for the proposition that criminal trial could follow Senate acquittal, that evidence is hardly decisive. Text and history ultimately leave the question unresolved. Given that basic uncertainty, three structural considerations lead us to conclude that acquittal by the Senate should not prevent regular prosecution. The first rests on the special function of impeachment within the scheme of separation of powers. The second and third rest on the distinctive qualities of impeachment verdicts by the Senate as compared to verdicts by criminal juries.

     The first structural consideration is perhaps the most fundamental. Impeachment and criminal prosecution serve entirely distinct goals. Impeachment is one of several tools placed in the hands of Congress in order to enable it to check the other branches and thus to maintain the proper separation of powers. The limitation on impeachment sanctions to removal and disqualification from office and the requirement that removal be mandatory upon conviction show that impeachment is designed to enable Congress to protect the nation against officers who have demonstrated that they are unfit to carry out important public responsibilities, not to penalize individuals for their criminal misdeeds. The limitation on sanctions imposable by the Senate reflects the conviction that the national legislature is not to be trusted with dispensing criminal punishments, sanctions aimed not at protecting the integrity of the government's operations but at penalizing individuals by taking away their life, liberty, or property. Thus the Impeachment Judgment Clause's limitation on Senatorial sanctions is of a piece with the Bill of Attainder Clause and the Ex Post Facto Clause, provisions in the Constitution also aimed at breaking decisively with the long English practice of legislatively imposed punishments. Under our constitutional system, the job of determining guilt that may result in criminal punishment is reserved to the courts, where both the original Constitution and the Bill of Rights ensure that individuals will not suffer those especially severe sanctions without being afforded a number of procedural protections. Impeachment serves the remedial and protective function of guarding the government's integrity and thus its effective functioning, a function appropriately entrusted to the legislature. Trials that may lead to the imposition of criminal punishments must be supervised by the courts, the branch of the national government both suited and required to guard the defendant's procedural rights.⁵⁶

     A second, closely related structural consideration favoring prosecution following acquittal by the Senate is that an acquittal by the Senate may well rest on a legal judgment rather than on a judgment that the respondent did not commit the acts alleged in the articles of impeachment, that is, a judgment that the respondent is not factually guilty. Most often that non-factual basis for acquittal will be that although the respondent carried out the charged acts, those acts do not amount to "high crimes or misdemeanors."⁵⁷ Sometimes, though, it may be that the Senate lacks the authority to try the respondent. Indeed, of the eight instances in which the Senate has failed to convict officers impeached by the House, most may fairly be attributed in significant part either to qualms about the charged conduct meeting the constitutional standard for impeachable offenses or to jurisdictional doubts.⁵⁸ It makes little sense for a judgment unrelated to factual guilt to prevent bringing a former official to justice for criminal conduct. As the Court has explained in justifying the distinction between re-trials following reversals of convictions due to trial errors and those due to evidentiary insufficiency, "It would be a high price indeed for society to pay were every accused granted immunity from punishment because of any defect sufficient to constitute reversible error in the proceedings leading to conviction." United States v. Tateo, 377 U.S. 463, 466 (1964); see Burks, 437 U.S. at 15-16. Similarly, it would be a high price indeed for society to pay for every accused official spared removal from office by the Senate's judgment that the offense fell short of the constitutional standard, or that it lacked the authority to try the official, to be free - unlike citizens possessing no federal office - from prosecution for criminal conduct.

     Of course, in the case of trials before the courts our double jeopardy jurisprudence does give jury verdicts of not guilty, regardless of their basis, an absolutely prohibitive effect on re-trials for the same offenses. See, e.g., Sanabria v. United States, 437 U.S. 54, 64, 75 (1978). While jury verdicts of not guilty are normally based on insufficiency of the government's proof, they may be based as well on jurors' judgments unconnected to the defendant's factual innocence, for example, on their disagreement with the judge's statement of the governing law, their belief that the likely punishment is excessive, or their disapproval of what they take to be improper prosecutorial motives or methods. Although juries lack the legal right to engage in such nullification absent legislative authorization, see Sparf & Hansen v. United States, 156 U.S. 51, 59-107 (1895), they undoubtedly possess the power to do so.⁵⁹ If juries' ability to acquit against the evidence does not diminish the effect of their acquittals as bars to successive prosecutions, why should the Senate's authority to acquit on legal grounds justify relaxing the double jeopardy effect of their acquittals?

     The difference between the two cases lies in the different functions served by the Senate in an impeachment trial and by a jury in a criminal trial. The Senate's verdict is different from a criminal jury's in two crucial respects.

     First, except in cases of treason or bribery, the Senate's judgment, unlike a jury's, inescapably involves a crucial legal judgment: whether the conduct charged constitutes a "high crime or misdemeanor." The jury in a criminal trial is above all a fact-finder; at least in the federal system, its ability to nullify based on its own view of the law is tolerated only because it is essential to preserving the independence of juries from judicial coercion and second-guessing. While the Senate in an impeachment trial takes on the jury's role of fact-finder, it also assumes the judge's role of interpreter of the governing law. Far from constituting a power necessary to protect another function, the Senate's judgment whether the charged offenses constitute "high crimes and misdemeanors" is an essential part of its function, one entrusted to it by the Constitution.

     Second, and more importantly, the Senate's verdict differs from a jury's because the legal judgment the Senate must make is also a special kind of political judgment. The drafters of the Constitution probably assigned the Senate, rather than the regular courts, the task of trying impeachments in part because they recognized that impeachment trials necessarily involve making political judgments. As Hamilton observed in Federalist 65, impeachable offenses "are of a nature which may with peculiar propriety be denominated POLITICAL, as they relate chiefly to injuries done immediately to the society itself."⁶⁰ The Senate's judgment is political in two senses. The uncertain contours of the phrase "high crimes and misdemeanors" mean they must in each case determine whether the charged conduct constitutes a sufficiently serious breach of the public trust to warrant conviction. That determination will appropriately draw on their knowledge of history, their understanding of the character of the office involved, and their realistic appraisal of the derelictions charged. Their determination will necessarily be shaped by the Constitution's mandate that conviction means removal from office. U.S. Const. art. II, § 4. In order to convict an officer, they must be convinced that his conduct merits his loss of position. In the case of the President, who has been elected by the entire nation (and who cannot remain in office for more than four years without again facing the electorate), they must decide whether to undo the will of the people.⁶¹ Moreover, the necessary link between conviction and removal introduces a second political dimension to the Senate's judgment as well. Even if they conclude that the charged conduct would normally merit removal, they must weigh the strength of that conviction against their judgment about the harmful consequences for the nation of removal at a particular moment in our nation's history. If, for example, our country were in the midst of a war, the Senate might well conclude that an acquittal of the President would be the wiser course simply because his removal would be too costly to the successful prosecution of the war.

     The necessarily legal and political judgment embodied in a Senate acquittal is distinct from a determination whether the charged conduct violates the regular criminal laws and does not turn on the determination of factual guilt or innocence. It is ultimately the unreviewability of the jury's making of that factual determination that drives the absoluteness of the ban on re-trial for offences of which a jury has acquitted a defendant. No such institutional imperative requires a similar ban following Senate acquittals. On the contrary, the unavoidably legal and political character of Senate acquittals suggests the inappropriateness of such a ban.⁶²

     A third structural reason that acquittal by the Senate should not prevent criminal prosecution flows from the framers' concern that partisan loyalties or popular sentiment might influence the Senate's decision to convict or acquit. One of the reasons the framers limited the punishments for conviction on impeachment was their fear that impeachments were liable to partisan abuse. As Hamilton noted in Federalist 65, "[t]he prosecution of them . . . will seldom fail to agitate the passions of the whole community, and to divide it into parties, more or less friendly or inimical, to the accused. In many cases, it will connect itself with pre-existing factions, and will inlist all their animosities, partialities, influence and interest on one side, or on the other; and in such cases there will always be the greatest danger, that the decision will be regulated more by the comparitive strength of parties than by the real demonstrations of innocence or guilt."⁶³ The Constitution's requirements that Senators take an oath before convening as an impeachment court and that a two-thirds vote is necessary for conviction were designed to guard against the influence of these political forces. See U.S. Const. art. I, § 3, cl. 6. Its specification that the Chief Justice rather than the Vice President should preside when the President is tried reflects a similar concern with impeachment verdicts being swayed by immediate political interests. See id. If the Vice President presided, he might encourage conviction so as to boost himself into the Presidency, especially if the Vice President and President were rivals, a realistic possibility before the 12^th Amendment reformed the electoral college in 1804. But, as a number of participants in the ratification debates pointed out, partisanship and transitory political passions may sway the Senate to acquit as well as to convict.⁶⁴ Just as the possibility of partisan convictions helps explain the limitation on impeachment punishments and the lifting of the double jeopardy bar for Senate convictions, so the possibility of partisan acquittals supports the lifting of the double jeopardy bar for Senate acquittals.

     C. The Double Jeopardy Clause

     The Double Jeopardy Clause of the Fifth Amendment provides that "[n]o person . . . shall . . . be subject for the same offence to be twice put in jeopardy of life or limb." U.S. Const. amend. V. For several reasons, we think a party acquitted by the Senate may not rely on the Double Jeopardy Clause as a bar to prosecution in the courts for the same offenses.

     1. Original Understandings

     First, the history of the Double Jeopardy Clause suggests that its drafters understood the phrase "in jeopardy of life or limb" to exclude impeachment proceedings. The Clause's legislative history, like that of the Bill of Rights amendments as a whole, is sparse. We know that in Madison's proposal to the House, what became the Double Jeopardy Clause was expressed in these terms: "No person shall be subject, except in cases of impeachment, to more than one punishment or one trial for the same offence."⁶⁵ Several House members suggested deleting the phrase "or one trial," but their motion was defeated.⁶⁶ The version adopted by the House followed Madison's phrasing.⁶⁷ In response, the Senate initially adopted a version of the clause that deleted the reference to impeachment and added the phrase "life or limb": "No person shall be subject to be twice put in jeopardy of life or limb by any public prosecution for the same offence, . . . ."⁶⁸ The Senate adopted the ultimate wording by omitting "by any public prosecution" when it combined the double jeopardy provision with the other clauses that make up what became the Fifth Amendment.⁶⁹

     One might argue that the Senate's deletion of the House's exception for impeachments suggests an intent to include impeachments within the Double Jeopardy Clause's scope.⁷⁰ But while we lack direct evidence of the purpose of the Senate's change in language, that explanation seems unlikely. The wording of related amendments suggests that a more likely explanation for the removal of the exception for impeachments was a recognition that the use of the phrase "life or limb" by itself restricted the reach of the clause to a subset of ordinary criminal cases. In Madison's original proposal, the jury trial and grand jury guarantees had been grouped together in an amendment separate from the double jeopardy guarantee. The jury trial provision included an express exception for impeachments and the grand jury clause an implicit one: "The trial of all crimes (except in cases of impeachments, and cases arising in the land or naval forces, or the militia when on actual service in time of war, or public danger,) shall be by an impartial jury of the vicinage . . . ; and in all crimes punishable with loss of life or member, presentment or indictment by a grand jury shall be an essential preliminary, . . . ."⁷¹ Clearly impeachments, not indictments, were the preliminary step toward trial before the Senate. The emphasized phrase seems to have been understood to exclude impeachment proceedings⁷² and to identify a group of serious crimes, probably most if not all felonies (when tried in the regular courts).⁷³ The Senate's substitute for the House version of the Double Jeopardy Clause omitted the express exception for impeachments and added the phrase "life or limb" in one fell swoop. Given Madison's earlier restrictive use of the similar phrase "loss of life or member," it makes more sense to understand the Senate's deletion of the impeachment exception as an acknowledgment that the use of "life or limb" made the express exception for impeachments unnecessary than to view the deletion in isolation as an attempt to bring impeachment within the Double Jeopardy Clause's reach.

     Second, our interpretation of the legislative history of the Double Jeopardy Clause fits with the dominant understanding of the reach of the double jeopardy rule at the time of the founding. As we explained above, under that understanding the rule was limited to proceedings that placed the defendant in risk of at least liberty if not life, and thus a trial in which removal and disqualification are the only possible sanctions does not fit within the rule.⁷⁴

     2. Current Double Jeopardy Doctrine

     The Court uses a two-step approach to determining whether a proceeding constitutes an instance of jeopardy. First, it looks to the legislature's intent. See, e.g., Hudson, supra, 522 U.S. at 103. If the legislature intended the proceeding to be criminal, then the Double Jeopardy Clause applies. If the legislature intended the proceeding to be civil, then the Court looks to a series of factors designed to identify criminal punishments. If those factors clearly show that the legislature has provided for the imposition of criminal punishment, the Double Jeopardy Clause will apply despite the legislature's claim that the proceeding is civil.

     At both the first and second steps of this method, we think the better view is that an impeachment trial does not constitute an instance of jeopardy within the meaning of the Double Jeopardy Clause.

     At the first step, one might argue that the references to impeachment in the Constitution suggest that it is a criminal proceeding. Article III, Section 2, Clause 3 mandates that the "Trial of all Crimes, except in cases of impeachment, shall be by Jury." Article II, Section 4's definition of impeachable offenses limits that group to treason, bribery "or other high Crimes and Misdemeanors." The President's pardon power, in Article II, Section 2, Clause 1, extends to all "Offenses against the United States, except in cases of impeachment."⁷⁵

     We find this view unconvincing for several reasons. First, the uses of the term "crimes" in connection with impeachments occur precisely in contexts that distinguish impeachments from regular criminal proceedings. The reference in Article III, Section 2, Clause 3 establishes that parties who have been impeached, unlike regular criminal defendants, are not entitled to a jury, one of the most fundamental safeguards in our system of criminal justice. The definition of impeachable offenses in Article II, Section 4 was designed to capture more than ordinary crimes. Second, as we have tried to show above, the framers and ratifiers understood the limited nature of the sanctions available to the Senate as marking out impeachments as distinct from regular criminal proceedings.⁷⁶

     Third, the practice of the Senate under the Constitution suggests that, while impeachment trials are akin to criminal trials in many respects, they are fundamentally different from criminal trials in ways that remove Senate trials from the reach of the double jeopardy rule. The clearest examples of this are perhaps the Senate's standard of proof and its methods for taking evidence. Senators have not considered themselves bound to apply the beyond-a-reasonable-doubt standard of proof required in criminal trials.⁷⁷ See In re Winship, 397 U.S. 358, 361-364 (1970) (Due Process Clause mandates beyond-a-reasonable-doubt standard of proof in criminal trials). In one recent impeachment, the Senate overwhelmingly rejected a motion requiring that standard.⁷⁸ Since the early part of this century, moreover, the Senate has empowered a committee to take evidence on its behalf rather than hearing the evidence itself, and the Senate has now employed that method on three occasions.⁷⁹ Such a delegation of the responsibility to hear the evidence conflicts with our understanding of the factfinder's essential role in a criminal trial.

     The text of the Constitution, the evidence concerning the founders' understanding of the new process of impeachment they were creating, and the Senate's practice suggest that the framers and ratifiers conceived of impeachment trials, as Judge Gesell has observed, as sui generis proceedings, bearing some characteristics of criminal trials but clearly lacking many others. Hastings v. United States Senate, 716 F. Supp. 38, 41 (D.D.C. 1989). Although the evidence is hardly unmixed, we think it weighs in favor of the view that the framers and ratifiers did not consider an impeachment trial an instance of jeopardy within the meaning of the double jeopardy rule.

     In the regular case of legislatively created proceedings, the Court has developed and employed the second step of its two-step test in order to prevent legislators from evading the requirements of the Double Jeopardy Clause simply by labeling a proceeding civil rather than criminal or calling a monetary sanction a tax rather than a fine.⁸⁰ But when it comes to the framers' establishment of a new and distinctive process of impeachment, this need to second-guess legislative judgments by looking behind direct evidence of intent simply does not arise. As a result, we believe, when examining a special proceeding whose relationship to regular criminal proceedings the framers defined, the first step of the process should end our analysis (especially if the evidence at that step is clear).

     Even if one were to go on to the second step of current double jeopardy analysis and judge whether an impeachment trial is a criminal proceeding by determining whether the sanctions upon conviction are criminal punishments,⁸¹ the result would only confirm the conclusion reached so far: that an impeachment trial is not a criminal proceeding within the meaning of the Double Jeopardy Clause. With the possible exception of a few years in the early 1990s, the Supreme Court has for several decades applied an open-ended multi-factor test to determine whether a sanction constitutes criminal punishment. Originally developed in a non-Double Jeopardy case, Kennedy v. Mendoza-Martinez, 372 U.S. 144 (1963), the seven factors are, in the Court's view, "neither exhaustive nor dispositive," United States v. Ward, 448 U.S. 242, 249 (1980), but "useful guideposts," Hudson, 522 U.S. at 99. They are: "(1) "[w]hether the sanction involves an affirmative disability or restraint"; (2) "whether it has historically been regarded as a punishment"; (3) "whether it comes into play only on a finding of scienter"; (4) "whether its operation will promote the traditional aims of punishment-retribution and deterrence"; (5) "whether the behavior to which it applies is already a crime"; (6) "whether an alternative purpose to which it may rationally be connected is assignable for it"; and (7) "whether it appears excessive in relation to the alternative purpose assigned." Id. at 99-100 (internal quotation marks omitted).

     Of the seven Mendoza-Martinez factors, five strongly indicate that removal is not criminal punishment, one points more tentatively in that direction, and one points tentatively towards treating removal as a criminal sanction. Disqualification presents a much closer question because at least one, and possibly two, of the factors that favor treating removal as a non-criminal sanction suggest that disqualification is a criminal punishment; moreover, in a post-Civil War decision, the Supreme Court in dictum characterized disqualification in an impeachment judgment as punishment at least for purposes of bill of attainder and ex post facto analysis. See Cummings v. Missouri, 71 U.S. (4 Wall.) 277, 320 (1867). Still, we believe that those factors in the case of disqualification are not dispositive and that the Mendoza-Martinez factors as a whole still support classifying disqualification as a non-criminal sanction.

     The first Mendoza-Martinez factor, whether the sanction is an "affirmative disability or restraint," is the one that weighs in favor of treating removal as a non-criminal sanction while its significance for disqualification is less clear. Neither removal nor disqualification imposes an affirmative restraint because neither restricts the physical liberty of the sanctioned individual. In addition, removal clearly does not constitute an affirmative disability because it imposes no lasting restrictions on the offender.

     The question whether disqualification from all federal offices is an affirmative disability is a close one, and we think the better view is that it does constitute such a disability. The difficulty of the question stems in part from a degree of inconsistency between the Court's bill of attainder and ex post facto cases, in which it developed the notion of disability as punishment, and its double jeopardy decisions.

     The Court first used the phrase "affirmative disability or restraint" three years before Mendoza-Martinez in Flemming v. Nestor, 363 U.S. 603 (1960), a challenge to a provision of the Social Security Act taking away Social Security benefits from all individuals who were deported for certain reasons, including (in Nestor's case) past membership in the Communist Party. The Supreme Court upheld the law, rejecting, among other contentions, claims that the statute constituted a bill of attainder or an ex post facto law. Necessary to both contentions was the proposition that the sanction constituted punishment. The Court explained that the punitive character of a sanction is a question of legislative purpose. See id. at 616; cf., e.g., DeVeau v. Braisted, 363 U.S. 144, 160 (1960). In determining that the statute before it did not have a punitive purpose, the Court considered several circumstances, the first of which was that "the sanction is the mere denial of a noncontractual governmental benefit. No affirmative disability or restraint is imposed . . . ." Flemming, 363 U.S. at 617.

     The Flemming Court looked back to two post-Civil War decisions striking down laws on bill of attainder and ex post facto grounds. In Cummings v. Missouri, 71 U.S. (4 Wall.) 277 (1867), the Court invalidated a provision of the Missouri constitution requiring all those who would hold a state office, teach, be an officer of a corporation, an attorney, or a clergyman to take an oath affirming, among other things, that they had never aided or expressed sympathy for those engaged in rebellion against the United States or evaded the draft. Cummings was a Catholic priest who had not taken the oath and yet was serving a church in the state, and he had been convicted and fined. Referring to the "disabilities" imposed by the state constitution, the Court rejected Missouri's contention that punishment was restricted to deprivations of life, liberty, or property:

     The deprivation of any rights, civil or political, previously enjoyed, may be punishment, the circumstances attending and the causes of the deprivation determining this fact. Disqualification from office may be punishment, as in cases of conviction upon impeachment. Disqualification from the pursuits of a lawful avocation, or from positions of trust, or from the privilege of appearing in the courts, or acting as an executor, administrator, or guardian, may also, and often has been, imposed as punishment. . . . .

     The theory upon which our political institutions rest is, that all men have certain inalienable rights--that among these are life, liberty, and the pursuit of happiness; and that in the pursuit of happiness all avocations, all honors, all positions, are alike open to every one, and that in the protection of these rights all are equal before the law. Any deprivation or suspension of any of these rights for past conduct is punishment, and can be in no otherwise defined.

Id. at 320, 321-22 (emphasis added). Punishment, in the Court's view, therefore "embrac[ed] deprivation or suspension of political or civil rights." Id. at 322.⁸²

     In Ex parte Garland, 71 U.S. (4 Wall.) 333 (1867), decided the same day, the Court struck down for similar reasons a federal law making the taking of a similar oath concerning participation in or support for the Confederate cause a condition for practice of law in federal court. The Court stated that "exclusion from any of the professions or any of the ordinary avocations of life for past conduct can be regarded in no other light than as punishment for such conduct." Id. at 377.⁸³

     The Court next addressed these issues in United States v. Lovett, 328 U.S. 303 (1946), in which the Court invalidated as a bill of attainder an appropriations act that prohibited any federal agency from paying any further compensation to three particular federal employees, apparently because of the belief that they were, in the words of the act's principal sponsor, "'crackpot, radical bureaucrats' and affiliates of 'Communist front organizations.'" Id. at 308-09. After an examination of the act's origins, the Court concluded that its purpose was "permanently to bar them from government service," id. at 313, and so it determined to judge the act on that basis. The Court likened the act to those voided in Cummings and Garland because it "'operate[d] as a legislative decree of perpetual exclusion' from a chosen vocation. Ex Parte Garland, supra, [71 U.S.] at 377. This permanent proscription from any opportunity to serve the Government is punishment, and of a most severe type. It is a type of punishment which Congress has only invoked for special types of odious and dangerous crimes, such as treason, 18 U.S.C. 2; acceptance of bribes by members of Congress, 18 U.S.C. 199, 202, 203; or by other government officials, 18 U.S.C. 207; and interference with elections by Army and Navy officers, 18 U.S.C. 58." Id. at 316.⁸⁴

     The broad statements in Cummings, Garland, and Lovett that permanent exclusion from a profession or federal office or employment constitute disabilities and punishments stand in some tension with the Court's pronouncements in two of its leading double jeopardy decisions. In Helvering v. Mitchell, 303 U.S. 391 (1938), the Court's seminal New Deal decision marking its willingness to give Congress greater leeway to impose civil sanctions free from the constraints of the Double Jeopardy Clause, the Court found a special "tax" imposed on those who fraudulently underreported their income on their federal tax return to be a civil sanction and thus imposable despite the defendant's prior acquittal of a criminal charge, based on the same acts, of fraudulently evading payment of his full income tax bill. In assessing whether the special tax was a punitive or remedial sanction, the Court observed that one remedial sanction "which is characteristically free of the punitive criminal element is revocation of a privilege voluntarily granted." Id. at 399. As examples, the Court gave deportation of aliens and disqualification of attorneys to practice before certain courts. Id. at 399 n.2. Sixty years later, in its most recent decision to address these issues, the Court expressly endorsed that conclusion. In Hudson v. United States, 522 U.S. 93 (1997), the Court held that permanent exclusion from employment by any federally insured bank did not constitute criminal punishment. It reached that conclusion by applying the Mendoza-Martinez factors, and it stated that "the sanctions imposed do not involve an 'affirmative disability or restraint,' as that term is normally understood. While petitioners have been prohibited from further participating in the banking industry, this is 'certainly nothing approaching the "infamous punishment" of imprisonment.' Flemming v. Nestor, 363 U.S. 603, 617 (1960)." Id. at 104.⁸⁵

     Whatever tension may exist between the more sweeping language in Cummings, Garland, and Lovett, on the one hand, and Helvering and Hudson, on the other, the latter decisions do not directly reject the Court's earlier statements as applied to disqualification from federal office. Even if one took the view (supported perhaps by Garland, but not Lovett) that the right to hold congressionally established federal offices is a "privilege voluntarily granted," Helvering, 303 U.S. at 399, it would be much more difficult to characterize the right to run for those elective offices created by Constitution in similar terms. The qualifications for those offices are established by the Constitution, and may not be modified by either Congress or the states. See Powell v. McCormack, 395 U.S. 486 (1969); U.S. Terms Limit, Inc. v. Thornton, 514 U.S. 779 (1995). Disqualification from those constitutionally created offices, if not from legislatively created ones, constitutes an affirmative disability.⁸⁶

     The second Mendoza-Martinez factor is whether the sanction "has historically been regarded as punishment." Although the historical record is not unambiguous, we think that, as discussed earlier in this memorandum, both the evidence concerning the framing and ratification of the Constitution and the predominant views expressed by participants in impeachment trials support the judgment that removal and disqualification for conviction upon impeachment have been seen not as criminal punishments but as sanctions with principally remedial goals. The actions by the House and Senate in the 1980s judicial impeachments discussed in the next part of this memorandum, each of which involved a defendant previously prosecuted in the courts, also support that conclusion.

     At least two considerations may be raised against this view, however. First, while removal has an obvious remedial goal and effect, disqualification's remedial function may be less clear. As the record of impeachment trials suggests, though, Representatives and Senators have seen disqualification's non-punitive purpose as preventive or protective. Disqualification prevents those who have abused positions of public trust from doing so again and thus protects the integrity of the government's activities. Admittedly, in one of its bill of attainder cases, the Court has expressed some skepticism about this sort of argument. In United States v. Brown, 381 U.S. 437 (1965), the Court undid as a bill of attainder a criminal statute prohibiting anyone who had been a member of the Communist Party within the past five years from being an officer or employee of a labor union. The Solicitor General argued that the statute's prohibition on union employment or officership did not constitute punishment because it "was enacted for preventive rather than retributive reasons - that its aim is not to punish Communists for what they have done in the past, but rather to keep them from positions where they will in the future be able to bring about undesirable events," id. at 456-457, an argument the Court had apparently embraced fifteen years earlier in American Communications Ass'n v. Douds, 339 U.S. 382, 413-14 (1950). This time around, the Court was unwilling to follow the Government's reasoning: "It would be archaic to limit the definition of 'punishment' to 'retribution.' Punishment serves several purposes; retributive, rehabilitative, deterrent--and preventive. One of the reasons society imprisons those convicted of crimes is to keep them from inflicting future harm, but that does not make imprisonment any the less punishment." Brown, 381 U.S. at 458. But this statement from Brown is inapposite to the question before us. That a criminal punishment may aim to prevent further criminality does not mean that all sanctions with preventive ends are criminal. Indeed, most regulatory sanctions count prevention among their prominent goals. It is the centrality of prevention, as compared to retribution and deterrence, that helps mark disqualification by the Senate as a non-criminal sanction. See Hudson, 522 U.S. at 105 (the presence of one arguably punitive purpose is insufficient to brand a sanction as criminal).

     A second, more historically grounded, objection to the view that removal and disqualification is not punishment rests on the significant number of federal criminal statutes that have authorized removal or disqualification from federal office as a punishment for crime.⁸⁷ At the state level, statutory or constitutional provisions for removal and disqualification of officials convicted of crime are even more common.⁸⁸ None of these federal statutes provides for disqualification from office as the sole result of a conviction,⁸⁹ and all but one of them may properly be viewed, as is true of similar state-law provisions, as mandating collateral and remedial consequences of criminal conviction rather than as defining one of the punishments for the specified crimes. See United States v. Waddell, 112 U.S. 76, 82 (1884) ("this language . . . is not the sentence of the court, but an indelible disgrace affixed to the party convicted by the declaration of the law itself").⁹⁰ Indeed, so learned a jurist as Justice Story wrote in 1833 that "[i]n the ordinary course of the administration of criminal justice, no court is authorized to remove, or disqualify an offender, as a part of its regular judgment. If it results at all, it results as a consequence, and not as a part of the sentence."⁹¹ If there have been one or two instances in which disqualification was made part of the punishment itself, they are exceptions to the general pattern of disqualification as a legislatively mandated collateral consequence of criminal conviction, designed to protect the public from unfit officers rather than to punish the offender convicted of such a crime.

     The import of the third Mendoza-Martinez criterion is uncertain. On the one hand, several considerations suggest that a finding of scienter is not absolutely necessary for impeachment. Other than by implication in the definition of impeachable offenses, the Constitution does not impose a scienter requirement. Moreover, in the second federal impeachment, the Senate convicted and removed a federal judge for drunkenness on the bench and for flagrantly erroneous rulings in a forfeiture proceeding despite the fact that it heard evidence submitted by the judge's son that the judge was insane and had been at the time of the charged conduct. The conviction of Judge Pickering has been seen by some as an instance of conviction of a defendant lacking a criminal mental state.⁹² Furthermore, given that the ultimate touchstone for conviction upon impeachment is conduct that clearly demonstrates unfitness for office, before more modern solutions, see U.S. Const. amend. XXV (providing procedures for coping with Presidential incapacity); Judicial Councils Reform and Judicial Disability Act, 28 U.S.C. § 372 et seq. (authorizing methods short of removal to cope with judicial incapacity), impeachment might well have been the only avenue for removal of officers who were clearly incapable of carrying out their duties.⁹³ The difficulty of determining whether an impeachable offense must include an element of scienter stems in part from the fact that conduct need not be previously defined as criminal in order to support an impeachment charge and in part from the somewhat uncertain meaning of the term "scienter."

     On the other hand, two considerations support the conclusion that scienter is a necessary element of an impeachable offense. First, the evolution of the language defining impeachable offenses at the Constitutional Convention suggests that the framers sought to exclude mere negligence in the meeting of official responsibilities.⁹⁴ The phrase originally adopted to define the scope of impeachable conduct was "malpractice or neglect of duty."⁹⁵ Later on the Convention considered limiting impeachable offenses to treason and bribery, or perhaps "corruption" as well.⁹⁶ Near the end of their meetings, several delegates thought this definition was too limited and suggested adding "maladministration." Madison, however, objected that this term was too loose and would leave the President serving at the "pleasure of the Senate."⁹⁷ The Convention then settled on "other high crimes and misdemeanors" apparently as a compromise, broadening the impeachable offenses beyond treason and bribery but restricting them more narrowly than mere "maladministration." That progression suggests that the framers considered something beyond negligence in the handling of official responsibilities as necessary to impeachable conduct, trusting that elections would provide sufficient check against the less culpable forms of misconduct.

     Second, as Madison's comment about the danger of impeachment being wielded as a tool of political control suggests, impeachment should not be used as a means to punish officials for reasonable, good-faith disagreements over the reach of statutory or constitutional provisions. The acquittal of Justice Chase, for example, stands for the proposition that impeachment should not lie simply because Congress concludes that a judge has taken an erroneous view of the law.⁹⁸ The acquittal of President Johnson similarly stands for the proposition that a President should not be impeached simply because he refuses to carry out a law that he reasonably believes is unconstitutional. Cf. Presidential Authority to Decline to Execute Unconstitutional Statutes, 18 Op. O.L.C. 199 (1994) (outlining circumstances in which President may appropriately decline to execute statutory provisions he believes are unconstitutional).

     The fourth Mendoza-Martinez factor is whether the sanction will "promote the traditional aims of punishment - retribution and deterrence." We think the answer here is "no." As the discussion of the Impeachment Judgment Clause during the ratification debates suggests, contemporaries understood the regular criminal punishments available in addition to removal and disqualification as the vehicles for exacting retribution. While removal and disqualification are likely to have, and were intended to have, some deterrent effect, that is true of virtually any governmental exaction. Accordingly, the Court has reasoned, "the mere presence of [a deterrent] purpose is insufficient to render a sanction criminal, as deterrence 'may serve civil as well as criminal goals.'" Hudson, 522 U.S. at 105 (citations omitted); see also United States v. Ursery, 518 U.S. 267, 292 (1996).

     Under the fifth Mendoza-Martinez factor, "whether the behavior to which [they] appl[y] is already a crime," the sanctions that the Senate may impose are not criminal punishments. Of course, only conduct that is already defined as criminal will provide a basis for subsequent criminal prosecution of an impeached official, and thus only cases involving criminal conduct will raise the double jeopardy issue addressed in this memorandum. But as the development of impeachment law before the Constitution, the debates at the time of the Founding, and the history of impeachments under the Constitution show, despite the protestations of many impeachment defendants to the contrary, officials may be impeached and convicted for conduct that is not prohibited by the regular criminal laws.⁹⁹

     The sixth and seventh Mendoza-Martinez factors are whether a purpose other than punishment may "rationally" be assigned to the sanction and whether the sanction "appears excessive in relation to the alternative purpose assigned." In our view, these are the most important considerations for they go most directly to the ultimate question of legislative (or drafters' and ratifiers') purpose. The same sanction may have either a punitive or a non-punitive purpose and thus may be characterized as criminal punishment in one circumstance and as a civil sanction in another. Compare, e.g., Flemming, 363 U.S. at 617 (imprisonment as punishment); Mendoza-Martinez, 372 U.S. at 165 (deprivation of nationality in one section of the Immigration and Nationality Act as punishment) with Bell v. Wolfish, 441 U.S. 520, 533-39 (1979) (imprisonment in the context of reasonable pre-trial detention not punishment); Mendoza-Martinez, 372 U.S. at 164 (interpreting Perez v. Brownell, 356 U.S. 44 (1958) as based on view that deprivation of nationality in another section of Immigration and Nationality Act not punishment). Ultimately it is the purpose for which the sanction is applied that will determine its character.

     And when it comes to disqualification, the Court has emphasized, from its post-Civil War bill of attainder decisions to the modern era, that it is the closeness of the fit between the causes of disqualification and the positions from which the individual is disqualified that most clearly reveals a non-punitive purpose. In Cummings, the Court concluded that the disqualifying provision in the Missouri constitution was a penalty largely because it was "evident from the nature of the pursuits and professions of the parties, placed under disabilities by the constitution of Missouri, that many of the acts, from the taint of which they must purge themselves, have no possible relation to their fitness for those pursuits and professions." Cummings, 71 U.S. at 319; see also Garland, 71 U.S. at 379-80. In Dent v. West Virginia, 129 U.S. 114 (1889), a decision upholding West Virginia's medical licensure statute, Justice Field, who had written the majority opinions in both Cummings and Garland, distinguished those decisions by explaining that they turned on the conclusion that "as many of the acts from which the parties were obliged to purge themselves by the oath had no relation to their fitness for the pursuits and professions designated, . . . the oath was not required as a means of ascertaining whether the parties were qualified for those pursuits and professions, but was exacted because it was thought that the acts deserved punishment ." Id. at 126; see also Hawker v. New York, 170 U.S. 189, 198-199 (1898). More recently, in Flemming v. Nestor, 363 U.S. 603 (1960), the Court endorsed the same view: "Where the source of legislative concern can be thought to be the activity or status from which the individual is barred, the disqualification is not punishment even though it may bear harshly upon one affected. The contrary is the case where the statute in question is evidently aimed at the person or class of persons disqualified." Id. at 614.

     Here, the core "source of the legislative concern," abuse of federal office, is precisely "the activity or status from which the individual is barred." The non-punitive purpose which may rationally be assigned to removal and disqualification is keeping government authority out of the hands of those who have demonstrated their disregard for the obligations of public office. In relation to that purpose, these sanctions, far from being excessive, are deftly tailored. Unlike the prohibitions in Cummings and Garland, they do not reach beyond the exact sphere of the misconduct and thus the threat: federal office.

     The Court's statement in Cummings that disqualification in an impeachment judgment constitutes punishment does not dissuade from concluding that such disqualification is not punishment within the meaning of the Double Jeopardy Clause. The statement in Cummings was dictum unsupported by any reasoning concerning the special character or function of impeachment proceedings.¹⁰⁰ Moreover, as the Court's more recent bill of attainder decisions suggest, the range of sanctions that count as punishment for purposes of the Bill of Attainder Clause may well be broader than the range of penalties that amount to criminal punishment under the Double Jeopardy Clause. In Selective Service System v. Minnesota Public Interest Research Group, 468 U.S. 841 (1988), the Court stated that it looks to three considerations in determining whether a statute inflicts punishment for bill of attainder purposes: "(1) whether the challenged statute falls within the historical meaning of legislative punishment; (2) whether the statute, 'viewed in terms of the type and severity of burdens imposed, reasonably can be said to further nonpunitive legislative purposes'; and (3) whether the legislative record 'evinces a congressional intent to punish.'" Id. at 852 (quoting Nixon v. General Services Administration, 433 U.S. 425, 473, 475-76 (1977)). The second of those criteria is quite similar to the sixth Mendoza-Martinez factor. But the Court's recent bill of attainder criteria leave out a number of the Mendoza-Martinez factors that would tend to narrow the class of punitive sanctions - whether the sanction constitutes an affirmative disability or restraint, whether a finding of scienter is necessary, and whether the conduct to which it applies is already criminal. Recognizing that criminal punishments under the Double Jeopardy Clause may form a subset of punishments under the Bill of Attainder Clause also helps relieve the apparent tension between, on the one hand, the bill of attainder decisions' assertions that disqualification from a profession constitutes punishment, and, on the other, Helvering's and Hudson's holdings that bars on participation in particular professions did not amount to punishment within the meaning of the Double Jeopardy Clause.

     On balance, then, we conclude that removal and disqualification when imposed by the court of impeachments are best seen as special civil sanctions rather than as criminal punishments. The historical evidence demonstrating the founders' intent to break with the English tradition of criminal punishments and to codify the American practice of limited impeachment sanctions, the record of impeachment trials showing the House's and Senate's endorsement of that view, and even the criteria of current double jeopardy law all support the conclusion that the sanctions the Constitution places in the Senate's hands are not criminal punishments within the meaning of the Double Jeopardy Clause.¹⁰¹

     III. The 1980s Impeachment Trials

     The "case law" that gives meaning to the constitutionally defined process of impeachment is made largely by Congress. The three impeachment trials carried out during the 1980s bolster the proposition that the Constitution permits prosecution of a former official for the same offenses of which he has been acquitted by the Senate.

     After a 50-year hiatus as a court of impeachments, the Senate tried and convicted three district court judges during the 1980s. In each case, the defendant had previously been prosecuted in the courts. In each case, the defendant challenged the propriety of his impeachment both in court and before Congress. As a result, these proceedings gave the courts and Congress an opportunity to address whether former conviction or acquittal in the courts should bar trial before the Senate for the same offenses. One district court and both houses of Congress concluded that prior criminal judgments did not preclude impeachment and conviction for the same offenses.

     Judge Harry E. Claiborne and Judge Walter L. Nixon were both tried and convicted of federal offenses.¹⁰² Although they were not the first federal judges to be found guilty of crimes while in office, they were the first to refuse to resign their judicial posts.¹⁰³ The House thus impeached them for the offenses of which they had already been convicted (as well as other conduct) and the Senate tried and convicted them and removed them from office.

     Neither Claiborne nor Nixon directly argued to the House or Senate that double jeopardy should bar their impeachment and trial. On the contrary, in Claiborne's case the House managers contended that the House and Senate should each be bound by the guilty verdicts rendered by the jury that had sat in the judge's criminal trial, and Claiborne argued that the impeachment process was distinct from regular prosecution and that separation of powers and due process concerns required the House and Senate to do their own factfinding.¹⁰⁴ In both cases, the Senate accepted the evidence from the prior criminal trials, took some evidence of its own, and apparently did not consider itself bound by the juries' verdicts.¹⁰⁵

     Although neither the House nor the Senate squarely passed on double jeopardy challenges in the Claiborne and Hastings cases, the fact that they impeached, tried, and convicted the defendants indicates that they found that prior conviction was no bar to trial before the Senate on the same charges.¹⁰⁶ The House's and Senate's actions thus suggest that they did not consider trial before the Senate an instance of jeopardy within the meaning of the Double Jeopardy Clause.

     It may be argued that, although trial before the Senate is an instance of jeopardy, the Impeachment Judgment Clause permits such trial following criminal conviction. That Clause expressly allows for criminal trial after conviction by the Senate. So, one might argue, it permits the reverse sequence as well: trial before the Senate following criminal conviction. By similar logic, if the Impeachment Judgment Clause bars prosecution following Senate acquittal, it should bar trial in the Senate following acquittal in the courts. In carrying out the impeachment trial of Judge Alcee Hastings, however, the Senate rejected this view that the relationship between criminal prosecution and impeachment trials could turn on whether the prior judgment was a conviction or an acquittal.

     Following a jury trial, Judge Alcee Hastings was acquitted in 1983 of conspiring to take a bribe and of obstructing justice. In 1988, pursuant to 28 U.S.C. § 372(c), the 11^th Circuit Judicial Council certified to the Judicial Conference of the United States that Hastings had engaged in conduct that might constitute an impeachable offense and the Judicial Conference made a similar certification to the House of Representatives. The House impeached Hastings in 17 articles, the first of which was in substance the bribery charge upon which he had been acquitted and 14 of which alleged that he had repeatedly lied under oath at his criminal trial. In 1989, the Senate tried and convicted Hastings on the first article and eight of the ones charging lying at his criminal trial.

     The Investigating Committee of the Eleventh Circuit Judicial Council, which consisted of three circuit judges and two district court judges, briefly considered whether Judge Hastings' acquittal should bar his impeachment by the House and trial by the Senate (and thus the Committee's making of a recommendation of impeachment). They concluded that it should not for three principal reasons.¹⁰⁷ First, the Committee thought it obvious that a conviction in the courts would not bar impeachment and legislative trial, and they could see no distinction between convictions and acquittals in this respect. Second, they reasoned that "impeachment does not serve the same purpose as a criminal prosecution. Impeachment is remedial and designed to protect the institution of government from corrupt conduct . . . ."¹⁰⁸ Third, they noted that the standard of proof was higher in a criminal prosecution than in an impeachment trial.¹⁰⁹

     The House Judiciary Committee also found no double jeopardy bar. The Committee took the view that "impeachment is not a criminal proceeding" because the possible sanctions upon conviction are "remedial or prophylactic, rather than criminal or punitive."¹¹⁰ The House adopted the articles by a vote of 413-3.¹¹¹

     Just before the Senate took up the House's charges, Judge Hastings brought suit against the Senate and some of its officers seeking to enjoin his impeachment trial on double jeopardy grounds, among others. District Judge Gerhard Gesell rejected Hastings' double jeopardy contention and dismissed the action. Judge Gesell reasoned as follows:

     Impeachment trials are sui generis: in several instances in the Constitution, impeachment is distinguished from criminal proceedings. The accused has no right to a jury, and the President may not pardon a person convicted by impeachment. The Framers understood that impeachment trials were fundamentally political, which seems to indicate that impartiality--however much it has been present and is to be desired--is not guaranteed. It is clear that the federal rules of evidence do not apply in impeachment trials, and the Constitution itself does not require unanimity among the Senators sitting in judgment. Senators determine their own burdens of proof: they need not be persuaded beyond a reasonable doubt that the defendant committed each and every element of every Article. Deviating from English precedent, the Framers sharply limited the remedies or punishment available upon conviction to disqualification and removal from office . . . .

Hastings v. United States Senate, 716 F. Supp. 38, 41 (D.D.C. 1989). Judge Gesell read the Impeachment Judgment Clause as "acknowledg[ing] separate and different roles for the executive's power of prosecution and the legislature's impeachment powers. It is unthinkable that the executive branch could effectively prevent an impeachment by purporting to try a judge or that the judiciary could prevent an impeachment by accepting a plea. Rather, the executive and legislative branches have different roles to play if a judge engages in criminal behavior." Id. at 42. The court of appeals affirmed on non-justiciability grounds rather than reaching the merits of any of Judge Hastings' contentions. Hastings v. United States Senate, 887 F.2d 332 (D.C. Cir. 1989) (unpublished).

     Judge Hastings renewed his double jeopardy argument before the Senate in a motion to dismiss.¹¹² He made the expressio unius argument based on the Impeachment Judgment Clause, urging that the Clause "creates an express exception for a 'party convicted' of an impeachable offense[, but] no exception for a party acquitted."¹¹³ He pointed out that Madison's proposed double jeopardy clause had included an exception for impeachments, which had been deleted by the Senate. He noted the constitutional provisions suggesting that an impeachment trial is a criminal proceeding, and he argued that the Mendoza-Martinez factors pointed in the direction of treating impeachment trials as criminal proceedings. Finally, he argued that the "core policies" promoted by the double jeopardy rule favored prohibiting Senate trials following acquittal in the courts. As the Supreme Court stated in Green v. United States, 355 U.S. 184, 187-88 (1957), "[t]he underlying idea, one that is deeply ingrained in at least the Anglo-American system of jurisprudence, is that the State with all its resources and power should not be allowed to make repeated attempts to convict an individual for an alleged offense, thereby subjecting him to embarrassment, expense and ordeal and compelling him to live in a continuing state of anxiety and insecurity, as well as enhancing the possibility that even though innocent he may be found guilty."

     The Senate denied Hastings' motion by a vote of 92-1.¹¹⁴ In statements inserted into the record following the final vote to convict, several Senators addressed the double jeopardy issue. They explained their judgment that trial by the Senate was not a criminal proceeding and that it therefore did not constitute an instance of jeopardy within the meaning of the Double Jeopardy Clause.¹¹⁵

     The Hastings impeachment trial provides additional support for the notion that an impeachment trial is not a jeopardy within the meaning of the Double Jeopardy Clause and that an acquittal by the Senate should not block a criminal prosecution for the same offenses.

     We recognize that several arguments might be made to limit the significance of the Hastings case (and of the Claiborne and Nixon cases) for the question we are addressing, but we find none of them convincing.

     First, one might argue that trial in the Senate following acquittal in the courts (as in the Hastings case) is different from trial in the courts following acquittal in the Senate (the situation we are considering) because of the different standards of proof required in the two proceedings. A jury verdict of not guilty means the prosecution has failed to prove beyond a reasonable doubt that the defendant is guilty. The Senate might conclude that such a verdict presented no obstacle to their trial of the defendant on the same charges because, quite consistently with the jury's verdict, they might conclude that the House managers had shown, under some lower standard of proof (whether preponderance or clear and convincing), that the defendant had committed the charged acts. The reverse sequence would still be impermissible because a verdict of not guilty in the Senate under the lower standard of proof would be inconsistent with a finding of guilty under the more demanding beyond-a-reasonable doubt standard required in court.

     We find this explanation of the significance of the Hastings case unconvincing for two reasons. First, the argument concerns collateral estoppel - the principle that an issue finally resolved in one proceeding as between two parties may not be re-examined in a subsequent proceeding - not double jeopardy. It is true that the resolution of a factual issue in favor of a defendant under the beyond-a-reasonable-doubt standard is no bar to consideration of the same issue under a more lenient standard of proof. Thus, for example, collateral estoppel is no bar to a civil proceeding alleging that a defendant committed certain acts following acquittal of the same defendant on criminal charges requiring proof of the same acts. See, e.g., Helvering v. Mitchell, 303 U.S. 391, 397-398 (1938). But, as the Supreme Court has repeatedly explained, even where collateral estoppel creates no obstacle to a successive trial, double jeopardy still may. That is because it is the risk of criminal punishment, regardless of the form of the proceeding or the standard of proof, that determines whether the two proceedings constitute impermissible successive jeopardies. See, e.g., id. at 398-405. Thus, for the Senate to try Judge Hastings after his criminal trial, it would not have been enough for the members of that body to have concluded that the reduced standard of proof removed any collateral estoppel problem. They would also have to have concluded that trial before the Senate was not an instance of jeopardy.

     Second, we find little, if any, evidence in the record of the Senate trial of Judge Hastings suggesting that the Senators relied on this argument. Judge Hastings presented the double jeopardy issue squarely to the Senate, which considered it both in a motion to dismiss and in its final judgment.¹¹⁶ Of the fifteen Senators who inserted statements in the record explaining their final votes, several addressed the double jeopardy question, but none did so in terms of the difference in the standards of proof.¹¹⁷

     One might also try to cabin the significance of the Hastings case by contending that the Senate's decision to try Judge Hastings turned on the judicial character of his office and that the decision therefore does not serve as a precedent for the treatment of executive branch officials. The argument would go as follows. The "good behavior" standard governing judicial tenure imposes standards of propriety, and of the appearance of propriety, on federal judges that do not apply to executive officials. Because of these particularly rigorous standards of behavior, conduct short of the criminal may nonetheless be outside the bounds of judicial good behavior. Thus acquittal of serious crimes might still leave a judge open to fair condemnation as having deviated from the path of good behavior and thus as meriting removal from office.¹¹⁸

     Even if this argument for the significance of the good behavior standard were correct as a theoretical possibility, the records of the Hastings impeachment proceedings offer little, if any, evidence suggesting that the standard influenced the Senate's resolution of the double jeopardy issue it confronted.¹¹⁹

     Finally, regarding the Claiborne and Nixon cases, one might argue that the Senate's decision to proceed rested not on a judgment that Senate trial did not constitute an instance of jeopardy but on a decision that the need to remove federal judges who had been convicted of felonies was so imperative that it outweighed otherwise applicable double jeopardy principles. After all, federal judges, unlike federal executive officials, have life tenure, so impeachment provides the only political mechanism to remove them from office. If the Senate had proceeded on that basis, we would expect to find some discussion of the dilemma involved. We are aware of none in the record of those proceedings.

     The three judicial impeachment trials of the 1980s support the conclusion that the Senate does not view impeachment trials as instances of jeopardy within the meaning of the Double Jeopardy Clause. The Hastings case, moreover, demonstrates that the Senate sees no difference between prior acquittal and prior conviction in this regard.

     IV. Conclusion

     We conclude that the Constitution permits a former President to be criminally prosecuted for the same offenses for which he was impeached by the House and acquitted by the Senate while in office.

     As the length of this memorandum indicates, we think the question is more complicated than it might first appear. In particular, we think that there is a reasonable argument that the Impeachment Judgment Clause should be read to bar prosecutions following acquittal by the Senate and that disqualification from federal office upon conviction by the Senate bears some of the markers of criminal punishment. Nonetheless, we think our conclusion accords with the text of the Constitution, reflects the founders' understanding of the new process of impeachment they were creating, fits the Senate's understanding of its role as the impeachment tribunal, and makes for a sensible and fair system of responding to the misdeeds of federal officials.

FOOTNOTES:

1. In the context of successive trials in the courts, double jeopardy claims often raise the preliminary question whether the offenses charged in the second proceeding are the same as those that formed the basis for the first proceeding. See, e.g., United States v. Dixon, 509 U.S. 688, 696 (1993); Blockburger v. United States, 284 U.S. 299 (1932). We understand the question posed to assume that this issue has been resolved, and thus we express no view on how the issue might arise or be resolved in the circumstance of criminal prosecution following an impeachment trial.

2. The brief does not explain why the House's failure to impeach would, on any reading of the Impeachment Judgment Clause, act as a bar. Even if one took the view that the Impeachment Judgment Clause's reference to "the party convicted" implied that acquitted parties could not be criminally prosecuted, that implication would naturally extend only to individuals who had been impeached by the House and acquitted by the Senate. (In regular criminal proceedings, jeopardy does not attach until the jury has been sworn, see, e.g., Crist v. Bretz, 437 U.S. 28, 35-38 (1978), or, in a bench trial, the first witness has taken the stand, see, e.g., id. at 37 n.15. At the time of the drafting of the Constitution, the common law rule was that jeopardy did not attach until the jury had rendered a verdict. See, e.g., 2 William Hawkins, A Treatise of the Pleas of the Crown 527 (6th Ed. 1787).) The brief appears to treat an impeachment investigation and a rejection of articles of impeachment by the House as a type of acquittal. We are unaware of any commentator or Member of Congress who has adopted this position.

3. The citation is 14 Annals of Congress 432 (1805). Martin had been a delegate from Maryland at the Constitutional Convention. The memorandum cited a portion of Martin's speech at the Chase trial for the proposition that "Article 1, section 3, clause 7 was designed to overcome a claim of double jeopardy rather than to require that impeachment must precede any criminal proceedings." 1973 OLC Memo at 3. In support of his larger argument that impeachable offenses were limited to indictable offenses, Martin imputed to the House managers the view that "a judge is only removable from office on account of crimes committed by him as a judge, and not for those for which he would be punishable as a private individual." 14 Annals of Cong. 431 (1805). If that were true, Martin argued, a judge might be convicted and punished in the courts for burglary or receiving stolen goods and "yet he could not be removed from office, because the offence was not committed by him in his judicial capacity, and because he could not be punished twice for the same offence." Id. That implication, Martin explained, must be wrong:

     The truth is, the framers of the Constitution, for many reasons, which influenced them, did not think proper to place the officers of the Government in the power of the two branches of the Legislature, further than the tenure of their office. Nor did they choose to permit the tenure of their offices to depend upon passions or prejudices of jurors. The very clause in the Constitution, of itself, shows that it was intended the persons impeached and removed from office might still be indicted and punished for the same offence, else the provision would have been not only nugatory, but a reflection on the enlightened body who framed the Constitution; since no person ever could have dreamed that a conviction on impeachment and a removal from office, in consequence, for one offence, could prevent the same person from being indicted and punished for another and different offence.

Id. at 432.

4. We discuss the Story passage infra pp. 19-20 & n.44.

5. See Joseph Isenberg, Impeachment and Presidential Immunity from Judicial Process, 18 Yale L. & Pol'y Rev. 53, 92-93 (1999); Jay S. Bybee, Who Executes the Executioner? Impeachment, Indictment and Other Alternatives to Assassination, 2 Nexus: J. Opin. 53, 58-59, 63 (1997).

6. See Impeachment of Judge Alcee L. Hastings: Motions of Judge Alcee L. Hastings to Dismiss Articles I-XV and XVII of the Articles of Impeachment Against Him and Supporting and Opposing Memoranda, S. Doc. No. 101-4, at 48-57 (1989) ("Hastings Motions to Dismiss").

7. See Conn. Const. art. 9, § 3; Del. Const. art. 6, § 2; Haw. Const. art. III, § 19; Ky. Const. § 68; Mass. Const. ch. I, § 2, art. 8; Mich. Const. art. 11, § 7, par. 4; Minn. Const. art. 8, § 2; Miss. Const. § 51; N.H. Const. art. 39; N.J. Const. art. 7, § 3, par. 3; R.I. Const. art. XI, § 3; Tex. Const. art. 15, § 4; Vt. Const. § 58; Va. Const. art IV, § 17; W.Va. Const. art. IV, § 9.

8. See Ala. Const. art. 7, § 176; Alaska Const. art. 2, § 20; Ariz. Const. art. 7, pt. 2, § 2; Ark. Const. art. 15, § 1; Cal. Const. art. IV, § 18; Colo. Const. art. XIII, § 2; Fla. Const. art. III, § 17; Ga. Const. art. 3, § 7, par. 3; Idaho Const. art. V, § 3; Ill. Const. art. IV, § 14; Iowa Const. art. III, § 20; La. Const. art. X, § 24; Me. Const. art. III, § 7; Mo. Const. art. VII, § 3; Mont. Const. art. V, § 13; Nev. Const. art. 7, § 2; N.M. Const. art. IV, § 36; N.Y. Const. art. VI, § 24; N.C. Const. art. IV, § 4; N.D. Const. art. XI, § 10; Okl. Const. art. VIII, § 5; Penn. Const. art. VI, § 6; S.C. Const. art. XV, § 3; S.D. Const. art. XVI, § 3; Tenn. Const. art. V, § 4; Utah Const. art. VI, § 19; Wash. Const. art. V, § 2; W.Va. Const. art. IV, § 9; Wisc. Const. art. VII, § 1; Wyo. Const. § 18.

     We have found references to the difference between the wording of the federal clause and that of many of the state constitutions in only two judicial decisions, one of which relies upon the other: State ex. rel. Christian v. Rudd, 302 So.2d 821, 825 (Fla Dist. Ct. App. 1974), vacated in part on other grounds, Rudd v. State ex. rel. Christian, 310 So.2d 295 (Fla.1975); In re Investigation by Dauphin County Grand Jury, 2 A.2d 804, 808 (Pa. 1938). In the Pennsylvania case, a district attorney began a grand jury investigation of several state officials, the state House of Representatives initiated an impeachment investigation of the same officials, and the House investigating committee then sought a writ of prohi