The simple and undisputed facts of this case present interesting questions regarding the limits of prosecutorial discretion. Specifically, we are called upon to apply the doctrine of "prosecutorial vindictiveness" outlined in Blackledge v. Perry, 417 U.S. 21, 94 S. Ct. 2098, 40 L. Ed. 2d 628 (1974). A panel of this Court considered this question and produced three separate opinions without any consensus. United States v. Andrews, 612 F.2d 235 (6th Cir. 1979). Given the importance of the issue and the division of the panel, we granted the parties' cross-petitions for rehearing en banc.

Tallice Andrews, Thurston Brooks and Fannie Braswell were arrested at Detroit's Metropolitan Airport on November 16, 1975.1 On November 8, 1976, they were jointly indicted for narcotics and firearms offenses.2 On December 29, 1976, defendants Andrews and Brooks were arraigned before a Magistrate on the offenses. Both requested that they be released on bail pending trial, but the government was strongly opposed. Apparently, the third defendant, Fannie Braswell, had turned state's evidence and had been threatened. The government responded by placing Ms. Braswell in the federal witness protection program and urging that defendants Brooks and Andrews not be released on bail. Based on testimony by the Assistant United States Attorney regarding the threats, the Magistrate denied bail and remanded the two defendants to federal custody. On January 7, 1977, the jailed defendants appealed the Magistrate's decision to the district judge to whom the case was assigned. The judge overturned the Magistrate's decision and the defendants were released on bond on January 11, 1977. Two days later, on January 13, 1977, the Assistant United States Attorney sought and obtained a superseding indictment charging defendants with an additional conspiracy count.3

The district court, in a comprehensive opinion reported at 444 F. Supp. 1238 (E.D.Mich.1978), concluded that the conspiracy charge, added after defendants successfully pressed their bail appeal, was impermissible under Blackledge v. Perry, 417 U.S. 21, 94 S. Ct. 2098, 40 L. Ed. 2d 628 (1974), and North Carolina v. Pearce, 395 U.S. 711, 89 S. Ct. 2072, 23 L. Ed. 2d 656 (1969). The government appealed.

A divided panel of this court reversed. 612 F.2d 253 (6th Cir. 1979). There was no unifying rationale among the members of the panel. Judge Green advanced a rebuttable prima-facie vindictiveness test and remanded to the district court for reconsideration in light of that test. Judge Merritt concurred "in the result and much of the reasoning of Judge Green's opinion for the court," but stated that prosecutorial vindictiveness issues should be restricted to "post trial prosecutorial conduct which undermines double jeopardy values." Judge Merritt's view called for outright reversal, not remand. The author of this opinion dissented. I advanced an overall balancing test, and, applying that test, would have affirmed the district court.

Upon rehearing en banc, a majority of the court has agreed on the views expressed herein.

There are three Supreme Court cases which are relevant to our inquiry, all authored by Mr. Justice Stewart. First, is North Carolina v. Pearce, 395 U.S. 711, 89 S. Ct. 2072, 23 L. Ed. 2d 656 (1969). There, the Court ruled that a trial judge could not resentence a defendant to a longer sentence than had been originally imposed where the defendant was convicted, sentenced, won reversal on appeal and was then reconvicted and resentenced. The Court advanced two reasons for its holding. First, "(d)ue process of law ... requires that vindictiveness against a defendant for having successfully attacked his first conviction must play no part in the sentence he receives after a new trial." Second, "since the fear of such vindictiveness may unconstitutionally deter a defendant's exercise of the right to appeal or collaterally attack his first conviction, due process also requires that a defendant be freed of apprehension of such a retaliatory motivation on the part of the sentencing judge." Pearce, supra, 395 U.S. at 725, 89 S. Ct. at 2080. In order to effectively police vindictive judicial behavior, the Court advanced a "prophylactic rule." A stiffer re-sentencing could only occur where there were objective, on the record facts concerning conduct which took place after the initial sentencing. Id. at 726, 89 S.Ct. at 2081.

In two later cases, the Court declined to apply Pearce. In Colten v. Kentucky, 407 U.S. 104, 92 S. Ct. 1953, 32 L. Ed. 2d 584 (1972), the defendant appealed a misdemeanor conviction to a trial de novo before a different judge. The Court had no problem with a stiffer sentence on retrial since "(t)he possibility of vindictiveness, found to exist in Pearce, is not inherent in the Kentucky two-tier system." Id. at 116, 92 S.Ct. at 1960. The Court thought that there was little chance of "personal vindictiveness" before a different judge. In Chaffin v. Stynchcombe, 412 U.S. 17, 93 S. Ct. 1977, 36 L. Ed. 2d 714 (1973), the Court considered a situation where the petitioner was reconvicted and resentenced by a different jury after having an initial conviction overturned on appeal. The Court concluded that Pearce was inapplicable because "(t)he potential for such abuse of the sentencing process by the jury is ... de minimis in a properly controlled retrial." Id. at 26, 93 S.Ct. at 1982. The Court explained that "the jury, unlike the judge who has been reversed, will have no personal stake in the prior conviction and no motivation to engage in self-vindication." Id. at 27, 93 S.Ct. at 1983.

After this, the Court decided Blackledge v. Perry, 417 U.S. 21, 94 S. Ct. 2098, 40 L. Ed. 2d 628 (1974). There, a defendant was convicted on a misdemeanor assault charge in a state court of limited jurisdiction. The defendant exercised his statutory right to appeal to a trial de novo. The prosecutor responded by bringing a more serious felony assault charge against the defendant. This charge encompassed the same conduct for which the defendant had been charged with a misdemeanor in the lower court. The Supreme Court held that this conduct by the prosecutor violated due process. The Court contrasted Colten and Chaffin by noting that in those cases there was no "realistic likelihood of vindictiveness." Id. at 27, 94 S.Ct. at 2102. In the case before it, however, the Court concluded that the prosecutor had a "considerable stake in discouraging convicted misdemeanants from appealing and thus obtaining a trial de novo in the Superior Court." Id. This was because "such an appeal will clearly require increased expenditures of prosecutorial resources before the defendant's conviction becomes final, and may even result in a formerly convicted defendant's going free." Id. The Court concluded by following the rationale of Pearce and applying it to the prosecutor's conduct:

There is, of course, no evidence that the prosecutor in this case acted in bad faith or maliciously in seeking a felony indictment against Perry. The rationale of our judgment in the Pearce case, however, was not grounded upon the proposition that actual retaliatory motivation must inevitably exist. Rather, we emphasized that "since the fear of such vindictiveness may unconstitutionally deter a defendant's exercise of the right to appeal or collaterally attack his first conviction, due process also requires that a defendant be freed of apprehension of such a retaliatory motivation on the part of the sentencing judge." 395 U.S., at 725, (89 S.Ct. at 2080). We think it clear that the same considerations apply here. A person convicted of an offense is entitled to pursue his statutory right to a trial de novo without apprehension that the state will retaliate by substituting a more serious charge for the original one, thus subjecting him to a significantly increased potential period of incarceration. Blackledge, supra, 417 U.S. at 28, 94 S. Ct. at 2102.

In the case at bar, the factual situation is different from that in Blackledge. First, this case presents a pre-trial situation where the government and the defendants wrangled over a bail motion. Second, in this case, the government added an additional conspiracy charge which encompassed the same basic conduct for which the defendants had been originally charged. In Pearce, the prosecutor substituted a more severe charge.

The policy questions which this case presents were the subject of a lively debate among the members of the original panel. Prosecutorial vindictiveness issues have similarly divided other courts.4 We will not repeat the extensive discussion conducted in the panel opinions. Simply stated, we must reconcile two conflicting rules of law: 1) prosecutors have and need broad discretion to file charges where there is probable cause that someone has broken the law; 2) vindictive conduct by persons with the awesome power of prosecutors (and judges) is unacceptable and requires control.

We think that Blackledge points the way to the standard which a court must test for in prosecutorial vindictiveness situations. That standard is whether, in the particular factual situation presented, there existed a "realistic likelihood of vindictiveness" for the prosecutor's augmentation of the charges.

Blackledge was a clear case. The prosecutor had not just an incentive but a "considerable stake" in deterring a defendant from appealing to a trial de novo. Such an appeal meant that the prosecutor had to present his case over again. In addition, the prosecutor substituted the felony assault charge after charging the defendant with the misdemeanor assault charge in the lower court. Unlike a situation where a prosecutor could argue that he forgot to add an additional count where a defendant committed multiple criminal acts, the prosecutor in Blackledge brought a felony charge only after he exercised his discretion to simply charge the defendant with a lesser offense at the first trial. Finally the prosecutor did not try to explain his actions. Anyone who assessed these facts objectively would conclude that the prosecutor had retaliated against the defendant. Blackledge presented the clearest possible case of prosecutorial vindictiveness short of an actual confession by the prosecutor.

Despite this, the Court did not base its ruling in Blackledge on a finding that actual vindictive conduct had taken place. Rather, the Court found it sufficient that there existed a "realistic likelihood of vindictiveness." Put another way, the Court concluded that "the opportunities for vindictiveness in this situation (were) such as to impel the conclusion that due process of law requires a rule analogous to that of the Pearce case." Id. at 27, 94 S.Ct. at 2102.

We think that a court faced with vindictiveness allegations must assess the fact situation before it to see if the above-stated standard is met. In order to make this assessment, a court must weigh two factors. First, there is the prosecutor's "stake" in deterring the exercise of some right. In Blackledge the Court described the prosecutor's "stake" in deterring appeals to a trial de novo as "considerable" since the prosecutor would have to try the case all over again. In this case, the prosecutor's interest in deterring a bail motion is not as great as a prosecutor's interest in deterring appeals to a trial de novo. A prosecutor who loses a bail motion does not have to retry a case. However, we cannot accept the government's argument that the prosecutorial stake in a pretrial setting is always so de minimus that there can never be a "realistic likelihood of vindictiveness" in a pre-trial setting.5 Each situation will necessarily turn on its own facts.

Second, there is the prosecutor's conduct. A prosecutor who adds on extra charges after the exercise of a procedural right is arguably acting less vindictively than a prosecutor who substitutes a more severe charge for a less severe one. In the first situation, a prosecutor might well have made an honest mistake. This is especially true where a defendant has committed multiple criminal acts. However, in the second situation, the prosecutor will have already exercised his discretion, and the probability that the prosecutor acted vindictively is higher. See United States v. Andrews, 612 F.2d at 240-41 (opinion of Green, J.). Again, there is no per se rule here.

By carefully analyzing the circumstances, the district court can make the determination whether there existed a realistic likelihood of vindictiveness. The standard itself, however, is an objective one-whether a reasonable person would think there existed a realistic likelihood of vindictiveness. The proper standard does not depend on a defendant's subjective impressions.

We emphasize that Blackledge does not require that a defendant show that a prosecutor was actually vindictive. Blackledge's proscriptions are triggered where a defendant shows that there is a realistic likelihood that a prosecutor acted vindictively.6 There are sound reasons for this. First, this takes into account the due process value that defendants be "freed of apprehension of such a retaliatory motivation" on the part of a judge or prosecutor. Blackledge, supra 417 U.S. at 28, 94 S. Ct. at 2102, quoting Pearce, supra 395 U.S. at 725, 89 S. Ct. at 2080. Second, such a standard is the only realistic way to police vindictiveness. As the Court noted in Pearce, supra at 724 n.20, 89 S. Ct. at 2080 n.20, "the existence of a retaliatory motivation would, of course, be extremely difficult to prove in any individual case."7 Third, there is an important policy consideration. Prosecutors and criminal defendants are adversaries and prosecutors make honest mistakes. If the standard to be applied was proof of actual vindictiveness, a trial judge would have the Hobson's choice of either not barring the extra charge or of saying that a prosecutor acted wrongly. In some cases, a trial judge would, in effect, be calling a prosecutor a liar where the prosecutor claimed inadvertence and the judge ruled against him. We do not think that such confrontations before the judiciary and the executive branch are desirable. A standard of "realistic likelihood of vindictiveness" allows the barring of charges in appropriate situations without the need to find that the prosecutor acted in bad faith.8

The government relies on a statement in the Supreme Court's later opinion in Bordenkircher v. Hayes, 434 U.S. 357, 98 S. Ct. 663, 54 L. Ed. 2d 604 (1978), a decision which will be more fully discussed below. In Bordenkircher the Court stated:

The court has emphasized that due process violation in cases such as Pearce and Perry lay not in the possibility that a defendant might be deterred from the exercise of a legal right ... but rather in the danger that the state might be retaliating against the accused for lawfully attacking his conviction. Bordenkircher, supra at 363, 98 S. Ct. at 667-68. (emphasis added)

From this, the government argues that the sole consideration for a court faced with prosecutorial vindictiveness questions is whether the prosecutor, in fact, acted vindictively. In other words, the government is saying that the alternative due process rationale of Blackledge, halting defendants' apprehension of vindictiveness, does not survive Bordenkircher.9

We agree with the government that the mere possibility that prosecutorial or judicial conduct may be vindictive is insufficient to trigger judicial sanctions. It is only where the possibility of vindictiveness is substantial that Pearce-Blackledge sanctions are triggered. The simple fear on the part of a defendant that a prosecutor may be retaliating against him is insufficient, by itself. Put another way, the mere appearance of vindictiveness is not enough to trigger Pearce-Blackledge sanctions. The factual situation must pose a realistic likelihood of vindictiveness. See Chaffin v. Stynchcombe, 412 U.S. 17, 29, 93 S. Ct. 1977, 1984, 36 L. Ed. 2d 714 (1973) ("the court (in Pearce) intimated no doubt about the constitutional validity of higher sentences in the absence of vindictiveness despite whatever incidental deterrent effect they might have on the right to appeal.")

The due process principle that defendants not be deterred from exercising their rights is a limited one. As the Court explained in Chaffin, many situations present difficult choices for defendants and arguably chill the exercise of a right. By itself, however, defendants' apprehension does not justify due process protection. Only when there is a situation presenting a probability of vindictiveness does this due process interest come into play.

If, after carefully assessing a prosecutor's seemingly retaliatory adding of charges, a court finds that the situation before it presents a realistic likelihood of vindictiveness, the ordinary remedy is to bar the augmented charge. This was the remedy used in Blackledge.

In Blackledge, however, the prosecutor did not try to justify his conduct. In this case the government strongly argues that it did not retaliate against the defendant. The district court did not pass on the government's proffered explanations. The court simply ruled that the explanations were legally inadequate. 444 F. Supp. at 1243. We think that the government should be given an opportunity to rebut a finding of realistic likelihood of vindictiveness. In this case we are dealing with a pre-trial bail motion; not, as in Pearce and Blackledge, with post-trial efforts to secure a new trial. Whatever "rebuttal rights" exist in other settings, we think that the government should have a right of rebuttal here.

We emphasize that once a court has found the existence of a realistic likelihood of vindictiveness the burden of disproving it is on the government. For reasons outlined above, we do not think that judges should pass on subjective good faith assertions by prosecutors. Both Pearce and Blackledge went out of their way to avoid such difficult and unpleasant decision-making. At the same time, in Blackledge the Court noted in a footnote that "(t)his would clearly be a different case if the state had shown that it was impossible to proceed on the more serious charge at the outset." Id. 417 U.S. at 29 n.7, 94 S. Ct. at 2103 n.7. From this, we think that only objective, on-the-record explanations can suffice to rebut a finding of realistic likelihood of vindictiveness.10 In this case, the fact that the prosecutor said she made a mistake is not, by itself, sufficient to rebut. However, her claims that the Grand Jury was unavailable and that she was inexperienced do have an objective basis and deserve consideration. If, in light of the prosecutor's explanation, the district court no longer believes that there exists a realistic likelihood of vindictiveness, then the charge will not be barred.

The only remaining question for us to examine is the effect of Bordenkircher v. Hayes, 434 U.S. 357, 98 S. Ct. 663, 54 L. Ed. 2d 604 (1978), on this case. In Bordenkircher, the defendant had been charged with forgery. The state prosecutor threatened the defendant that unless he pleaded guilty to the forgery charge, he (the prosecutor) would bring habitual criminal offender charges which carried a mandatory life imprisonment term. When the defendant refused to plead guilty, the prosecutor lived up to his threat. The defendant was duly convicted of the more severe charges. This court granted the writ of habeas corpus. Hayes v. Cowan, 547 F.2d 42 (6th Cir. 1976). Not only was there a realistic likelihood of vindictiveness, the prosecutor openly admitted that he had acted vindictively. Id. at 43 n.2 and 45. The case thus presented the truly rare situation where actual retaliatory vindictiveness was clear on the record.

The Supreme Court reversed. Bordenkircher v. Hayes, supra. The Court focused on the plea bargaining situation before it and concluded that "there is no such element of punishment or retaliation as long as the accused is free to accept or reject the prosecutor's offer." 434 U.S. at 363, 98 S.Ct. at 668. The Court contrasted the "give and take" of plea bargaining where a defendant could choose to accept or reject a plea bargain with the "unilateral imposition of a penalty upon a defendant who had chosen to exercise a legal right." 434 U.S. at 362, 98 S.Ct. at 667. Significantly, however, the Court stated that "(t)o punish a person because he has done what the law plainly allows him to do is a due process violation of the most basic sort ..." 434 U.S. at 363, 98 S. Ct. at 668.

We think that Bordenkircher must be confined to the plea bargaining context in which it arose. Pearce and Blackledge hold that actual retaliatory behavior is unacceptable under the due process clause. Bordenkircher holds that actual retaliatory behavior is acceptable, at least in the plea bargaining context. An expansive reading of Bordenkircher makes that decision irreconcilable with Blackledge and Pearce. The dissent has no difficulty ruling that Bordenkircher effectively overrules Blackledge.11 Given the narrow language of the Court's opinion in Bordenkircher, we cannot agree. We do not think that we, as a lower federal court, should nullify a doctrine established by the Supreme Court absent far stronger evidence than we have before us that the Court itself would no longer apply that doctrine.

We do not need to discuss any policy arguments regarding prosecutorial vindictiveness. These arguments were the subject of much debate among the members of the panel. We think that deference to the Supreme Court mandates a holding that Blackledge is alive and well outside of a plea bargaining situation. We regard Bordenkircher as a unique case where a prosecutor acted vindictively, but an independent factor-the defendant's ability to choose-eliminated any due process problems.

The district court applied a near per se appearance of vindictiveness standard, i. e. where the prosecutor adds charges after the defendant's exercise of a procedural right, there arises an appearance of vindictiveness which the government has the "heavy burden" to rebut. 444 F. Supp. at 1244. We think that this standard is too harsh and operates to unduly limit prosecutorial discretion. The proper standard, in the opinion of the en banc court is not whether there is an appearance of vindictiveness, but whether there exists a "realistic likelihood of vindictiveness." Only then should the government be required to assume the burden of rebutting with objective facts.

Accordingly, the judgment of the district court is vacated and the case remanded for further proceedings consistent with this opinion. The district court should conduct an evidentiary hearing where the government's explanations can be formally presented and tested.

I concur in the well-written opinion of Judge Keith. I would like to emphasize that on the remand, in addition to conducting an evidentiary hearing, the District Court should adopt findings of fact and conclusions of law on the issues involved and particularly with respect to the reasons given by the prosecution for its failure to charge a conspiracy initially and whether but for the moratorium on the work of the Grand Jury due to challenges to its composition and the scheduling difficulties in the prosecutor's office, the conspiracy evidence would have been presented by the prosecution to the Grand Jury prior to the time the defendants filed their bond motion.

MERRITT, Circuit Judge, dissenting.

The Supreme Court has never applied the concept of "vindictiveness" announced in Pearce and Perry to a prosecutor's actions during the trial and pretrial stages of a criminal case. It declined to do so in Bordenkircher, where the prosecutor's conduct was clearly vindictive. Strong policy considerations lie behind what seem to some to be contradictory conclusions reached by the Supreme Court in those three cases. It seems likely that these strong policy considerations will lead the Court to limit the application of the doctrine of prosecutorial vindictiveness to prosecutorial conduct which occurs after conviction. Even if this view of the direction of the Supreme Court in the application of this doctrine is in error, the standard proposed by our Court, which focuses on the subjective intent of the prosecutor-the "realistic likelihood of prosecutorial vindictiveness"-misses the point and is not workable. Any intelligible standard must focus more clearly on the nature of the constitutional or other legal right which the defendant has asserted and on whether the prosecutor's conduct, without regard to his motives, denies to the accused the effective enjoyment of the particular right asserted.

In Pearce, the Supreme Court held under the due process clause that a state trial judge may not double the punishment imposed at the first trial when the defendant is reconvicted of the same offense after a successful appeal and new trial. In Perry, the Court held that after a defendant is tried for a misdemeanor and appeals his conviction, the prosecutor violates due process at the new trial when he seeks to reindict and try the defendant for a felony for the same criminal event. In both cases, the problem arose because a majority of the Court, in a long line of cases, has historically read the protections of the double jeopardy clause narrowly to allow the state to "up-the-ante" on a criminal defendant at a new trial following an appeal. In Pearce and Perry, a majority of the Court was unwilling to retreat from its prior double jeopardy rulings or to reinterpret the double jeopardy clause along the broader lines suggested in the concurring opinions of Justices Douglas and Harlan in Pearce, 395 U.S. 711, 726, 744, 89 S. Ct. 2072, 2081, 2085, 23 L. Ed. 2d 656 (1969). So in order to remedy the wrong, the Court conceptualized the problem more broadly under the due process clause. It found that "vindictiveness" by the state, or the apprehension of retaliation, is the key principle and declined to base its holding on a double jeopardy analysis. But the cases retain their double jeopardy flavor.

Bordenkircher, it seems to me, suggests that Pearce and Perry are limited to the post-conviction stage of criminal proceedings where the element of finality becomes a consideration and double jeopardy values come into play. It is difficult to find a clearer case of "vindictiveness" than Bordenkircher. There the Court was faced not simply with the appearance of prosecutorial vindictiveness but with the reality. The prosecutor told the defendant that unless he gave up his right to a trial and pled guilty to fraud, he would be reindicted as a habitual criminal, an offense carrying life imprisonment. When the defendant refused to enter such a plea, the prosecutor made good his threat, and the Supreme Court upheld that habitual criminal conviction. It is hard to argue against Justice Blackmun's observation in dissent in Bordenkircher that the Court "is departing from, or at least restricting, the principles established" in Pearce and Perry. 434 U.S. at 365, 98 S.Ct. at 669. In Bordenkircher, the Court itself says that the interest it was concerned about in Pearce and Perry was "the danger that the state might be retaliating against the accused for lawfully attacking his conviction." 434 U.S. at 363, 98 S. Ct. at 668.

It is difficult to reconcile Pearce and Perry with Bordenkircher if we take an expansive view of the concept of "vindictiveness" announced in the first two cases. But the cases become more understandable and consistent if we put aside the concept of general prosecutorial vindictiveness under the due process clause and look at Pearce and Perry as due process cases designed to protect double jeopardy values involving post-conviction "vindictiveness."

In both cases, the state tried to "up-the-ante" by imposing a harsher penalty on retrial after the appeal. They are both cases of successive prosecutions for the same criminal event. Except for the very broad due process "vindictiveness" language used by the Court, the two cases fit much better the double jeopardy mold, for, after all, the double jeopardy clause was designed in part to restrain a certain species of prosecutorial vindictiveness, the kind that leads a dissatisfied prosecutor to want to get the defendant again, or put him away longer, after the first trial is over. In the competitive, adversary environment of the criminal trial, it is not unnatural for a dissatisfied prosecutor to want to try the defendant again, but that is the precise species of vindictiveness the double jeopardy clause is designed to prevent. That is what happened in Pearce and Perry and that, it seems to me, is the only species of vindictiveness which the Supreme Court struck down in those cases.

In Bordenkircher, on the other hand, like the instant case, there was no threat of a second trial, nor any threat of greater punishment upon reconviction, and double jeopardy-type, post-trial vindictiveness did not come into play. In Bordenkircher, the Court found that prosecutorial retribution is acceptable during the preliminary and trial stages of the criminal process before double jeopardy values come into play at the end. It is not, as Judge Keith suggests, the plea bargaining context of Bordenkircher that distinguishes it from other "vindictiveness" claims. It is the double jeopardy context of Pearce and Perry that limit and control the vindictiveness principles announced in those two cases.

The concept of prosecutorial "vindictiveness" or the "likelihood of vindictiveness" is theoretically endless and unmanageable unless limited to the post-conviction, double jeopardy context. Otherwise, the concept is so elusive that nobody-judge, prosecutor or defendant-can determine where it starts or stops. The concept also appears inconsistent with the competitive, adversary nature of the pretrial and trial process.

By necessity the public prosecutor must take hard positions contrary to the liberty interests of the accused. Our society and our jurisprudence have never reached agreement on whether the purpose of the penal system is to deter, to treat or to avenge. Public prosecutors, like the public at large, are sometimes motivated by a desire for retribution in criminal cases. Whether he is motivated by deterence or retribution or some more complicated but less explicable state of mind, the prosecutor's attitude toward the defendant in a hard-fought criminal case is seldom benign or neutral.

During the pretrial and trial process, the prosecutor must decide what position to take on an endless variety of procedural, evidentiary, substantive and tactical questions. He may oppose motions to suppress evidence or for the appointment of counsel or refuse to agree to discovery, severance, bail, or plea bargaining; he may try to get into evidence prior criminal conduct or various co-conspirator and other kinds of hearsay; he may be harsh in his characterization of the defendant's conduct to the jury; he may recommend probation or refuse to prosecute altogether; or he may make a deal with a co-defendant in exchange for incriminating testimony and on and on.

Once a defendant has successfully asserted a particular legal right in the course of the criminal process over the prosecutor's objection, is the prosecutor arguably guilty of unconstitutional vindictive conduct, which "chills" the exercise of the legal right asserted, each time the prosecutor thereafter takes a position contrary to the interests of the defendant? If not, why not, and what is the standard of measurement? What difference does it make that the prosecutor's conduct took place after rather than before the defendant asserted the right? The "exercise" of a legal right can be more effectively "chilled" before it is asserted than after. What difference should it make that the defendant was unsuccessful rather than successful in asserting the legal right? On the facts of this case, would it make any difference that the defendant lost his motion for bail rather than won it? If we are talking about the "exercise" of a legal right, should it make a difference that its exercise happened to be unsuccessful in the particular case?

Here retaliatory motives, conscious or unconscious, probably played a part in the prosecutor's decision. If the defendant had consented to stay in jail, the prosecutor might not have added a conspiracy count. But the prosecutor believed that the defendants had threatened to harm a government witness. The prosecutor retaliated by opposing bail. Having lost on that issue, the prosecutor probably added the conspiracy count for reasons similar to those that led the government to put the witness under the Federal Protection Act-in order to show the defendants that the government intended to fight harder to protect its witness and get a conviction now that the defendants were back on the street. The fact that the prosecutor may have retaliated in this situation only indicates to me diligence in the performance of the prosecutorial role.

Many constitutional, common law and evidentiary principles protect the accused and should be given an expansive meaning. But to require, as a general principle, the general absence of "prosecutorial vindictiveness" in the heat of battle is an unworkable goal in our adversary system. It seems impossible to isolate and eliminate motives of prosecutorial retribution in a process that is itself based, in part, on a theory of retribution.

For these reasons I would hold that the doctrine of prosecutorial vindictiveness announced in the Pearce and Perry cases, and rejected in Bordenkircher, is limited to prosecutorial behavior after the first trial is over. That means that in the instant case the doctrine is inapplicable, and I would therefore reverse the judgment of the District Court.

BOYCE F. MARTIN, Jr., Circuit Judge, dissenting.

I join in Judge Merritt's dissent. From the conflicting concurrences and dissents that have been filed it is obvious that the Court is widely divided on the issue we face. To me the facts of this case are very simple. As outlined in the decision of the District Court at 444 F. Supp. 1238 (E.D.Michigan 1978), and in the majority opinion, we are faced with a single question of prosecutorial error. No one has alleged that the additional conspiracy count was any more than a correction on the part of an inexperienced Assistant United States Attorney. To characterize this prosecutorial error as vindictiveness is to me unwarranted. Nowhere in Blackledge v. Perry, 417 U.S. 21, 94 S. Ct. 2098, 40 L. Ed. 2d 628 (1974), and North Carolina v. Pierce, 395 U.S. 711, 89 S. Ct. 2072, 23 L. Ed. 2d 656 (1969), do I find support for the conclusion reached by the District Court and now confirmed by a majority of this Court.

As Judge Merritt pointed out, our system of government vests the United States Attorney with wide discretion in handling criminal charges. So long as he possesses facts sufficient to convince a Grand Jury that a crime has been committed and that the charged defendant has committed that crime, he may proceed to seek an indictment. Once charged, the prosecutor's responsibility becomes more difficult. Guilt of the charge must be proven beyond a reasonable doubt. The United States Attorney may attempt to charge an alleged offender as often as he reasonably can.

As Judge Merritt points out so well in Part II of his dissent, the majority has lost sight of the role of a public prosecutor in our society. To impose an additional hearing upon the District Court each time a superseding indictment is filed does nothing more than further defeat the ends of justice. There are many other ways to harness an overzealous prosecutor.

I would reverse the decision of the District Court, reinstate the second complete indictment including the conspiracy count, and direct that the defendants be tried on the charges as quickly as possible.

CORNELIA G. KENNEDY, Circuit Judge, dissenting.

I am unable to concur in either Judge Keith's opinion for the majority or Judge Merritt's dissent. I cannot join the majority since I would find that the prosecutor's stake in preventing a defendant's release on bond is so slight that where it is the only fact before the Court tending to show vindictiveness, the Court cannot find a realistic likelihood of vindictiveness (which I would define to be an improper retaliation for the exercise of a constitutional right). Particularly where, as here, if any vindictiveness or retaliatory motive prompted the prosecutor to add a count, it could equally be due to the belief that defendant had threatened witnesses and should therefore be charged to the fullest extent possible, rather than because defendant had sought bond. This is retaliation for a threat against the witnesses; it is not retaliation for the exercise of the right to bond, and it is a type of retaliation that I think is a proper exercise of prosecutorial discretion. The fervor with which a prosecutor opposes bond should not be a factor since it is a prosecutor's obligation to perform vigorously.

My judgment is tempered by the fact that the prosecutor is only one player in the drama leading up to trial. The grand jury exists to protect against prosecutorial abuse. Whatever the prosecutor's motive in bringing or substituting charges, the defendant is protected from charges that lack a probable cause basis.

Further, I would hold that the prosecutor could in all instances refute allegations of vindictiveness. See Jackson v. Walker, 585 F.2d 139, 148 (5th Cir. 1978). The majority has expressed concern that the Court should not be placed in the undesirable position of having to determine the actual motive of the prosecutor in a particular instance. However difficult the task of determining the credibility of a prosecutor and the unseemliness of finding a prosecutor untruthful, dismissal with prejudice of a charge for which a grand jury has returned an indictment is such a serious sanction that it should not be imposed where the prosecutor was guilty of nothing more than oversight or inadvertence.

Moreover, I would limit our holding to circumstances similar to those before us. The instant case involves bond set after the indictment. Bail is also often set following arrest on a complaint and warrant or on probable cause. Would the majority hold that the addition of charges after the setting of that initial bail also requires objective on-the-record explanation? Would the majority confine its holding to the constitutional right to bail? If an indictment is dismissed for failure to properly advise the defendant of the charge or a count is substituted because a defendant has prevailed on a motion to suppress evidence required to support some charge, is the prosecutor foreclosed from substituting a charge? In such instances, the prosecutor might well admit that the addition or substitution of these charges resulted from the defendant's exercise of a constitutional right. Surely such retaliation on the part of the prosecutor is permissible, even expected; what we are called upon to do is define exactly when retaliation by the prosecutor is impermissible.

For that we look to the due process clause of the fifth amendment. The prosecutorial conduct that we must proscribe under that clause is conduct which is fundamentally unfair. We must examine the prosecutor's actions while mindful of the fact that we expect prosecutors to exercise broad discretion in deciding when to charge a particular defendant, what charges to bring against him, whether to deal with him through the plea bargaining process, etc. It is inevitable when a prosecutor is exercising this much discretion that much of what he does can be seen as vindictive or retaliatory for the defendant's exercise of some right. A test for prosecutorial retaliation that too easily lets a defendant challenge prosecutorial conduct will have the undesirable effect of curbing prosecutorial discretion to a significant degree. For example, as a prophylactic measure, the prosecutor may be forced to bring every charge he conceivably can against each defendant at the initiation of the proceedings to avoid later becoming embroiled in a "retaliation" controversy. Breakdowns are inevitable in the administration of a busy prosecutor's office. The majority's opinion would make bringing legitimate charges after the initial proceedings extremely difficult.

I do not think that this is a good or a necessary result. I would find very few cases in the pre-trial stage in which the prosecutor's conduct appeared impermissibly to retaliate for the defendant's exercise of a constitutional right, such that the prosecutor should be called upon to explain his action. As I stated, I would not require explanation here.

One of my chief concerns with the majority opinion is that it provides no guidance to the district courts. The Court does not even say whether it would find impermissible retaliation on the facts alleged in this case. After considerable thought, I can only conclude that the phrase "realistic likelihood of vindictiveness" will have to be given content on a case-by-case basis. I am unable to produce, and I do not think that the majority has produced, a test that is useful in general application.

Judge Merritt's bright line rule is attractive for its pragmatic advantages. However, tempting as it is, I am not prepared to take the position that a prosecutor can never be impermissibly retaliatory before trial. I conceive of situations, not present here, where I would find that retaliatory conduct by a prosecutor violates due process (for example, explicit verbalized threats that should a defendant seek bond more serious charges would be substituted or added, or possibly a retaliatory pattern or practice of adding or substituting more serious charges).

I would, therefore, reverse the judgment of the district court and order reinstatement of the count of the indictment.

I respectfully dissent. I would remand for reinstatement of the dismissed count.

The remedy applied by the majority is more drastic than that provided by the exclusionary rule, for it would exclude not merely evidence of crime but the government's right to seek prosecution on the charge itself. It is a charge which a duly constituted grand jury has found to have been supported by probable cause and of which Andrews may actually be guilty.

There is no suggestion in the record here that the grand jury was not properly constituted, or that the indictment was not proper on its face or that it was procured by fraud. It is true that case law suggests such a right where there is evidence of selective prosecution, Yick Wo v. Hopkins, 118 U.S. 356, 6 S. Ct. 1064, 30 L. Ed. 220 (1886). However, no racial, religious or other invidious class-based discrimination is even hinted at here. Oyler v. Boles, 368 U.S. 448, 456, 82 S. Ct. 501, 505, 7 L. Ed. 2d 446 (1962).

The majority's decision here represents an intrusion into an area historically reserved to the executive branch, and which in my view can be justified only by concerns more fundamental and conduct more egregious than are present here. The Constitution itself provides the necessary checks and balances to guard against prosecutorial vindictiveness, real or imagined, by interposing the grand jury system between persons accused of crime and their accusers:

The institution of the grand jury is deeply rooted in Anglo-American history. In England, the grand jury served for centuries both as a body of accusers sworn to discover and present for trial persons suspected of criminal wrongdoing and as a protector of citizens against arbitrary and oppressive governmental action. In this country the Founders thought the grand jury so essential to basic liberties that they provided in the Fifth Amendment that federal prosecution for serious crimes can only be instituted by "a presentment or indictment of a Grand Jury." Cf. Costello v. United States, 350 U.S. 359, 361-62, 76 S. Ct. 406, 407-408, 100 L. Ed. 397 (1956). The grand jury's historic functions survive to this day. Its responsibilities continue to include both the determination whether there is probable cause to believe a crime has been committed and the protection of citizens against unfounded criminal prosecutions. Branzburg v. Hayes, 408 U.S. 665, 686-687, 92 S. Ct. 2646, 2659, 33 L. Ed. 2d 626 (1972).

Traditionally the grand jury has been accorded wide latitude to inquire into violations of criminal law. No judge presides to monitor its proceedings. It deliberates in secret and may determine alone the course of its inquiry. The grand jury may compel the production of evidence or the testimony of witnesses as it considers appropriate, and its operation generally is unrestrained by the technical procedural and evidentiary rules governing the conduct of criminal trials. "It is a grand inquest, a body with powers of investigation and inquisition, the scope of whose inquiries is not to be limited narrowly by questions of propriety or forecasts of the probable result of the investigation, or by doubts whether any particular individual will be found properly subject to an accusation of crime." Blair v. United States, 250 U.S. 273, 282, 39 S. Ct. 468, 471, 63 L. Ed. 979 (1919).

The scope of the grand jury's powers reflects its special role in insuring fair and effective law enforcement. A grand jury proceeding is not an adversary hearing in which the guilt or innocence of the accused is adjudicated. Rather it is an ex parte investigation to determine whether a crime has been committed and whether the criminal proceedings should be instituted against any person. The grand jury's investigative power must be broad if its public responsibility is adequately to be discharged. Branzburg v. Hayes, supra, at 700, 92 S. Ct. at 2666; Costello v. United States, supra, at 364, 76 S.Ct. at 409.

United States v. Calandra, supra, 414 U.S. at 342-44, 94 S. Ct. 613, at 617-18, 38 L. Ed. 2d 561 (footnotes omitted).

I would not rule out the possibility that the problem perceived here might become so pervasive or systemic as to invoke the protection of the Due Process Clause. I am therefore unable to agree with my colleague, Judge Merritt, that Pearce and Blackledge are necessarily limited to post-conviction proceedings or to cases where double jeopardy values are implicated. Neither am I, however, willing to confine Bordenkircher to its facts, as would my colleague, Judge Keith. I believe that we should consult the wisdom of both, but that we are bound in this case by neither.

The right allegedly chilled here is the Eighth Amendment right to bail. It is a right which is purely collateral to the right of the defendant to a fair trial. It is a precious right to be sure, but a limited one. It has nothing to do with guilt or innocence. It is also a right whose protection already has a plain, speedy, efficient remedy.

The right to bail is not unqualified. The Eighth Amendment only forbids "excessive bail." It has never been thought to accord a right to bail in all cases, but merely to provide that bail shall not be excessive in those cases where it is proper to grant bail. Carlson v. Landon, 342 U.S. 524, 545-46, 72 S. Ct. 525, 536-537, 96 L. Ed. 547 (1952). Thus while a defendant has an unquestionable right under the Eighth Amendment and a statutory right under 18 U.S.C. § 3146(d) to petition the court for release in non-capital cases, bail may properly be denied altogether if the defendant has threatened the lives of the witnesses against him, a circumstance at least suggested at the bond determination proceedings here. Mr. Justice Douglas recognized the need to deny bail on such occasions:

In my view the safety of witnesses, should a new trial be ordered, has a relevancy to the bail issue. Keeping a defendant in custody during the trial "to render fruitless" any attempt to interfere with witnesses or jurors may, in the extreme or unusual case, justify denial of bail. (Citations and footnotes omitted).

Justice Douglas, sitting as Circuit Justice in Carbo v. United States, 82 S. Ct. 662, 668, 7 L. Ed. 2d 769, rehearing denied, 369 U.S. 868, 82 S. Ct. 1137, 8 L. Ed. 2d 274 (1962). To like effect in our circuit, see United States v. Wind, 527 F.2d 672 (6th Cir. 1975); United States v. Bigelow, 544 F.2d 904 (6th Cir. 1976). See also, Mastrian v. Hedman, 326 F.2d 708 (8th Cir. 1964), cert. denied 376 U.S. 965, 84 S. Ct. 1128, 11 L. Ed. 2d 982.

While the underlying facts in Bordenkircher v. Hayes sufficiently differ from those here to persuade me that it is not controlling, still its logic and language are highly persuasive. The chilling effect of the allegedly vindictive conduct in Pearce, Blackledge and Hayes was upon the defendant's right to contest by trial the truth of the charges against him. If there are different levels of sensitivity to different constitutional rights, surely those which affect truth-finding are foremost, and yet Bordenkircher v. Hayes clearly tolerates such a chilling effect when necessary to preserve the perceived benefits of plea bargaining. In this sense, at least, I find some merit in the government's view that the government's stake in whether a defendant is free on bond pending trial is not as great as it was in Blackledge, and hence that due process does not require the rule of Blackledge or Pearce here. The majority opinion acknowledges this to some extent but concludes that each case must turn on its facts.

In short, I am not convinced that a Supreme Court which decided Pearce, Blackledge and Hayes would exact such a sanction here where the exercise of the right is less likely to affect the prosecution's vital interest and thus invite abuse.

The majority opinion assumes that a defendant will be less likely to seek bail if he believes that by so doing the prosecution will retaliate by increasing the charges against him. I respectfully suggest that while such a result is possible, reality says he will not be chilled. While the majority's concern is with the prospective chilling effect upon future defendants, it is still noteworthy that Andrews and Brooks were in fact admitted to bail and that the prosecutrix did not seek to revoke it after the superceding indictment was returned.

In Chaffin v. Stynchcombe, 412 U.S. 17, 29, 93 S. Ct. 1977, 1984, 36 L. Ed. 2d 714 (1973), Justice Powell carefully considered the "chilling effect" in determining whether the Due Process Clause should bar a jury from rendering a higher sentence on re-trial following a reversal of a prior conviction. There, speaking for the majority, Justice Powell held that jury sentencing was, like plea bargaining, a legitimate practice not barred merely because it might have a "discouraging effect on the defendant's assertion of his trial rights." 412 U.S. at 31, 93 S.Ct. at 1985. At the same time he realistically analyzed the "chill factor" from the viewpoint of the defendant. He concluded, as I would here, that "we doubt that the 'chill factor' will often be a deterrent of any significance." 412 U.S. at 33, 93 S.Ct. at 1986. The realities here pose, in my judgment, a deterrent of even lesser significance.

First, to be chilled, the defendant, in hesitating to seek bail, must anticipate that if he is successful, the prosecutor will seek additional, or enhanced, charges. Next, he must anticipate that the grand jury will find probable cause and return an indictment. Then he must anticipate that in trial the jury will convict him of the additional charge. Finally, he must anticipate that having been found guilty, the judge will give him a more severe sentence because of the additional count in the indictment. It is true that competent defense counsel may advise him of these possibilities, much as he would advise the defendant when deciding whether to plead or stand trial, to testify or to remain silent, to cooperate with the police or to follow any number of alternative courses of action. I seriously doubt, however, that this possible consequence, so far down the road, will actually chill and thus deter his action, when measured against the more immediate benefit of freedom pending trial.

On the other hand, while the compelling desire for immediate freedom will no doubt be the strongest instinct in the defendant, there is at least one other strongly persuasive consideration of great strategic appeal to defendant's counsel which can reasonably be expected to offset the fear of heavier charges. Every defense lawyer knows that each day of freedom for the client, accompanied by demonstrated capability for reform, enhances the chances for ultimate freedom or at least adds to the weight of evidence that the defendant is not a probation risk should he be convicted. Thus it is altogether realistic to expect that the defendant's counsel will urge that defendant seek bail even at the risk of retaliation. The long period of time which normally exists between arrest and ultimate sentence, if spent on bail, provides the defendant the most compelling argument for a lighter sentence if he can, by obtaining employment, supporting his family, and engaging in therapy, persuade the sentencing judge that he is a reformed man and that it would be both unnecessary and unprofitable to return him to confinement.

My judicial instinct says that almost invariably the defendant will opt for immediate freedom, if he can get it. He will not be chilled or deterred from seeking bail by the implied threat of additional charges. He will take the cash and let the credit go, and his attorney will in all likelihood recommend that course. Justice Powell's observation in Chaffin have even more compelling logic here.

Apart from whether a defendant is likely to be chilled or discouraged from seeking admission to bail by the threat of vindictiveness, a further question is whether the prosecutor is likely to be deterred from such a course of conduct by the prophylactic rule of dismissal. He may. A more likely result, however, is that the proposed rule will have the undesired effect either of blunting prosecutorial discretion by fostering timidity, or of providing a strong incentive to prosecutors to "throw the book" at a defendant at the first opportunity as the safest way to avoid a charge that a later addition of counts was vindictive. Either reaction is less than desirable. For every defendant whose rights the majority may wish to protect, there will be many others who stand to be prejudiced by the almost inevitable effect of the rule which the majority would create.

The dilemma which the majority rule poses for the prosecutor is real indeed. If he honestly believes his own motives are free from vindictiveness, how may they appear to the trial judge who must apply an objective standard? How will the court of appeals view it? Why should he risk such accusations and enmesh himself in collateral evidentiary hearings when he can avoid it all by charging the defendant to the hilt at the outset?

The record in this case provides a classic example of the dilemma. There was testimony at the bond revocation hearing that a confidential source had reported that the defendant had threatened a witness. It does not appear whether the trial judge credited this testimony for, though he heard it, he made no finding with respect to it. We do not know if in fact the defendants threatened the witnesses, or if the confidential informer was reliable, or if the prosecutor completely believed him. Nevertheless, this intelligence had to create in the prosecutrix's mind some apprehension of danger to the witness and to the prosecution of her case.

The government has also claimed that the Assistant United States Attorney did not originally seek to indict on a conspiracy because of her inexperience. It contends that she would in all events have sought the new indictment before the defendants, appealed denial of bail but could not because the composition of the grand jury itself was then under challenge and because, after that problem was resolved, her scheduled time before the grand jury had been pre-empted by higher ranking attorneys in her office. Although the trial judge expressly declined to find her action to be malicious or to question her integrity, he made no findings concerning the truth of these claims.

The facts here demonstrate as well as any others the problem of attempting to draw the fine line between what is vindictiveness or retaliation and what is righteous indignation or an appropriate discretionary response to a perceived danger. It is the essence of discretion that it may properly be exercised in different ways and likewise appear differently to different eyes. The discretion to charge persons accused of crime is the function of the executive branch acting through the prosecutor, historically protected from abuse by the interposition of the grand jury. There are undoubtedly many occasions when the exercise of that discretion may subject the executive branch to the danger of appearing vindictive or arbitrary or unjust. For the most part, however, correction should lie in the political process to which the executive is subject. Correction in the area of executive discretion should only rarely be a judicial function.

Chief Justice Burger addressed this precise question as a circuit judge in Newman v. United States, 382 F.2d 479 (D.C.Cir.1967):

An attorney for the United States, as any other attorney, however, appears in a dual role. He is at once an officer of the court and the agent and attorney for a client; in the first capacity he is responsible to the Court for the manner of his conduct of a case, i. e., his demeanor, deportment and ethical conduct; but in his second capacity, as agent and attorney for the Executive, he is responsible to his principal and the courts have no power over the exercise of his discretion or his motives as they relate to the execution of his duty within the framework of his professional employment.

* * *

To say that the United States Attorney must literally treat every offense and every offender alike is to delegate him an impossible task; of course this concept would negate discretion. Myriad factors can enter into the prosecutor's decision. Two persons may have committed what is precisely the same legal offense but the prosecutor is not compelled by law, duty or tradition to treat them the same as to charges. On the contrary, he is expected to exercise discretion and common sense to the end that if, for example, when one is a young first offender and the other older, with a criminal record, or one played a lesser and the other a dominate role, one the instigator and the other a follower, the prosecutor can and should take such factors into account; no court has any jurisdiction to inquire into or review his decision.

It is assumed that the United States Attorney will perform his duties and exercise his powers consistent with his oath; and while this discretion is subject to abuse or misuse just as is judicial discretion, deviations from his duty as an agent of the Executive are to be dealt with by his superiors.

The remedy lies ultimately within the establishment where power and discretion reside. The President has abundant supervisory and disciplinary powers-including summary dismissal-to deal with misconduct of his subordinates; it is not the function of the judiciary to review the exercise of executive discretion whether it be that of the President himself or those to whom he has delegated certain of his powers.

382 F.2d 481-82 (footnotes omitted).

The application of the foregoing logic to the facts here is plain indeed. Even if she misconceived the danger which the defendants might pose to government witnesses, it was not surprising for the prosecutrix to desire to avoid that risk by opposing bail or by seeking the additional charge if she believed it was warranted. As Newman teaches, a proper exercise of prosecutorial discretion necessarily includes a personal, even subjective, evaluation of culpability in deciding what charges to seek from the grand jury. If public policy of lenity in seeking grand jury indictments is to be encouraged, there must be some room for further repair to the grand jury where the initial impression of the defendant's conduct has proved too optimistic. A prosecutor may be quite willing to believe the law is adequately vindicated by one or two substantive drug charges but may reasonably decide it does not when it later appears that the defendants are willing to threaten or harm witnesses to avoid conviction.

My most fundamental objection is that the pre-trial procedures mandated by the majority opinion are wasteful of time, will unnecessarily delay the trial and are altogether out of proportion to the problem which is perceived.

If there were present here some genuine issue of selective prosecution based upon racial or religious or other invidious classwide bias, or some question of fraud in the procurement of the indictment, I could join in the creation of a rule to prevent it, however difficult it might be of application. Even the problem of defining vindictiveness should not deter us from creating the rule, if it is otherwise justified. However, we should not embark on such a course without first measuring its need and effect in terms of the overall impact such a rule would have upon the administration of justice generally. The truth is, however, that the majority opinion creates needless complication and further uncertainty in the already overburdened processes leading from arrest to trial.

Defense counsel, to avoid a charge of ineffective assistance, will be induced to challenge every superseding indictment for fear of waiving any claim that it was procured because of vindictiveness. Where not raised by the defense, the trial court, fearful of the plain error rule or of later collateral attack, may raise one issue sua sponte.

Applying the rule will be difficult in practice. For all practical purposes, I am unable to distinguish the standard of "appearance of vindictiveness" applied by the district court from "reasonable likelihood of vindictiveness" where the trial judge, following Blackledge, must apply an objective standard. I can see little or no difference between the "heavy burden" which the district court said was upon the government and the majority rule which somehow rules out "mistake," and requires "only objective, on-the-record explanations." In my opinion, an "appearance of vindictiveness" means no more nor less than "realistic likelihood of vindictiveness" judged on an objective basis.

If the defendants prevail following the hearing as here, the government has a right of appeal under 18 U.S.C. § 3731. The result, regardless of success, is delay. As this opinion is written, nearly five years have transpired since the alleged offenses and arrest of Tallice Andrews and Thurston Brooks in November, 1975. It is true that not all delays are due to this litigation, but it has since March, 1978 played a substantial role in that delay. The majority rule places a heavy burden upon district courts without sufficient promise of benefit to justify it.

While a corresponding right to interlocutory appeal by the defendant probably does not exist where the trial judge denies such motions, there remains the serious problem of diversion of judicial and legal resources to determine the issue, with its own concomitant delay. More important, there remains the uncertainty which the issue interjects into the trial and into the appeal which, with mechanical certainty, now follows nearly all convictions. The issue must be reviewed. Three judges of the court of appeals must consider whether the trial judge's factual findings are clearly erroneous; if not, whether he abused his discretion; and finally, whether he applied the correct standard of law in a field where substantial disagreement obviously exists.

If the court of appeals concludes that the count should have been dismissed, there remains the further question whether trial on the excluded count, with its different evidence and instructions, so prejudiced the defendants that they must be retried on the remaining counts. Those who may view this possibility as theoretical need only to look to the original panel decision in United States v. Sutton, 605 F.2d 260 (6th Cir. 1979), rehearing en banc pending, which would have reversed all convictions because of what the majority conceived was the transferred prejudice of the impermissible count. The problem is real indeed for everyone at the trial level, trial judge, prosecutor and defense counsel.

The defendants already have fully adequate remedies to protect their right to bail. These remedies should largely dissipate any effect of truly unjust retaliatory action by the prosecutor.

Part I of this dissent points out that the defendants have the constitutional protection provided by the grand jury system to stand between them and their accusers. That system guarantees that, regardless of the motives of the prosecutor, the charges against them will be supported by probable cause before they must face them. Thereafter, the defendants can seek dismissal if the proofs are insufficient to support the charge. The jury, being unpersuaded, can acquit them. Finally, if the defendants are convicted, the Court itself stands as an effective shield against the effects of any perceived prosecutorial vindictiveness.

Likewise the defendants have a speedy and effective remedy if the prosecutor should seek to increase bail because of the increased charges. There is some significance, at least, in the fact that the prosecutrix did not do so here and that so far as this record shows, defendants remain free on bail. Defendants quickly appealed from the magistrate's remand without bail, were granted a hearing, and received prompt relief from the district court. 18 U.S.C. § 3147 (a) provides for district court review of an order of detention. Section 3147(b) provides as well for expedited appellate review. A denial of a motion to reduce bail under section 3147(a) is a final decision for purposes of immediate appeal. 28 U.S.C. § 1291. Stack v. Boyle, 342 U.S. 1, 72 S. Ct. 1, 96 L. Ed. 3 (1951).

Existing statutory provisions protecting the rights to bail are both speedy and effective, as shown by the facts of this case. They accord a fair hearing at which any impermissible motive of the prosecutor can be challenged. Speedy appeal is provided for and does not interrupt the orderly advance of litigation to trial. Finally, the trial does not need to proceed under the apprehension that error in determining the motion to quash may render the validity of the ultimate verdict uncertain.

I would vacate the order of the district court and remand for reinstatement of the conspiracy count.