Summary
In Cervantes-Pacheco, the Fifth Circuit reversed a conviction for conspiring to possess and import marijuana on due process grounds, because of the government's use of a contingent fee arrangement under which "the quality of the informant's [trial] testimony determined not only whether he would be paid but also how much he would be paid."
Summary of this case from U.S. v. SwiatekOpinion
No. 84-2687.
September 5, 1986. Rehearing En Banc Granted September 25, 1986.
Melvyn Kessler, Miami, Fla., for Pacheco and Jerry Wayne Nelson.
Oscar J. Pena, Laredo, Tex., for W. Nelson.
Susan L. Yarbrough, James R. Gough, Asst. U.S. Attys., Henry K. Oncken, U.S. Atty., Houston, Tex., Ann T. Wallace, Atty., Appellate Section, Crim. Div., U.S. Dept. of Justice, Washington, D.C., for U.S.
Appeals from the United States District Court for the Southern District of Texas.
Before GOLDBERG, WILLIAMS and DAVIS, Circuit Judges.
(Opinion July 7, 1986, 5th Cir. 1986, 793 F.2d 689)
The court substitutes this opinion for its earlier opinion reported at 793 F.2d 689 (5th Cir. 1986).
This is a criminal appeal involving the characterization of a conspiracy and a claim that the Government hired and compensated its main witness on an impermissible contingent fee basis. We dispose of the conspiracy issue in summary fashion but conclude that the contingent fee claim is a substantial one. Because of its important implications for the standards that govern criminal investigation procedures, we treat this issue in greater depth and ultimately decide that the contingent fee arrangement employed in this case offends principles of due process.
I. Factual and Procedural Background
In April 1984 Frank Kelly, a/k/a Frank Kennedy, a Government informer who had previously worked on more than thirty-five cases, was contacted by his supervisor, D.E.A. Special Agent Jimmy Bradley, and told to "infiltrate, gather information for intelligence, and report" on William Nelson. Tr. at 297. The activities of Nelson and a number of co-conspirators concluded on or about June 6, 1984. Defendants Nelson and Adalberto Cervantes-Pacheco were arrested, indicted, tried, and convicted of conspiring to possess 200,000 pounds of marijuana with intent to distribute it in violation of 21 U.S.C. §§ 841(a)(1) and 846, and conspiring to import marijuana in violation of 21 U.S.C. §§ 952(a), 960(a)(1), and 963. They each received two four-year prison terms to run concurrently; defendant Jerry Wayne Nelson was put on probation. They jointly filed a timely notice of appeal.
Defendants make two points on appeal: (1) The Government proved at most the existence of multiple conspiracies, not the single conspiracy charged in the indictment; all defendants thereby suffered substantial prejudice and their motion for a new trial should therefore have been granted. (2) The testimony of informant Frank Kelly should have been excluded because he had been assigned to make a case against a specific defendant and was paid on a contingent fee basis; Kelly thus had a financial incentive to give damaging testimony, since the amount of his fee was based entirely on the Government's evaluation of his performance in the case, including his testimony.
II. Single vs. Multiple Conspiracies
Appellants' first contention is without merit. The trial court carefully and thoughtfully instructed the jury on the difference between single and multiple conspiracies and left it free to make a well informed decision on the matter. In a full and fair charge to the jury covering more than five pages in the trial transcript, see Tr. at 1550-55, the court dealt with the same issues appellants now raise on appeal and also provided several illustrative examples taken from the trial, see id. at 1552-55. And, in fact, the jury acquitted two of the co-defendants, thereby indicating that it retained its powers of discrimination. We are therefore satisfied that appellants' first issue does not merit further discussion.
III. Was the Informant's Fee "Contingent"?
Appellants' second issue requires a more detailed analysis. After examining the record in this case we conclude that the Government's dealings with informant Kelly did indeed describe an impermissible "contingent fee" arrangement that deprived appellants of their fifth amendment right to due process.
A. Deciphering the Law of Contingent Informant Fees
Our cases on the law of contingent informant fees are based mainly on principles enunciated in Williamson v. United States, 311 F.2d 441 (5th Cir. 1962), in which the court reversed a conviction because the Government's informant had been hired under a contingent fee arrangement. Williamson held that
See also United States v. Waterman, 732 F.2d 1527 (8th Cir. 1984).
Without some . . . justification or explanation, we cannot sanction a contingent fee agreement to produce evidence against particular named defendants as to crimes not yet committed. Such an arrangement might tend to a "frame up," or to cause an informer to induce or persuade innocent persons to commit crimes which they had no previous intent or purpose to commit. The opportunities for abuse are too obvious to require elaboration.
Id. at 444. (footnote omitted).
The court noted that under McNabb v. United States, 318 U.S. 332, 340, 63 S.Ct. 608, 612, 87 L.Ed. 819 (1943), and its progeny, "it becomes the duty of the courts in federal criminal cases to require fair and lawful conduct from federal agents in the furnishing of evidence of crimes." Williamson, 311 F.2d at 444.
The court also elaborated on its reference to possible "justification or explanation" as follows:
It may possibly be that the Government investigators had such certain knowledge that Williamson and Lowrey were engaged in illicit liquor dealings that they were justified in contracting with Moye on a contingent fee basis, $200.00 for Williamson and $100.00 for Lowrey, to produce the legally admissible evidence against each of them. It may be also that the investigators carefully instructed Moye on the rules against entrapment and had it clearly understood that Moye would not induce them to commit a crime, but would simply offer them an opportunity for a sale. None of these facts or circumstances were developed in the evidence, though Moye's deposition had been taken months before the trial.
Id.
In Williamson a former convict, Robert Harris Moye, who had previously specialized in bootleg whiskey operations, contacted Government agents and proposed to put some of his hard-earned experience to good use because, as he succinctly put it, "I wanted to make some money, I had been in the penitentiary three years and not a bootlegger helped me one bit in the world and not a one of them paid me a nickel he owed me, or paid my wife." Id. at 442 (Moye deposition). Moye explained his terms of employment as follows:
Q. How were they to pay you? Let me put it that way.
A. How was he to pay?
Q. Yes, sir, how were you to be paid, what was the agreement?
A. I was to be paid $200.00 for Big Boy, $200.00 for James McBride, $100.00 for Hogie, he's Big Boy's half brother.
. . . .
Q. What did you agree to do with Mr. Rainer and Mr. Morris? What did they tell you they wanted you to do?
A. They told me to give them the major violators and told me that they'd pay for them.
Q. For you to go out and catch them, in other words?
A. If I could catch them.
Q. And they agreed to pay you $200.00 as to Jack Williamson?
A. Yes, sir.
Q. You all called him Big Boy, I believe?
A. Yes, sir.
Q. And $100.00 as to his half brother, the Lowrey boy?
A. Yes, sir.
Q. When were they to pay you the money?
A. Wasn't nothing said about when it was to be paid.
. . . .
Q. After you had made this arrangement with Mr. Rainer and with Mr. Morris, was there any kind of a written arrangement or anything on it?
A. No, sir.
Q. Just a verbal understanding with those two gentlemen?
A. Yes, sir.
Id. (Moye deposition).
After the usual surreptitious arrangements, Moye managed to purchase some illicit whiskey from Jack Williamson in the presence of an undercover agent. As the court put it, "Moye made the purchase from Williamson and produced the evidence against both Williamson and Lowrey." Id. at 444-45. (emphasis added).
In a brief concurrence Judge Brown noted that Williamson was not an entrapment case as such. Rather, it was one in which "the means used to `make' the case are essentially revolting to an ordered society." Id. at 445 (Brown, J., concurring). Judge Brown also emphasized that a contingent fee arrangement need not be explicit in order to warrant condemnation: "the `contingent fee' for a narcotics addict hardly needs to be spelled out in the terms used to trap this moonshiner. To get the stuff to feed his uncontrollable appetite, he knows that he must produce results or he will no longer be `hired'. . . ." Id. On this point Judge Brown was joined by Judge Cameron, who acknowledged that
Cf. id. (Brown, J., concurring):
For Government to offer a specific sum of money to convict a specified suspect is really more than civilized sensibilities can stand.
. . . .
What we hold is that, recognized as is the role of informer in the enforcement of criminal laws, there comes a time when enough is more than enough — it is just too much. When that occurs, the law must condemn it as offensive whether the method used is refined or crude, subtle or spectacular.
Every such informer knows that his day to day arrangement with the Government will continue only if he delivers the goods. . . . If the addict succeeds in landing some of the criminals the Government is after, he is well paid and his services will continue. If he does not, he is dropped.
The interpretation of Williamson in later cases has been uneven and its significance has thereby been obscured. In United States v. Garcia, 528 F.2d 580 (5th Cir. 1976), this court cited Williamson for the proposition that "we have condemned the use of a contingent fee where it involves the promised payment of a specified sum to convict a specified suspect." Id. at 586 (emphasis added). In United States v. Onori, 535 F.2d 938 (5th Cir. 1976), the court stated that in Williamson "we reversed a conviction for possession of illicit liquor because the informer who made the government's case had been promised a specified sum of money for successfully incriminating the indicated defendants." Id. at 942 (emphasis added). In United States v. Lane, 693 F.2d 385 (5th Cir. 1982), we provided the following gloss on Williamson: "In Williamson v. United States, this court reversed a conviction in which the informer assisting the Government was paid a contingent fee to implicate certain targeted defendants." Id. at 387 (emphasis added). Similarly, in United States v. Yater, 756 F.2d 1058 (5th Cir. 1985), the court noted that "We have limited the Williamson holding . . . to cases in which the government directs the informant to implicate government-pretargeted specific defendants." Id. at 1067 (emphasis added).
These somewhat inconsistent cases interpreting Williamson draw attention to a decisive issue: Upon what is an informant's fee impermissibly "contingent"? The following possibilities suggest themselves:
(1) payment after informant's testimony is over;
(2) payment after case is over;
(3) payment dependent on informant's performance;
(4) payment dependent on outcome of case.
These possibilities are listed in the order of increasing "strength" of the contingency. # 4 describes the strongest contingency, and if payment is conditioned upon a conviction as the "outcome," then this sort of arrangement is clearly impermissible under our cases. # 3 describes a somewhat weaker form of contingency; as detailed below, we conclude that this is the sort of contingency proscribed in Williamson and presented by the facts of the present case.
Since we come to that conclusion, we need not decide whether payment of an informant's fee is impermissibly "contingent" as a matter of law merely because it takes place after the informant's testimony (# 1) or after the trial (#2). But see note 8 infra.
In Williamson, informant Moye testified that he agreed to "go out and catch" the defendants in return for specified sums of money. Williamson, 311 F.2d at 442. According to Moye, "They told me to give them the major violators and told me what they'd pay for them." Id. Moye "caught" the defendants "in the act," so to speak, by arranging for them to consummate an illegal transaction with him in the presence of federal agents. The Williamson court concluded that Moye contracted with the Government agents "to produce . . . legally admissible evidence against each of" the defendants. Id. at 444. There is no suggestion here, or elsewhere in Williamson, that the informant's payments were conditioned upon the conviction of the defendants. Instead, it would be more accurate to say, as we put it in Lane and Yater, that Moye was paid to "implicate" the defendants. Lane, 693 F.2d at 387; Yater, 756 F.2d at 1067. Thus the Williamson situation comes squarely within the arrangement described in #3 above: payment dependent upon the informant's "performance." Williamson prohibits the Government from picking out particular individuals for investigation and paying informants if they can implicate those suspects.
Even were Williamson to apply only in a case where the informer's fee is contingent on the government garnering a conviction, as the dissent and the Government assert, this is such a case.
A fair and unbiased reading of the record inexorably leads to the conclusion that at least a portion of the informer's fee was contingent on conviction of the defendants. The agent who had primary responsibility for establishing the fee testified unequivocally that the informer's fee was "dependent on a lot of things," including "how he testified." The Government's agent, moreover, when asked whether the informer's fee was in part contingent on the ultimate result in trial, replied "sure." Supp.Rec. 32 (4th Tr.).
The informer solicited and obtained $20,000 from one of the defendants. The informer's fee was $20,000 plus expenses for some 70 days of work. That his pay, with near exactitude, matched the amount confiscated is an unlikely, and unseemly, coincidence. Had the defendants been acquitted, the Government would have had to return the $20,000.
In oral argument, the Government conceded as much. When asked "[w]hat about counsel's argument that the $20,000 pot would probably not have been there had the Government not gotten a conviction and . . . [Kelly] probably knew that," counsel for the Government failed to provide a salutary explanation:
I can't think of any response. I think that it's clear he would have gotten something as he had in every other case he had worked for the government on. I don't think there's any question that that was a convenient amount.
(emphasis added).
It may have been convenient to base the informer's pay on the amount of money he was able to extract from the defendants and to retain if they were convicted. But the enduring and timeless Constitutional principles of Due Process and fundamental fairness at trial cannot rest on evanescent conveniences of paying government informants. It is plain that at least a portion of informer-Kelly's fee was contingent on conviction of these defendants. Therefore, irrespective of whether the Williamson rule is triggered when the contingency fee is predicated on "implication" or "conviction," the fee paid to the informer in this case was impermissibly and unconstitutionally tainted.
B. Pre-targeting Specific Individuals
It emerges unmistakably from the record in this case that the Government's actions are of the sort proscribed in Williamson. As we have acknowledged elsewhere, " Williamson has been subsequently limited to require reversal of a conviction only when the specific defendant was picked out for the informer's efforts by a government agent." Onori, 535 F.2d at 942 (citing four 5th Circuit cases). That initial condition is easily met in the present case.
Informant Kelly testified at trial as follows:
Q. All right. What were your instructions? What were you asked to do?
A. I was asked to infiltrate, gather information for intelligence, and report.
Q. Any particular individual here in Laredo?
A. Yes. Bill Nelson
Tr. at 297. Later in his testimony Kelly stated that "my job was to infiltrate and report on Bill Nelson." Id. at 312. In the face of this uncontradicted testimony the Government wisely concedes that Nelson was indeed pre-targeted: "In the instant case, informant Kelly was instructed to `infiltrate, gather information for intelligence, and report' on appellant William Nelson (Tr. 297), and it therefore appears that Nelson was, in fact, a government-pretargeted potential defendant." Appellee's Brief at 31.
The dissent asserts that William Nelson's co-defendants, Jerry Wayne Nelson and Cervantes-Pacheco, should not in any event come within the Williamson rule because they were not pretargeted by the Government. We respectfully disagree.
The jury found that there was a single conspiracy. By the very nature of a conspiracy, all the defendants were inextricably linked, and therefore "sufficiently connected" so that the Williamson rule must be applied to exclude the tainted testimony for all the defendants. See United States v. Dickens, 524 F.2d 441, 446 (5th Cir. 1975), cert. denied sub nom., Glenos v. United States, 425 U.S. 994, 96 S.Ct. 2208, 48 L.Ed.2d 819 (1976).
C. Contingent Fee Payments
The record in this case also supports a finding that the Government set up an impermissible contingent fee arrangement with its informant. As in Williamson, the informant's payment depended on his performance. In fact, it appears that this case involves an even stronger form of contingency than Williamson did. Here, the informant's "performance" included not merely his undercover operations — as in Williamson — but also extended to his performance while testifying at trial, thereby directly tainting the factfinding process itself.
Informant Kelly testified as to his terms of employment as follows:
Q. And how are you paid, sir?
A. I'm paid after a case is over with, anywhere from when the case is over to a year later, depending on Washington. It's Washington's job to tell . . . the Agency that I'm working for would request an amount of money and Washington would approve or disapprove that amount, depending on the way they feel I testified or the way they feel about my infiltrating, gather information, report that information, and testify.
Tr. at 295 (emphasis added). Later Kelly clarified that testifying at trial was an integral part of his assignment, and one that determined whether he would be paid:
A. Initially, when they requested my services, he said, "Will you testify," and I said, "Yes." That's it.
Q. So you knew then that it's part of your job in this case to testify about your investigation?
A. Yes, sir.
Q. And you knew that you wouldn't be paid until the case was concluded; is that correct?
A. Yes, sir.
Q. And you knew that conclusion of this case meant your testimony in this case?
A. Yes, sir.
Id. at 418-19.
Later at trial agent Bradley testified on the factors he considered in determining how much to pay Kelly. This testimony makes clear that the quality of Kelly's testimony determined whether he would be paid:
Q. You pay for performance, what they perform?
A. That is correct.
Q. Now, when you figure out this recommendation, do you isolate the different things that the informer has done for you or do you consider it as a whole?
A. I evaluate different aspects of the case and what his participation was in such case.
Q. Do you weigh different aspects against each other when you formulate a final figure, or is there one that just overrides all of them?
A. No, I would say that you take them and combine them altogether and you come up with an evaluation of what you think it's worth.
Q. Okay. So, if I understand you correctly, you were going to weigh these factors in your own mind?
A. That is correct.
Q. Now, are there any set procedures or guidelines that you're aware of that would assist you in weighing these different factors of performance?
A. There are no set guidelines. It's strictly up to my discretion.
Q. Okay. I would imagine that there are numerous factors you're going to consider.
A. Yes. Each case is different.
Q. One would be, for instance . . . what, the availability of Mr. Kelly whenever you sought him? In other words, when you picked up the phone, Mr. Kelly answered and he was always there whenever you needed him? That would be a factor, wouldn't it?
A. That would be a factor, yes.
Q. Okay. And the responses that you got from Mr. Kelly, as to whether or not they were complete, that would be a factor?
A. Yes.
Q. And whether or not they were accurate, that would be a factor?
A. Yes.
Q. And his willingness to repeat these responses at a later time would be another factor, Mr. Kelly's performance?
A. You mean, testifying? Yes.
Q. To repeat it anywhere?
A. Right.
Q. You certainly don't want Mr. Kelly to have the authority to say when he will repeat his responses and when he will not; is that correct?
A. I believe he does not have that authority.
Id. at 660-62.
Further, it even appears that if Kelly performed well enough on the stand, his reward would be the very $20,000 that the Government seized in this case:
Q. (Mr. Zelman resuming) When you make your recommendation, you're going to wait until the end of the case?
A. Yes.
Q. This case is not over, is it?
A. The case is never over until the trial is over.
Q. Okay. And you are right now in the process of evaluating Mr. Kelly; is that correct?
A. I've already evaluated Mr. Kelly.
Q. How did he do?
A. He did great.
Q. You're going to recommend a lot of money for him?
A. I sure am.
Q. What figure?
A. I'm going to recommend that he gets that twenty thousand dollars he seized.
Q. Now, before Mr. Kelly testified, had you had any figure in mind?
A. No.
Q. So you were going to wait until Mr. Kelly's testimony before you figured how much you were going to recommend he get?
A. That is correct.
Id. at 662-63 (emphasis added). Thus, the quality of the informant's testimony determined not only whether he would be paid but also how much he would be paid. Needless to say, such an arrangement provides strong financial incentives for giving damaging testimony. Indeed, one might fairly remark that this case provided the entrepreneurial informant a splendid opportunity to "write his own ticket." As the Williamson court noted, "The opportunities for abuse are too obvious to require elaboration." 311 F.2d at 444.
The dissent posits that even if there were an impermissible contingency fee, the fee would be justified "[b]ecause of the government's reasonable suspicion that Nelson was smuggling drugs." At 697. It is true that the Government in this case could reasonably suspect that there was wrongdoing prior to launching its investigation. It is also true that in past cases we have recognized that reasonable suspicion may justify the use of a contingency fee. Harris v. United States, 400 F.2d 264, 266 (5th Cir. 1968); Sears v. United States, 343 F.2d 139, 144 (5th Cir. 1965); Hill v. United States, 328 F.2d 988, 989 (5th Cir. 1964), cert. denied, 379 U.S. 851, 85 S.Ct. 94, 13 L.Ed.2d 54 (1964); see Williamson, 311 F.2d at 444.
But we have also recognized that reasonable suspicion, standing alone, may not be enough to justify an otherwise impermissible contingency fee. Rather, in order to justify the fee, the investigation must be of unusual difficulty. See Sears, 343 F.2d at 144. Here, the Government has not argued, nor does the record support an argument, that the investigation was in some way unusually difficult.
Although reasonable suspicion standing alone may sometimes be sufficient to justify a fee contingent on successful investigation, this justification is irrelevant and inapposite when the fee is made contingent on the informer's testimony at trial. In none of our prior cases have we found it necessary to reach this question, since this is the only case since Williamson was decided in which the informer's fee was premised, at least in part, on the quality of the informer's testimony and on the defendant's conviction.
IV. Conclusion
We conclude on the basis of the foregoing analysis that the testimony of the Government's main witness, informant Kelly, is inherently untrustworthy and should have been excluded as such. We find no error in the trial court's handling of the conspiracy issue; but on the contingent fee issue we reverse and remand the case to the district court for a new trial.
We see nothing in Williamson or later cases that bars an arrangement under which an informant involves himself with a "targeted" individual and is later compensated, after the job and testimony are over, according to an objective assessment of the actual time and effort expended and of the quality of the services. We would not condemn a payment based on the quality of the informant's work, including the quality of the testimony, so long as that evaluation of quality was determined by the informer's ability to accurately and truthfully recall the facts. That is not what happened here.
The introduction of the tainted testimony was not harmless error. The standard for showing that a constitutional error was harmless is appropriately stringent. A "conviction should . . . be set aside if the reviewing court may confidently say, on the whole record, that the constitutional error was harmless beyond a reasonable doubt." Rose v. Clark, ___ U.S. ___, ___, 106 S.Ct. 3101, 3105, 92 L.Ed.2d 460 (1986) ( quoting Delaware v. Van Arsdale, ___ U.S. ___, ___, 106 S.Ct. 1431, 1436, 89 L.Ed.2d 674 (1986) (emphasis added); see Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 828, 17 L.Ed.2d 705 (1967).
It is true that much of the informer's testimony was corroborated. See at 697-699. But informer-Kelly was the Government's star performer at trial. The case revolved around Kelly and his bartered-for testimony. As Judge Davis notes, "Kelly remained on the witness stand for the better part of a day and testified in excruciating detail about his activity. . . ." At 700. The Government heavily relied on Kelly's testimony to convict the defendants. As the Government prosecutor put it at trial, "Kelly will be the one . . . that will tie it all up." Tr. 272. Reading the record as a whole, we cannot confidently say that without Kelly's testimony the Government would have proved beyond a reasonable doubt that the defendants conspired to bring marijuana into the United States.
One of the basics of our jurisprudence is the search for truth, and by this is meant not the purchased truth, the bartered-for truth, but the unvarnished truth that comes from the lips of a man who is known for his integrity. A corollary of our search for truth, a component thereof, is that the plaintiff or defendant, prosecution or defense, as the case may be, vouches for the truthfulness and the integrity of its witnesses. The government in its prosecutorial efforts should be like Caesar's wife, above or beyond reproach. At the very least, the court must tell a jury that the words of a witness have been in a sense purchased if he will be paid, more or less, depending upon how effective his putative truth-telling sells itself to the jury. It may be that we must live with informers. It may be that we must live with bargained-for pleas of guilty. But we do not have to give a receipt stamped "paid in full for your damaging testimony" or "you will be paid according to how well you can convince the jury even though it be in the face of lies."
It is true that the precedents we cite are timid in their approach to this problem, fearful that somehow the condemnation of contingent fee arrangements will destroy our criminal justice system. If that be true, then our system of finding the truth is pallid and weak and not to be trusted. Trustworthiness is a keystone and a hallmark of any judicial system that seeks recognition for its role in a civilized society. The time has come to announce boldly and firmly that our juridical search for the truth cannot be reconciled with the virtual purchase of perjury.
REVERSED AND REMANDED FOR A NEW TRIAL
Circuit Judge W. EUGENE DAVIS adheres to his original dissent reported at 793 F.2d 695.