Appeal from the United States District Court
for the Central District of Illinois. No. 93 CR 10059--Michael M. Mihm, Chief Judge.
Before Posner, Chief Judge, and Rovner and Diane P. Wood, Circuit Judges.
After we reversed Catton's false-claim conviction (18 U.S.C. sec. 287) because of trial errors and remanded for a new trial, 89 F.3d 387 (7th Cir. 1996), he moved to dismiss the case on the ground that retrial would violate the Fifth Amendment's double jeopardy clause. He appeals from the denial of his motion, arguing that the prosecutor suborned perjury, and concealed exculpatory evidence, in order to stave off a looming acquittal. Citing Oregon v. Kennedy, 456 U.S. 667 (1982), and United States v. Wallach, 979 F.2d 912, 916 (2d Cir. 1992), he urges us to hold that a defendant victimized by such tactics should not have to stand trial a second time.
Catton is a farmer who was prosecuted for having submitted false loss claims under a federal crop insurance program. The prosecution presented evidence that the loss of Catton's hybrid seed corn was due not to drought, as Catton had represented in his claims, but to his having deliberately ruined the crop. The evidence against Catton included testimony by an expert witness, Hale, who reported that he had spoken to an employee of Shissler Seed Company, a grower of hybrid seed corn in Catton's area. This employee had given Hale, according to Hale's testimony, information concerning Shissler's average yields of hybrid seed which showed that Shissler's crop had not been damaged by drought during the growing season in which Catton had lost his crop. The prosecutor invited the jury to infer that Catton's loss could not have been due to drought, since Shissler was farming the same area.
In fact Hale had spoken to nobody at Shissler. He had gotten the information about Shissler's yields from an inspector for the Department of Agriculture, Brinkman, who knew but did not tell Hale that Shissler had submitted loss claims, based on drought, for acreage that, so far as appears, was as close to Catton's as the acreage about which Hale testified. The bona fides of Shissler's claims had not been questioned, and the claims had been paid--and this in the same growing season to which Catton's claims pertained. Had Hale been truthful on the stand, and revealed Brinkman's role, Catton's lawyer would have questioned Brinkman and learned about Shissler's drought claims. We thought this concealment so material in a close case that in conjunction with the prosecutor's having in closing argument misrepresented a key part of Catton's testimony it entitled Catton to a new trial.
Catton now argues that the prosecutor concealed Brinkman's role because he knew that if the jury learned about the Shissler loss claims it would have acquitted him. The Supreme Court held in the Kennedy case that if the prosecutor goads the defense into moving for a mistrial, and the mistrial is granted, the double jeopardy clause bars a retrial. 456 U.S. at 676. The need for such a rule is easily seen. Suppose the trial is going very badly for the prosecutor. He anticipates an acquittal. To stave off the acquittal and thus be able to retry the defendant, the prosecutor commits an error in the hope that the defendant will move for a mistrial and that the motion will be granted. A prosecutor could try a defendant over and over again, inducing mistrials every time until finally he had enough evidence to have a good shot at a conviction. This is what Kennedy blocks.
But now suppose that the defendant's motion for a mistrial is denied, he is convicted, and his conviction is overturned on appeal. Kennedy does not speak directly to that case. But if retrial is permitted, the prosecutor will be better off than if the defendant had been acquitted at the first trial--and better off because of prosecutorial misconduct. So it can be argued that Kennedy should be extended to that case, although as yet no court has had to decide whether to make the extension. Greyson v. Kellam, 937 F.2d 1409, 1413-15 (9th Cir. 1991).
There is an argument for a further extension of Kennedy that would bring Catton's case within the range of the double jeopardy clause. Confined to cases in which the defendant is goaded into moving for a mistrial, whether the motion is granted or denied, Kennedy would leave a prosecutor with an unimpaired incentive to commit an error that would not be discovered until after the trial and hence could not provide the basis for a motion for a mistrial, yet would as effectively stave off an acquittal and thus preserve the possibility of a retrial. Suborning perjury would be a good example. It can be argued that if the prosecutor commits a covert error for the same purpose that he might have committed an open error calculated to evoke a motion for a mistrial (before Kennedy made this tactic unprofitable)-- namely, to prevent an acquittal and so preserve the possibility of retrying the defendant even if the error is sure to be discovered and result in a reversal of the conviction either on direct appeal or on collateral attack--the double jeopardy clause should protect the defendant against being retried. Wallach does not hold that the argument is sound, but in a considered dictum concludes that it may well be sound. See also United States v. Pavloyianis, 996 F.2d 1467, 1473-75 (2d Cir. 1993); United States v. Gary, 74 F.3d 304, 314-15 (1st Cir. 1996); State v. Colton, 663 A.2d 339, 346-48 (Conn. 1995); contra, State v. Swartz, 541 N.W.2d 533, 538-40 (Iowa App. 1995).
So at argument we asked the prosecutor whether there was a principled distinction between the open error, which might lead to a mistrial, and the covert error not discovered till after trial. He could not think of any. He could have pointed to language in Kennedy and other cases to the effect that, as we put it in United States v. Oseni, 996 F.2d 186 (7th Cir. 1993), the only prosecutorial intent that is relevant to double jeopardy is "intent to terminate the trial, not intent to prevail at this trial by impermissible means." Id. at 188. "It doesn't even matter that he knows he is acting improperly, provided that his aim is to get a conviction." Id.; see also Oregon v. Kennedy, supra, 456 U.S. at 676; United States v. Jozwiak, 954 F.2d 458, 460 (7th Cir. 1992). But this language, like that in Beringer v. Sheahan, 934 F.2d 110 (7th Cir. 1991), does not have reference to the case in which the prosecutor does not expect to prevail at this trial--the case in which he knows that his misconduct is likely to be discovered and that if it is discovered the verdict will be set aside either on direct appeal or, later, in a collateral attack on the conviction-- and what he is seeking to obtain by committing a reversible error is the opportunity to retry a defendant who but for the error would be acquitted. In such a case, the prosecutor's ultimate aim is not to obtain a conviction at this trial but to obtain a conviction at a subsequent trial, and that was not a consideration in the cases that we have just been citing.
Yet it would be a great burden on the courts if every reversal traceable to a prosecution-induced error at trial gave rise to a Kennedy-style inquest on the prosecutor's motives; and it is possible to read Kennedy as merely carving a narrow exception to the rule that by moving for a mistrial a defendant waives his defense of double jeopardy to a retrial. And so we have left open the question whether to adopt Wallach's dictum as the law of this circuit, United States v. Doyle, 121 F.3d 1078, 1085 (7th Cir. 1997), as has the Eighth Circuit. Jacob v. Clarke, 52 F.3d 178, 182 (8th Cir. 1995). We need not bite the bullet in this case either. For it is clear that a defendant who wants the district court (or this court on appeal from an adverse ruling by the district court) to block a retrial on the basis of prosecutorial error must show that the prosecutor committed the error because he thought that otherwise the jury would acquit and he would therefore be barred from retrying the defendant. It is not enough that there was an error; it is not enough that it was committed or procured by the prosecutor; it is not enough that it was deliberate prosecutorial misconduct; it must in addition have been committed for the purpose of preventing an acquittal that, even if there was enough evidence to convict, was likely if the prosecutor refrained from misconduct. United States v. Wallach, supra, 979 F.2d at 916. Any greater extension of Kennedy must be left to the Supreme Court, in view of the danger of adding a double jeopardy tail to every appellate-reversal dog.
When after his conviction and before his first appeal Catton moved for a new trial on the basis of the just-discovered falsity of Hale's testimony and the prosecutor's failure at the trial to blow the whistle on Hale, the judge conducted an evidentiary hearing. At this hearing the prosecutor testified that when he had asked Hale on the stand whether he had "been in contact with . . . a commercial seed grower by the name of Schissler [sic]," he hadn't meant to suggest that the contact had been direct and so he had not considered Hale's answer false. He explained that because Hale was a competitor of Shissler, Hale had thought that Shissler would refuse to give him the information he was seeking, so he needed an intermediary, and that was Brinkman. The prosecutor also pointed out that before the trial he had furnished Catton's lawyer with a computer printout showing loss claims made and allowed (including drought claims) in Catton's county during the relevant year, and that two of Shissler's drought claims were on the list. Only they ...