Opinion for the Court filed by Circuit Judge STEPHEN F. WILLIAMS.
Concurring opinion filed by Circuit Judge RANDOLPH.
Opinion dissenting in part and concurring in the judgment filed by Circuit Judge HARRY T. EDWARDS.
STEPHEN F. WILLIAMS, Circuit Judge:Walter L. Nixon, Jr., formerly the Chief Judge of the U.S. District Court for the Southern District of Mississippi, was impeached by the House of Representatives and convicted by the Senate for giving false testimony to a grand jury investigating allegations that he had been bribed. Nixon seeks judicial review of the Senate’s procedures — in particular, its use of a committee to take testimony and gather other evidence.
The Constitutional Convention, however, gave the Senate “the sole Power to try all Impeachments”, Art. I, § 3, cl. 6 (emphasis added). It not only rejected proposals to assign the power to the federal courts, but it did so for reasons that are almost impossible to square with any judicial role in the process. We find Nixon’s claim nonjusticiable.
After an investigation into reports that Nixon had asked a local district attorney to stop the prosecution of a man whose father had enriched Nixon through an investment scheme, a grand jury indicted Nixon on one count of receiving an illegal gratuity and three counts of perjury before the grand jury. At trial, Nixon was convicted on two counts of perjury and acquitted on the other two counts. He was sentenced to prison, and his conviction was affirmed on appeal. See United States v. Nixon, 816 F.2d 1022 (5th Cir.1987); see also United States v. Nixon, 881 F.2d 1305 (5th Cir.1989) (affirming the denial of Nixon’s motion for a new trial).
Even after this conviction, Walter Nixon refused to resign from his office as a United States district judge, and while serving time in prison he continued to draw his judicial salary. See H.R.Rep. No. 36, 101st Cong., 1st Sess. 13 (1989). The House of Representatives began impeachment proceedings, see id., and on May 10, 1989, it voted to impeach Nixon on three articles charging him with giving false testimony to the grand jury and bringing disrepute on the federal judiciary. See 135 Cong.Rec. H1811 (daily ed. May 10, 1989).
When these articles of impeachment were presented to the Senate, it invoked its own Impeachment Rule XI, under which the presiding officer appoints a committee of twelve senators “to receive evidence and take testimony”. S.Imp.R. XI, reprinted in Senate Manual, S.Doc. No. 1, 101st Cong., 1st Sess. 186 (1989); see S.Res. 128, 101st Cong., 1st Sess.,-135 Cong.Rec. S5199 (daily ed. May 11, 1989). The committee conducted four days of hearings, taking live testimony from ten witnesses, including Nixon himself. See S.Rep. No. 164, 101st Cong., 1st Sess. 4 (1989). It then transmitted to the full Senate a complete record of the evidence and a report, summarizing both the undisputed and disputed facts of the case without resolving contested issues or recommending any particular disposition of the charges. See id. at 3-4. *241After considering final briefs, hearing arguments on the Senate floor from both the impeachment managers and the defense, including a personal appeal from Nixon himself, and posing questions to the parties, see 135 Cong.Rec. S14,493-517 (daily ed. Nov. 1, 1989), the Senate voted by more than the constitutionally prescribed two-thirds majority to convict Nixon on two of the three articles. 135 Cong.Rec. S14,635 (daily ed. Nov. 3, 1989); see Art. I, § 3, cl. 6. The presiding officer entered judgment removing him from his office as a United States district judge. 135 Cong.Rec. at S14,636.
Nixon then sued in district court, arguing that the Senate’s failure to give him a full evidentiary hearing before the entire Senate violated its constitutional duty to “try” all impeachments. See Art. I, § 3, cl. 6. He sought a declaratory judgment that his conviction by the Senate was void and that his judicial salary and privileges should be reinstated from the date of his conviction. The district court held that his claim was nonjusticiable, see Nixon v. United States, 744 F.Supp. 9 (D.D.C.1990), and we agree.
“The House ... shall have the sole Power of Impeachment”, Art. I, § 2, cl. 5, and “The Senate shall have the sole Power to try all Impeachments”, Art. I, § 3, cl. 6. Nowhere else does the Constitution explicitly confer on a body the “sole” power to do anything. The only court to fully consider the issue before this case gave “sole” its full weight. It read the word to express an “intention that no other tribunal should have any jurisdiction of the cases tried under the provisions with reference to impeachment.” Ritter v. United States, 84 Ct.Cl. 293, 296 (1936). The court went on:
The dictionary definition of the word “sole” is “being or acting without another” and we think it was intended that the Senate should act without any other tribunal having anything to do with the case. This would be the ordinary signification of the words and this construction is supported by a consideration of the proceedings of the Constitutional Convention and the uniform opinion of the authorities which have considered this matter.
Id. Indeed, the unanimous rejection of judicial review to which the court refers seems not to have been breached until Raoul Berger 20 years ago used a rather casual reading of Powell v. McCormack, 395 U.S. 486, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969), to claim the availability of judicial review. See Raoul Berger, Impeachment: The Constitutional Problems (1973); Staff of Senate Comm, on Rules and Administration, 93d Cong., 2d Sess., Impeachment: Miscellaneous Documents 170-71 (1974) (“Committee Print”) (memorandum on judicial review of impeachment proceedings by Stephen F. Goldstein).
The history of the Constitution’s impeachment provisions bears out Ritters understanding. Both of the broad proposals that provided the foundation for the Convention delegates’ debates, Randolph’s “Virginia Plan” and Paterson’s “New Jersey Plan”, gave the power to “hear and determine” impeachments to the federal judiciary. See 1 The Records of the Federal Convention of 1787, at 21-22 (Max Farrand ed. 1966) (Virginia Plan); id. at 244 (New Jersey Plan); see also P. Hoffer & N. Hull, Impeachment in America, 1635-1805, at 97-100 (1984). Madison supported this assignment, specifically favoring the Supreme Court, see 2 Farrand at 551, while Hamilton proposed a special court composed of the chief judge from each state’s supreme court, see 1 Farrand at 292-93. However, once the Convention decided that a college of electors, rather than the Senate, should name the President, thereby eliminating a potential conflict between the Senate’s roles as both selector and remover of the President, it authorized the Senate to conduct impeachment trials and to render final judgments by two-thirds vote. Hoffer & Hull at 98-99; 2 Farrand at 500-01, 552-53.
In the surviving scraps of Convention debate on the issue, the focus was on presidential impeachment. While both Madison and Pinckney opposed use of the Senate, as tending to increase executive dependence *242on the legislative branch, Gouverneur Morris responded that the Supreme Court was “too few in number and might be warped or corrupted”, and Roger Sherman suggested that the Court was an improper forum to try the President “because the Judges would be appointed by him.” 2 Farrand at 551. These themes of conflict of interest — so typical of the framers’ concern over checks and balances — persisted into later discussions of applying the impeachment power to judges.
In The Federalist, Hamilton identified the impeachment power as the basis for constraining usurpation by judges. Thus, in Federalist No. 79 he wrote:
The precautions for their responsibility are comprised in the article respecting impeachments. They are liable to be impeached for mal-conduct by the house of representatives, and tried by the senate, and if convicted, may be dismissed from office and disqualified for holding any other. This is the only provision on the point, which is consistent with the necessary independence of the judicial character, and is the only one which we find in our own constitution in respect to our own judges.
The Federalist 532-33 (Jacob E. Cooke ed. 1961). A little later, though generally disparaging the risk of judicial aggrandizement as a mere “phantom”, he went on again to identify impeachment as “the important constitutional check” and to justify the assignment to the Senate as a key assurance of the remedy’s adequacy:
And the inference [that usurpations were improbable] is greatly fortified by the consideration of the important constitutional check, which the power of instituting impeachments, in one part of the legislative body, and of determining upon them in the other, would give to that body upon the members of the judicial department. This is alone a complete security. There never can be danger that the judges, by a series of deliberate usurpations on the authority of the legislature, would hazard the united resentment of the body entrusted with it, while this body was possessed of the means of punishing their presumption by degrading them from their stations. While this ought to remove all apprehensions on the subject, it affords at the same time a cogent argument for constituting the senate a court for the trial of impeachments.
Federalist No. 81, Cooke ed. at 545-46. Hamilton’s emphatic language would have fallen rather flat if candor had compelled him to add that, of course, the judges themselves would sit in final judgment over this check on their excesses.
The framers invoked an additional kind of conflict or bias to support use of the Senate rather than the judiciary for impeachment trials — the bias caused by a person’s having played a role in a prior phase of an extended process. Article I specifies that a person convicted in impeachment proceedings “shall nevertheless be liable and subject to Indictment, Trial, Judgment and Punishment, according to Law.” Art. I, § 3, cl. 7. Thus the Constitution explicitly anticipates two sets of proceedings for at least some officials who commit impeachable offenses — one in Congress and one in the courts. The Framers separated the two, trying to secure for the accused the benefit of independent judgments. Though Hamilton assumed (in line with other delegates’ comments) that impeachment trials would precede criminal trials, his insistence on the need for distinct, independent forums is no less compelling when the sequence is reversed, as was true of Nixon:
Would it be proper that the persons, who had disposed of his fame and his most valuable rights as a citizen in one trial, should in another trial, for the same of-fence, be also the disposers of his life and his fortune? Would there not be the greatest reason to apprehend, that error in the first sentence would be the parent of error in the second sentence? That the strong bias of one decision would be apt to overrule the influence of any new lights, which might be brought to vary the complexion of another decision?
Federalist No. 65, Cooke ed. at 442.
The risks from overlapping powers reach their apogee in a presidential impeachment *243trial, for which the Chief Justice presides over the Senate. Art. I, § 3, cl. 6. Whether the Chief Justice has clashed with the Senate over trial procedures, as did Chief Justice Chase during the trial of President Andrew Johnson, see Berger at 268-69; Committee Print at 6-8, or concurred all the way, he would be seen as prejudiced in review of the impeachment trial. That no one recognized this conflict in the framing and ratification debates argues the implausibility of such review.
That the Convention intended the impeachment power to be qualified only by political forces is also reflected in constitutional language limiting the executive’s authority. Just as Hamilton viewed impeachment as “the important constitutional check” on the judiciary (Federalist No. 81), he called the power “an essential check in the hands of [the legislative body] upon the encroachments of the executive”. Federalist No. 66, Cooke ed. at 446; see also Federalist No. 66, Cooke ed. at 441 (describing impeachments as “a bridle in the hands of the legislative body upon the executive servants of the government”). The delegates made sure that the executive would have no power to undermine the check when they expressly excepted impeachments from the President’s pardon power. Art. II, § 2, cl. 1. While the absence of any such express limitation on judicial interference might support an “ex-pressio unius” argument in favor of judicial review, a more plausible reading is that the framers simply assumed that courts had nothing whatever to do with impeachments.
If the Constitution’s text, backed by the historical evidence, prevents both the judiciary and the executive from constraining the legislative power of impeachment, did the framers just slip up, leaving an unchecked check? The answer can be found in Article I itself, which provides two safeguards within the legislative branch to control unwarranted use of impeachments: (1) the separation of impeachment powers between the House and the Senate, see Art. I, § 2, cl. 5; § 3, cl. 6, and (2) the requirement of a two-thirds vote in the Senate to convict, see Art. I, § 3, cl. 6. Hamilton’s response to fears of abuse, including concerns about mixing judicial and legislative power in one body, tracked the two limits found in Article I. First he argued that the division of the impeachment powers “guards against the danger of persecution from the prevalency of a factious spirit in either of [the two houses].” Federalist No. 66, Cooke ed. at 446. Then he concluded, “As the concurrence of two-thirds of the senate will be requisite to a condemnation, the security to innocence, from this additional circumstance, will be as complete as itself can desire.” Id.
For Hamilton, that was enough. He made no reference to judicial review as a check of the power, in sharp contrast with his discussion of ordinary legislative powers. Compare Federalist No. 78, Cooke ed. at 524-25 (judicial review as a check on unconstitutional bills of attainder, ex post facto laws, and statutes). In fact, the parties have not identified a single statement in either the framers’ or ratifiers’ debates alluding even to the possibility of judicial review, and Berger, its ardent proponent, acknowledges the absence of any such mention at the state conventions. See Berger at 116. To check the impeachment power, the framers quite naturally relied on the political accountability of members of Congress. Thus judges, who on so many issues have the last word, must rely on the public as the ultimate check on impeachment, itself the Constitution’s explicit check on their own excesses.
The broad scope of the Senate’s power is further supported by the grant to each house of the power to “determine the Rules of its Proceedings”. Art. I, § 5, cl. 2. This clause gives the Senate independent discretion to set procedural rules for impeachment trials, including the rule challenged by Nixon — Senate Impeachment Rule XI. The Supreme Court hinted, in a case where it refused to second-guess the House’s rule on establishing the presence of a quorum, that congressional rules of procedure may be judicially reviewable in some circumstances if they “ignore constitutional restraints”, see United States v. Ballin, 144 U.S. 1, 5, 12 S.Ct. 507, 509, 36 L.Ed. 321 *244(1892), but this court has refused (on prudential grounds) to review the House of Representative’s rules allocating seats on committees, despite a claim that those rules violated the Fifth Amendment rights of members of the minority party in the House. Vander Jagt v. O’Neill, 699 F.2d 1166, 1173 (D.C.Cir.1983). The rules clause provides at least indirect support for the view that the Senate’s “sole Power to try all Impeachments” includes the sole power to frame the rules it will follow in conducting such trials.
Constitutional exclusion of judicial review of impeachments would seem to be the end of the matter. But courts have long analyzed the justiciability of issues constitutionally committed to the other branches as part of the “political question” doctrine. See Marbury v. Madison, 5 U.S. (1 Cranch) 137, 164, 2 L.Ed. 60 (1803). If the Constitution makes a “textually demonstrable commitment” of any issue to “a coordinate political department”, see Baker v. Carr, 369 U.S. 186, 217, 82 S.Ct. 691, 710, 7 L.Ed.2d 663 (1962), it so commits the conduct of impeachment trials to the Senate. It remains to consider whether Powell v. McCormack, 395 U.S. 486, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969), and later cases, have so shrunk the political question doctrine as to permit courts to set the boundaries of permissible impeachment trial procedure despite such a textual commitment. A careful reading of Powell, which explicitly preserved the political question doctrine, see id. at 518, 89 S.Ct. at 1962 (“It is well established that the federal courts will not adjudicate political questions”), suggests the contrary.
Powell concerned the interplay between three constitutional clauses: “Each House shall be the Judge of the Elections, Returns and Qualifications of its own Members” (Art. I, § 5, cl. 1); the qualifications for Representatives specified in Art. I, § 2, cl. 2; and the expulsion clause (Art. I, § 5, cl. 2). The Court embarked on its constitutional analysis of justiciability with the point that “[i]n order to determine whether there has been a textual commitment to a co-ordinate department of the Government, we must interpret the Constitution.” 395 U.S. at 519, 89 S.Ct. at 1963. The Court then examined the text and history of Art. I, § 5 to determine whether the framers limited the scope of the textual commitment found in that clause by other constitutional provisions. It found such limits in Art. I, § 2’s narrow list of qualifications (age, period of citizenship of the United States, and residence when elected). See id. at 520-21, 89 S.Ct. at 1963-64. The Court found its reading of the qualifications and exclusion clauses confirmed by the expulsion clause’s requirement of a two-thirds vote, which would be wholly undercut by giving an unrestricted meaning to the exclusion clause. Id. at 547-48, 89 S.Ct. at 1977-78.
The Court’s political question analysis in Powell thus relies heavily on the ultimate conclusion that it was the framers’ intention “to deny either branch of Congress the authority to add to or otherwise vary the membership qualifications expressly set forth in the Constitution.” Id. at 532, 89 S.Ct. at 1969-70 (emphasis added); see also id. at 540, 89 S.Ct. at 1973-74 (“The debates at the state conventions also demonstrate the Framers’ understanding that the qualifications for members of Congress had been fixed in the Constitution.”) (emphasis added). The Court emphasized as a key piece of evidence this quotation from Federalist No. 60: “The qualifications of the persons who may choose or be chosen, as has been remarked upon another occasion, are defined and fixed in the constitution; and are unalterable by the legislature.” Cooke ed. at 409 (emphasis added), quoted in Powell, 395 U.S. at 539, 89 S.Ct. at 1973. The Federalist yields no parallel suggestion that the minimum procedures for conducting an impeachment trial are unalterable by the legislature and thus not textually committed. Thus application of Powell’s method — an analysis of the relevant constitutional text and history — leads here to a conclusion of nonjusticiability.
Later cases confirm this understanding of Powell. In Roudebush v. Hartke, 405 U.S. 15, 92 S.Ct. 804, 31 L.Ed.2d 1 (1972), the Supreme Court decided that a state’s recount of ballots in a senatorial election *245did not infringe on the Senate’s power to “be the Judge of the Elections ... of its own members”, Art. I, § 5, cl. 1, as the recount could “usurp” the Senate’s function “only if it frustrate[d] the Senate’s ability to make an independent final judgment”, 405 U.S. at 25, 92 S.Ct. at 810-11, which it could not do, id. at 25-26, 92 S.Ct. at 810-11. While the wording of the holding itself carries an implication of nonjusti-ciability, the Court was more explicit in its rejection of a claim that the candidate’s action had been mooted by the Senate’s decision to seat one of the candidates, subject to the outcome of the case before the Court. It stated that it was “without power to alter the Senate’s judgment”, id. at 18-19, 92 S.Ct. at 807-08,1 and that “[wjhich candidate is entitled to be seated in the Senate is, to be sure, a nonjusticiable political question”, id. at 19, 92 S.Ct. at 807-08 (citing Powell). The Court decided that the case was not moot because the Senate had seated the other candidate only temporarily until the conclusion of the suit, after which it would “be free to make an unconditional and final judgment under Art. 1, § 5.” Id. (emphasis added). This strong language indicates that the Court would find nonjusticiable any challenge to the Senate’s “final” judgment in a disputed election, even a procedural challenge that alleged, for example, that the senators refused to hear critical witnesses on the validity of disputed ballots.
Since Powell and Roudebush, this court has refused to entertain objections not only to the substance but also to the procedures used by the House of Representatives in the exercise of its ballot-counting authority under Art. I, § 5, cl. 1 (each house “shall be the Judge of the Elections, Returns and Qualifications of its own Members”). In Morgan v. United States, 801 F.2d 445 (D.C.Cir.1986), a unanimous panel examined the text and history of the clause, noting especially that although it was attacked in the ratification debates, none of its defenders “mention[ed] the safeguard of judicial review. Such a safeguard was evidently unthinkable, since the determination of the legislative House was itself deemed to be a judicial one.” 801 F.2d at 447 (emphasis in original). The point applies with equal force here. We concluded that “[t]he exclusion of others — and in particular of others who are judges — could not be more evident.” Id.; see also McIntyre v. Fallahay, 766 F.2d 1078, 1081 (7th Cir.1985) (“The House is not only ‘Judge’ but also final arbiter. Its decisions about which ballots count, and who won, are not reviewable in any court....”).
In Morgan we quite literally underscored that the elections clause makes each House “the Judge” of elections. 801 F.2d at 447 (emphasis in original opinion). If language making each house “the judge” of elections gives that house exclusive discretion to determine the procedures for making that judgment, it seems a fortiori that a clause granting the Senate “the sole Power to try all Impeachments” gives it sole discretion to choose its procedures. If the clause made the Senate “the sole trier of impeachments” the structure would be more parallel but the grant of exclusive authority no clearer.
Although the primary reason for invoking the political question doctrine in our case is the textual commitment of impeachment trials to the Senate, the need for finality also demands it. See Baker v. Carr, 369 U.S. at 210, 82 S.Ct. at 706. In the elections clause context, the Morgan court emphasized the need for “quick, decisive resolution of election controversies.” Morgan, 801 F.2d at 450. The need for finality in impeachments is even more acute. If claims such as Nixon’s were jus-ticiable, procedural appeals from every impeachment trial would become routine, as the Court of Claims observed even in the less litigious era of the Ritter decision. See 84 Ct.Cl. at 299. For the impeach-*246merits that are anything but routine, those of presidents and chief justices, the intrusion of the courts would expose the political life of the country to months, or perhaps years, of chaos. Even if the courts qualified a finding of justiciability with a rule against stays or specific relief of any kind, their review would undermine the new President’s or Chief Justice’s legitimacy for at least as long as the process took. And a declaratory action without final relief awarding the office to one person or the other could confound matters indefinitely.
If the political question doctrine has no force where the Constitution has explicitly committed a power to a coordinate branch and where the need for finality is extreme, then it is surely dead. But although the Supreme Court has rarely applied the doctrine in recent years, see Dissent at 258 & n. 11, it has also declined the several opportunities available to dispatch it. See Gilligan v. Morgan, 413 U.S. 1, 93 S.Ct. 2440, 37 L.Ed.2d 407 (1973); Roudebush v. Hartke, 405 U.S. 15, 19, 92 S.Ct. 804, 807-08, 31 L.Ed.2d 1 (1972); Powell, 395 U.S. at 518, 89 S.Ct. at 1962. We honor the doctrine and apply it here.
It does not help establish justiciability to pose hypotheticals of outrageous behavior by a coordinate branch, such as that the Senate might turn its impeachment trial responsibilities over “to a randomly chosen group of schoolchildren”, Dissent at 259, or even pass a rule “allowing conviction and removal of impeached officers by a majority vote”, id. at 256. If the Senate should ever be ready to abdicate its responsibilities to schoolchildren, or, moved by Caligula’s appointment of his horse as senator, to an elephant from the National Zoo, the republic will have sunk to depths from which no court could rescue it. And if the senators try to ignore the clear requirement of a two-thirds vote for conviction, they will have to contend with public outrage that will ultimately impose its sanction at the ballot box.2 Absent judicial review, the Senate takes sole responsibility for its impeachment procedures as a full-fledged constitutional actor, just as the framers intended.
It would be peculiarly ironic for the judiciary to take charge of defining the limits of permissible procedure out of concern over the Senate’s possible excesses. The exercise of any final power is by definition open to monstrous hypothetical abuse. But judges exercise such power daily, unre-viewably imposing procedural and substantive boundaries on almost every decision of the political branches. In all this we are free of political constraints, subject to correction solely by constitutional amendment and to sanction solely by impeachment. If the impeachment claims of a fellow judge were justiciable, the circle would be closed — the judiciary would have final, un-reviewable power over the one procedure established to restrain excesses in all its other final and unreviewable powers: checkmate.
Today we refuse to embark on setting limits for the procedures the Senate may choose for the trial of impeachments; the Constitution excludes us. Walter Nixon’s claim is not justiciable.3
Affirmed.
. The Court appended the following footnote:
6. See Reed v. County Comm’rs, 277 U.S. 376, 388, 48 S.Ct. 531, 532, 72 L.Ed. 924 (1928): "[The Senate] is the judge of the elections, returns and qualifications of its members. Art. I, § 5. It is fully empowered and may determine such matters without the aid of the House of Representatives or the Executive or Judicial Department."
Id. at 19 n. 6, 92 S.Ct. at 807 n. 6.
. Because the two-thirds vote requirement of Art. I, § 3, cl. 6 is so concrete, the argument that it serves as an unalterable limit on the textual commitment of impeachments, with judicial review available for at least some claims of Senate disregard, is far more plausible under Powell than Nixon's effort to find justiciable limits in the word "try”. However, we need not decide this issue, so we leave it for the unlikely day of its arising.
. Because Nixon’s claim is nonjusticiable, we need not address appellees’ argument that Nixon should have sued in Claims Court rather than district court.