dissenting.
Five years ago, I argued that Bracy and Collins ought to be given the chance to conduct discovery in an effort to marshal evidence that Judge Maloney’s serial bribe-taking had an impact on their trial. “If their discovery proves fruitless,” I said, “we can at least take comfort in the knowledge that we have given them every opportunity to prove that Maloney’s corruption deprived them of a fair trial.” Bracy v. Gramley, 81 F.3d 684, 699 (7th Cir.1996) (dissent), rev’d in part, 520 U.S. 899, 117 S.Ct. 1793, 138 L.Ed.2d 97 (1997). The petitioners have now had their chance at discovery. If proof that a corrupt judge is biased in cases where no bribe is tendered were easy to come by, the petitioners would no doubt have found it. But smoking guns are rare, and I agree with my colleagues that discovery has yielded no hard proof that Maloney engaged in camouflaging bias in their particular case. *610What discovery has produced, however, should make us feel less, rather than more, confident that the petitioners’ convictions and sentences were untainted by Malo-ney’s bribe-taking.
We know, of course, as we did before, that Maloney was engaged in serial bribe-taking within the time frame surrounding the petitioners’ convictions. The jury’s finding against Maloney reveals that he accepted bribes to fix at least four cases: People v. Lenny Chow, No. 81 C 4020, People v. Ronald Roby, No. 82 I 50244, People v. Owen Jones, No. 81 C 9832, and People v. Earl Hawkins & Nathson Fields, No. 85 C 6555. The prosecution’s case against Maloney posited that he accepted bribes in at least two other cases-People v. Frank Calistro, No. 82 C 8355, and People v. Wilfred Rosario, No. 79 C 2469—although the jury did not find the alleged bribe in Calistro to have been proved beyond a reasonable doubt and it was not asked to render a finding as to the alleged bribe in Rosario. Hearsay evidence in the government’s investigative file on Maloney suggests that he may also have accepted a bribe to acquit the defendant of murder in People v. Rocco LaMantia, No. 79 C 2623, in 1981. R. 162 Ex. 23 at 6, Ex. 24 at 1-2; see Collins v. Welborn, 79 F.Supp.2d 898, 904 ¶ 19 (N.D.Ill.1999).
The record now before the court also makes it both possible and reasonable to infer that Maloney accepted bribes in many other cases. An FBI analysis of Ma-loney’s finances reveals that his expenditures in the years 1978 to 1984 exceeded his known, legitimate income in those years by some $400,000. R. 162 Exs. 53, 54; 79 F.Supp.2d at 907 ¶ 40. Assuming that the typical bribe was in the $5,000 to $10,000 range, then Maloney may have fixed at least 40 cases, and perhaps as many as 80 during those years.1
Five of the identified acts of bribery took place between 1981 and 1983, and one in 1986. (Bracy and Collins, of course, were arraigned in March of 1981, and they were tried, convicted, and sentenced to death in July 1981.) Two of the known bribes were tendered to Maloney in proceedings that took place shortly before and shortly after Bracy and Collins were tried in July 1981: after accepting a bribe of $2,500 in People v. Rosario, Maloney suppressed the defendant’s confession on January 23, 1981, and dismissed the case on June 17, 1981; and in mid-August, 1981, after accepting an unknown share in a bribe of $100,000, Maloney acquitted all three defendants in People v. Chow. 79 F.Supp.2d at 903-04 ¶¶ 12, 16. In short, we know that Maloney’s bribe-taking was neither isolated nor sporadic, and that Bracy and Collins were tried in the midst of it.
The record also supplies some confirmation that Maloney engaged in acts designed to further and/or camouflage his bribe-taking. First, now part of the rec*611ord in this case is the transcript of Malo-ney’s trial (which previously was under seal and unavailable to the petitioners). That transcript includes the testimony of William Swano, recounted in my previous dissent. See 81 F.3d at 697. In short, Swano, who had bribed Maloney in other cases, had a case before Maloney — People v. James Davis — that he believed to be a sure winner for his client. On Swano’s advice, his client waived his right to a jury and the case was tried to Maloney, whom Swano did not think it necessary to bribe in view of the merits of the case. Maloney nonetheless convicted Swano’s client, which Swano construed as a lesson that “to practice in front of Judge Maloney ... we had to pay.” R. 162 Ex. 20, United States v. Maloney & McGee, No. 91 CR 477, Trial Tr. at 2530. Without more, one might say that this was nothing more than Swano’s belief, and that it is entirely possible that Maloney convicted Davis wholly on the merits of the prosecution’s case. But Swa-no subsequently met with bagman Robert McGee to discuss a bribe in Hawkins & Fields. McGee told Swano that Maloney had okayed the discussion, acknowledging that he had “screwed” Swano in the Davis case. R. 241, United States v. Maloney & McGee, No. 91 CR 477, Trial. Tr. at 2568. That acknowledgment constitutes confirmation that Maloney did engage in camouflaging bias.
Second, Maloney initially accepted a $10,000 bribe in People v. Hawkins & Fields, only to return the money and convict the defendants when he perceived (correctly) that the FBI had its eye on him. The Illinois Supreme Court subsequently concluded that Hawkins and Fields were entitled to a new trial, because Maloney had been motivated to convict them in order to deflect suspicion. People v. Hawkins, 181 Ill.2d 41, 228 Ill.Dec. 924, 690 N.E.2d 999, 1004 (1998).
Third, in People v. Dino Titone, No. 83 C 127, Maloney accepted a $10,000 bribe to acquit Titone, but convicted him anyway, again to deflect suspicion. That compensatory motive persuaded the postconviction judge to vacate the defendant’s conviction and grant him a new trial. R. 239, People v. Titone, No. 83 C 127, Post-conviction Tr. at 10-13 (Cir. Ct. Cook County July 25, 1997).
Fourth, we have the version of the offense that the government submitted to the court for purposes of Maloney’s sentencing, which wholly embraces and advo-. cates the notion that Maloney engaged in compensatory bias.
... THOMAS MALONEY’S corruption began at the time he was a criminal defense attorney paying off judges and court personnel to fix cases — including a notorious murder case — and continued through the time he was a judge working as a mafia factotum in the Cook County Circuit Court system and taking all manners of bribes on very serious criminal cases. Thomas Maloney’s reputation as a strict prosecution-oriented judge was not a mistake. By casting this image, Maloney sought to deflect suspicion from his criminal activity, while simultaneously giving select desperate defendants who knew the right people an incentive to pay him off. Thus, by using his position as a felony trial court judge to extract bribes from defendants who faee[d] long periods of imprisonment or execution, THOMAS MALONEY far surpassed the category of corrupt jurist to chart a new territory of defilement.
* * *
... [W]hen he got his turn on the bench, THOMAS MALONEY imposed a sinister system which had the dual effect of concealing and promoting his corruption. THOMAS MALONEY the former *612champion of the defendant became one of the most ruthless judges on the bench. Showing defendants little mercy had the effect of diverting any conceivable suspicion from MALONEY while at the same time giving defendants a strong motivation to cough up big bribery dollars.
R. 162 Ex. 1, United States v. Maloney & McGee, No. 91 CR 477, Government’s Official Version of the Offense at 54-55 (N.D. Ill. filed May 23, 1994). As my colleagues suggest, the government certainly had an incentive in submitting this memorandum to portray Maloney in the worst possible light and so to persuade Judge Leinenwe-ber, who sentenced Maloney, to mete out the harshest possible penalty. Yet, the government obviously had no interest in helping Bracy, Collins, or any other defendant convicted by Maloney to obtain relief from their convictions. It strikes me as doubtful that any prosecutor — federal or state — would admit that a judge deliberately favored the prosecution for nefarious reasons if the facts did not warrant that acknowledgment. Having investigated and tried Maloney, his prosecutors were in the best position to assess his modus oper-andi as a corrupt jurist and its ramifications for defendants who did not bribe him. Their considered judgment, as officers of the court, that Maloney did engage in purposeful and systematic compensatory bias deserves much more weight than my colleagues (ensconced as we all are within the sheltered environs of this court) are willing to give it.
We know one more fact. In 1994, Malo-ney stood before Judge Leinenweber defiantly proclaiming his innocence. By this time, of course, the proof of Maloney’s bribe-taking had already convinced a jury beyond a reasonable doubt that he was a criminal. Nonetheless, as he waited for Judge Leinenweber to pass sentence, Ma-loney continued to insist that he had been an honest judge with a distinguished career. Among the cases he cited as a credit to his career, and as proof that he was not corrupt — in addition to Hawkins & Fields, where we know that Maloney engaged in compensatory bias — were the convictions of Bracy and Collins. R. 162 Ex. 13, United States v. Maloney & McGee, No. 91 CR 477, Sentencing Tr. at 606-07 (N.D.Ill. July 21, 1994).
In sum, the record supplies us with confirmation that Maloney was engaged in serial bribe-taking that almost certainly was more extensive than the particular bribes underlying his conviction reveal; that he accepted bribes in cases immediately before and after Bracy and Collins were tried; that he was cognizant of the public scrutiny that his bribe-taking could engender and did not hesitate to take compensatory action — including the return of a bribe, and the conviction of an individual who had bribed him — in order to protect himself; and that the very government that built the case against Maloney believed that he cultivated a reputation as a fierce, law-and-order judge both to deflect suspicion from his bribe-taking and to encourage defendants to bribe him.
What we do not have, and what my colleagues insist is a prerequisite to relief, is hard proof that camouflaging bias was at work in this case. The theory that Bracy floated before the Supreme Court — that Robert McDonnell was Maloney’s former law partner, and that Maloney agreed to appoint him as Bracy’s counsel in exchange for a promise that he would accept a quick trial, so that the trial of Bracy and Collins might serve to camouflage the corrupt disposition of the Chow murder prosecution, see 520 U.S. at 907-08, 117 S.Ct. at 1798-99 — was not borne out by the *613facts.2 Maloney, who continues to deny that he ever accepted a bribe, of course will not admit that he ever engaged in camouflaging bias in any case, let alone this particular one. And although there is proof that Maloney may have taken compensatory action in other cases to conceal and further his bribe-taking franchise, there is no proof that he invariably engaged in camouflaging bias in cases where no bribe was tendered.
Proof along these lines could come from only two sources — Maloney, and those who associated with him in his bribe-taking. Maloney is the only individual who would necessarily know when he took compensatory actions to hide his bribe-taking, but the questioning of him quickly reached a dead end. Maloney continues to “vehemently and arrogantly den[y] all of the bribery charges clearly established by the jury findings and the evidence presented at his criminal trial.” 79 F.Supp.2d at 907 ¶ 41. Those who bribed Maloney or who facilitated his bribe-taking — people like Swano, who learned firsthand the need to bribe Maloney in order to obtain an acquittal, and Lucius Robinson, who served as both bailiff and bagman for Maloney— might have a glimpse into Maloney’s mind. But in the finest tradition of mobsters, swindlers, and certain campaign fund-raisers, Swano declined the opportunity to further incriminate himself, R. 218 Ex. 14, Deposition of William Swano, and Robinson had nothing more insightful to say than Maloney was known as “a hard fucking judge” who “more or less leaned toward [prosecutors and police officers] in cases,” R. 213 Ex. 10(b), Deposition of Lucius Robinson, at 47, 48.3
But the lack of such proof by no means rules out the possibility that Maloney did engage in camouflaging bias on a regular basis, and possibly in this case. Bribery itself is quite difficult to expose. The fact that an acquittal (or some other dispositive ruling for the defendant) has been paid for is rarely evident from the record. Individual judges often take widely disparate views of the facts and even the law, and legitimately so. Given the breadth of a judge’s discretion, and his nearly unfettered power to render credibility assessments, one can almost never say with certainty that a surprising ruling in favor of the defendant is necessarily the product of corruption. To show that money has changed hands often requires years of work using the full array of investigative resources (electronic surveillance, undercover agents, and cooperating witnesses). Proof that a corrupt judge engaged in camouflaging bias will be at least as difficult to obtain, if not more so, and persons in the position of Bracy and Collins must *614search for that proof without the resources and might of the government. If a judge decides to hide his bribe-taking by favoring the prosecution in a particular case, there will be little to mark that decision. No money will change hands, the prosecutor will not be given a whispered assurance that “things have been taken care of,” and there will be no one in the courtroom (save the judge) patiently awaiting a foreordained result. Absent a lopsided record, the only hard proof of compensatory bias can issue from the judge’s own mouth. Where, as here, the judge does not even acknowledge his own bribe-taking, and associates who may have been privy to his thinking plead either ignorance or the Fifth Amendment, then the proof is wholly inaccessible.
Yet, proof of actual bias has never been treated as the sine qua non of a due process claim. The Supreme Court has not hesitated to vacate a conviction where it is shown that the judge who presided at trial had an interest in the outcome, even if there is no proof that this interest manifested in actual bias against either party. Thus, in Tumey v. Ohio, 273 U.S. 510, 47 S.Ct. 437, 71 L.Ed. 749 (1927), the Court invalidated a statutory scheme that authorized town mayors to try persons accused of liquor offenses and provided for reimbursement of the mayors’ trial costs from the fines imposed on those they convicted of such offenses. Because the scheme in this way gave the mayor-judge a pecuniary interest, albeit a relatively modest one, in the outcome of the trial, the court reasoned, it was inconsistent with due process.
There are doubtless mayors who would not allow such a consideration as $12 costs in each case to affect their judgment in it, but the requirement of due process of law in judicial procedure is not satisfied by the argument that men of the highest honor and the greatest self-sacrifice could carry it on without danger of prejudice. Every procedure which would offer a possible temptation to the average man as a judge to forget the burden of proof required to convict the defendant, or which might lead him not to hold the balance nice, clear, and true between the state and the accused denies the latter due process of law.
273 U.S. at 532, 47 S.Ct. at 444 (emphasis supplied). Similarly, the Court in In re Murchison, 349 U.S. 133, 75 S.Ct. 623, 99 L.Ed. 942 (1955), concluded that due process would not permit the same judge who acted as a one-person grand jury under Michigan law to try contempt charges arising out of the grand jury proceedings. “Fairness of course requires an absence of actual bias in the trial of cases,” the Court observed. Id. at 136, 75 S.Ct. at 625. “But our system of law has always endeavored to prevent even the probability of unfairness.” Ibid, (emphasis supplied). And in Aetna Life Ins. Co. v. Lavoie, 475 U.S. 813, 106 S.Ct. 1580, 89 L.Ed.2d 823 (1986), the Court found that a due process violation occurred where a member of the Alabama Supreme Court had cast the deciding vote and written the decision affirming a $3.5 million punitive damage award against an insurance company for its bad-faith refusal to pay a valid claim at the same time that the justice had at least one similar lawsuit of his own pending against an insurer in a lower Alabama court. That pending action, the Court reasoned, gave the justice a pecuniary interest in the outcome of the decision that he authored. Id. at 824-25, 106 S.Ct. at 1586-87. The Court did not find it necessary to decide whether the justice was actually biased by virtue of his own litigation; the possibility that he might have been so was sufficient to establish a due process violation. Id. at 825, 106 S.Ct. at 1587.
*615Tbe Due Process Clause “may sometimes bar trial by judges who have no actual bias and who would do their very best to weigh the scales of justice equally between contending parties. But to perform its high function in the best way, ‘justice must satisfy the appearance of justice.’ ”
Ibid., quoting Murchison, 349 U.S. at 136, 75 S.Ct. at 625 (additional citation omitted).
The circumstances here likewise argue in favor of treating the judge’s temptation to lean in one party’s favor as disqualifying in and of itself. Common sense tells us that a corrupt judge who wishes to retain his office will do whatever he can to hide his bribe-taking. An obvious way to do that is to make it a practice to favor the State in cases where no bribe is tendered.4 There is an inherent difficulty in the attempt to peer into a judge’s mind to ascertain if, when, and how the judge engaged in compensatory actions to deflect suspicion away from his wrongdoing. But the possible temptation to cloak his bribe-taking in this way was present nonetheless. And the presence of that temptation in turn undermines confidence in the fairness of the trials over which the corrupt judge presided.
My colleagues suggest that the Supreme Court has already rejected the notion that a corrupt judge’s temptation to engage in compensatory bias is sufficient to place the case within the Tumey-Murchison-Aetna framework, but instead has conditioned relief on proof of actual bias. Ante at 606-07. It is true that the Court remanded this case so that the petitioners would have the chance to conduct discovery aimed at showing actual bias. It is also true that the Court acknowledged, in a footnote, that discovery would be unnecessary if proof of actual bias were not required. 520 U.S. at 903 n. 4, 117 S.Ct. at 1796 n. 4. But I think it wrong to suggest that the Court, by implication, has already decided that proof of actual bias must be shown on the part of a corrupt judge. The Court granted certiorari to decide one and only one question — whether the petitioners were entitled to discovery. Bracy v. Gramley, 519 U.S. 1074, 117 S.Ct. 726, 136 L.Ed.2d 643 (1997). Had the discovery yielded proof that Judge Maloney was actually biased against Bracy and Collins, then there would be no need to consider whether his bribe-taking might entitle them to relief even in the absence of that proof. It may be that when confronted with the question of whether actual bias must be shown, the Supreme Court will decide the issue as my colleagues have predicted. But we have no business attributing a holding to the Court on a question as to which it did not grant certiorari, it did not give the parties an opportunity to brief, and which was not the subject of argument. The Supreme Court does not make law by stealth. When it decides the issue, we will know it.
We have a duty at this juncture to consider not only whether Bracy and Collins have succeeded in proving actual bias, but also whether, in light of what their discov*616ery has (and has not) produced, the burden we have assigned to them is a realistic and appropriate one. Discovery, after all, has not proven that Maloney was in fact fair to any defendant who did not bribe him. It has simply failed to produce proof that he was actually biased against Bracy and Collins. It has, however, established that Ma-loney’s bribe-taking was not isolated, that he was sensitive to public detection of his bribe-taking, that on occasion he took compensatory actions in order to hide and further his bribe-taking, and that the successful prosecution of Bracy and Collins was a case that he could and did cite as proof that he was not corrupt. At the same time, the limits of discovery have been laid bare: the principal wrongdoer refused to acknowledge even taking a bribe, and others who might have provided some insight into his practices either claimed an inability to do so or refused to speak at all. There remains the very real possibility that Maloney might have engaged in compensatory bias against the petitioners, but without the glimpse into his mind that my colleagues agree would be difficult to obtain, ante at 606, we will never know it.
And so we return to the fundamental problem that I highlighted in my first dissent. My colleagues believe that in the absence of any proof establishing the taint of compensatory bias in this particular case, we should assume that Maloney gave Bracy and Collins a fair trial. I remain puzzled as to the justification for this. After all, as the Supreme Court has recognized, the presumption that Maloney properly discharged his duties as a judge “has been soundly rebuttedf.]” 520 U.S. at 909, 117 S.Ct. at 1799. Moreover, in contrast to the Tumey-Murchison-Aetna line of cases, we are not simply dealing here with a conflicted judge, but a corrupt one. Common sense suggests that the corrupt judge would be even more likely to yield to the temptation not to “hold the balance nice, clear, and true between the state and the accused,” Turney, 273 U.S. at 532, 47 S.Ct. at 444, than an honest one. Yet the court clings to the idea that Maloney’s bribe-taking was insular, and that he rendered acceptably fair trials when he did not take bribes. I doubt that my colleagues would be so complacent if they stood in the petitioners’ shoes. If a court system promised everyone a fair trial, but offered to stack the deck in a party’s favor for an appropriate fee, the cash machines and loan sharks nearest to the courthouse would be swamped with business, and not simply because everyone would want the favors that the fee might secure for them, but because no one could be confident in the fairness of supposedly impartial proceedings that others were buying their way out of. No one convicted by or before Maloney can feel sanguine, knowing that for as little as $2,500 he might have walked out of his courtroom a free person. And none can be sure that, in the absence of a bribe, Maloney did not take the opportunity to hide and/or promote his bribe-taking.
The driving factor behind the court’s presumption that Maloney was impartial when not bribed is the same one that has lurked in the background since the start of this case — our reluctance to set a precedent that might undo the convictions of all those persons who were tried and convicted before Maloney. See ante at 607. “To reverse [the petitioners’] convictions or sentences would merely compound Malo-ney’s wrongdoing,” my colleagues observe. Ante at 609. That view, however, underestimates the profound damage that a corrupt judge inflicts by shaking the public faith in the fairness of our judicial process. Ignoring the impact that a judge’s systematic acceptance of bribes has upon cases in which no bribe was tendered will simply magnify the distrust that results from offi*617cial corruption. The truth is, given the passage of time, most people tried and sentenced by Maloney have served their prison terms, so few people would go free were retrials ordered for defendants like Bracy and Collins. But confidence in the integrity of the judicial system would be promoted. And in a State in which bribery has by no means vanished, requiring such retrials might encourage the State and its citizens to scrutinize the conduct of our public officials with a keener eye.
It strikes me that the court’s disposition of this case is laden with irony. Defendants tried before judges of unquestioned integrity, with a possible temptation to favor the government, are entitled to new trials. Turney, 273 U.S. 510, 47 S.Ct. 437, 71 L.Ed. 749. Defendants who bribed Ma-loney, only to be convicted by him when he decided that fixing their cases was too risky, have won new trials. Hawkins, 181 Ill.2d 41, 228 Ill.Dec. 924, 690 N.E.2d 999. Even defendants who managed to initially escape conviction by successfully bribing Maloney are being given exactly what Bra-cy and Collins seek today — a fair trial before an honest and impartial judge. People v. Aleman, 1994 WL 684499 (Ill. Cir. Oct. 12, 1994), aff'd, 281 Ill.App.3d 991, 217 Ill.Dec. 526, 667 N.E.2d 615 (1996), cert. denied, 519 U.S. 1128, 117 S.Ct. 986, 136 L.Ed.2d 868 (1997); see also Aleman v. Honorable Judges of Circuit Court of Cook County, 138 F.3d 302 (7th Cir.), cert. denied, 525 U.S. 868, 119 S.Ct. 162, 142 L.Ed.2d 132 (1998). But these two petitioners, sentenced to death before a judge whose career as an attorney and judge was corrupt from beginning to end, are told that the judgment of a racketeer is perfectly satisfactory. It is a sad day indeed when defendants who attempted to purchase their way out of a conviction receive a greater measure of justice than those who did not.
I respectfully dissent.
Before FLAUM, Chief Judge, and POSNER, COFFEY, EASTERBROOK, RIPPLE, MANION, KANNE, ILANA DIAMOND ROVNER, DIANE P. WOOD, TERENCE T. EVANS and ANN CLAIRE WILLIAMS, Circuit Judges.ORDER
June 25, 2001
The petition for rehearing en banc in the above-entitled case is GRANTED, the panel decision is VACATED, and the appeal is restored to the calendar for reargument before the full court at a date and time to be announced.
. The known bribes amounted to less than $5,000 in several instances and in some cases $10,000 or more. The greatest known bribe tendered to Maloney was an undisclosed share of the $100,000 payoff in Chow for the acquittal of three gang members in a notorious shooting in Chicago’s Chinatown. The other known bribes include the tender of $10,000 in Hawkins & Fields to acquit one or both of two El Rukn gang members of a double murder (a bribe that Maloney subsequently returned when he suspected that the FBI was monitoring the case), the $2,300 Maloney accepted in Roby in exchange for a sentence of probation, and the $4,000 to $5,000 Maloney accepted to acquit the defendant in Jones of felony murder and sentence him to a term of nine years on the lesser charge of voluntary manslaughter. The alleged bribes tendered in Calistro (for a sentence of probation) and Rosario (where the defendant's confession was suppressed and the case against him was eventually dismissed) were each in the amount of $2,500.
. I would note, however, that discovery has hardly given us reassurance as to McDonnell’s credentials. It turns out that McDonnell was a twice-convicted felon who managed to regain his law license in 1980, the year before Bracy and Collins were tried. R. 213 Ex. 8, Deposition of Robert McDonnell, at 26-32; see United States v. Mitro, 435 F.2d 839 (7th Cir.1970); In re McDonnell, 82 Ill.2d 481, 45 Ill.Dec. 897, 413 N.E.2d 375 (1980). Coincidentally, McDonnell also had a reputation among his colleagues as a fixer of cases. R. 213, Ex. 11, Deposition of Terrence Hake, at 16. Years later, McDonnell would add yet another conviction to his curriculum vitae. See United States v. Taglia & McDonnell, 922 F.2d 413 (7th Cir.), cert. denied, 500 U.S. 927, 111 S.Ct. 2040, 114 L.Ed.2d 125 (1991).
. The only other readily apparent means of establishing actual compensatory bias — one that the state’s counsel (who with commendable grace and patience endured my pointed questions) identified at oral argument — would be a pattern of rulings and conduct so overtly hostile to the defense that the judge’s bias would be readily apparent from the record. That is obviously not the case here, or the petitioners’ convictions would long ago have been vacated.
. My colleagues, of course, remain skeptical that a judge would engage in such compensatory bias. They have suggested that a corrupt judge might do better to cultivate a defense-friendly reputation, so that paid-for acquittals look less out of place. See ante at 606. That skepticism, however, overlooks the proof that Maloney did not hesitate to engage in compensatory bias on at least two occasions, when he convicted defendants who had bribed him in order to deflect the government's suspicions — notwithstanding the incentive that his turnabout gave to those defendants to blow the whistle on him. It also overlooks Swano's testimony, which suggests that Maloney engaged in compensatory bias not simply to avoid suspicion, but to cultivate additional bribes.