Robert Chevoor was indicted under 18 U.S.C. § 1623 for knowingly making false declarations to a federal grand jury. The district court, finding fundamental unfairness in the totality of the circumstances surrounding Chevoor's grand jury appearance, suppressed the grand jury transcript and dismissed the indictment. 392 F. Supp. 436 (D.Mass.1975).

In January, 1973, an intensive government investigation was launched into alleged gambling and loansharking by one Michael Pellicci, activities for which he was subsequently indicted. This effort involved the use of undercover agents and of court-authorized interceptions of wire and oral communications. From these sources, the government obtained evidence which indicated that Chevoor was a victim of Pellicci's loansharking, owed Pellicci $7,000, was being pressed to do political favors for Pellicci, and also was involved in the repayment by a third party of a larger loan to Pellicci. Particularly damning in the government's view was an intercepted December 9, 1973 conversation between Pellicci and Chevoor in which 'payments' by the alleged third party were discussed.

On the evening of January 8, 1974, two F.B.I. agents went to Chevoor's house to serve upon him a subpoena calling for his appearance before the grand jury on January 9. Prior to serving the subpoena, agent Vaules told Chevoor that he was not a grand jury target, but that the grand jury was investigating Pellicci, and he asked Chevoor a number of questions about his association with Pellicci, including whether Chevoor owed Pellicci money or knew about others who did. Chevoor answered these questions in the negative. The district court subsequently found that his responses were clearly inconsistent with the transcript of the December 9 intercepted conversation. The district court also found that during the course of this meeting, the F.B.I. agent told Chevoor that he 'had to testify', or at least 'had to appear', before the grand jury.

The subpoena (and the agent) directed Chevoor to report the next morning to Strike Force attorney Friedman in connection with his grand jury appearance. Friedman told Chevoor that the grand jury was investigating Pellicci's loansharking, and that Chevoor was not a target, but that he could expect to be prosecuted if he gave false testimony before the grand jury. Then, with a transcript of the December 9 conversation before him (not identified as such to Chevoor), Friedman questioned Chevoor about whether he had ever borrowed money from Pellicci, whether he owed money to Pellicci, whether he knew others who did, and whether he had ever discussed with Pellicci repayments of such loans. Chevoor again responded negatively. Friedman said he knew Chevoor was not cooperating and not telling the truth; Chevoor asked in what respects this was so. Vaules (who was there throughout this interview) interrupted: 'We are not going to tell you what we have on you.' Despite this pressure, Chevoor did not change his story. The district court found that on the way to the grand jury room, Friedman told Chevoor that he 'just got six or seven guys like you for perjury' in New York or New Jersey.

Before the grand jury, the same questions, among others, were asked, and the same negative responses given by Chevoor. His later indictments were for material false statements with respect to whether Chevoor had ever owed money to Pellicci; whether he had ever discussed with Pellicci payments due to the latter; and whether he had ever discussed with Pellicci the fact that other individuals owed Pellicci money.

The district court found as a fact that during the January 9 office conversation, neither Friedman nor Vaules informed Chevoor that the December 9 conversation had been intercepted,1 that the questioning was based on that conversation, or that the answers given both on the 8th and the 9th conflicted specifically with it. The district court also found that on January 9, Chevoor was not given Miranda warnings or told of any of his rights under the Fifth and Sixth Amendments. At the time of his appearance, Chevoor was a 40 year old school teacher and real estate broker, and a member of the Watertown Redevelopment Authority. He had a B.A. in business administration and had recently received a Master's degree in Education. He was not accompanied by a lawyer.

The district court's assessment of this case relied heavily upon the existence of 18 U.S.C. § 1001, which makes it a crime to make false statements in matters within the jurisdiction of federal departments or agencies.2 The court felt that Chevoor's responses prior to entering the grand jury room, first on January 8 to the F.B.I. agents, then on January 9 to Vaules and Friedman, had violated this statute.3 These violations, in the opinion of the district court, made government assurances to Chevoor that 'he was not a target' at least misleading. 'Both (Friedman) and Vaules knew, or should have known, that the defendant was subject to possible indictment for having made false statements to a federal officer.' 392 F. Supp. at 440.

The district court, moreover, found that Chevoor was a target in another sense. '(I)t would have been unrealistic for Vaules and Friedman to assume that defendant would do anything but give the same, apparently untruthful, answers before the grand jury.' Id. Although Pellicci was their primary interest and the subpoena originally was issued to Chevoor as a witness in the investigation of Pellicci, the 'prime purpose of putting him on the (grand jury) stand', in light of the intervening denials, was, in the view of the district court, to obtain a perjury indictment. Id. at 442.

Putting these two factors together, the district court saw Chevoor as being on the horns of a dilemma. The government had put him into a situation where his choices were to tell the truth, which would provide evidence against him under a subsequent § 1001 indictment, or to lie, which would open him up to an indictment for perjury. In this connection, the agent's indication the night before that Chevoor 'had to testify', coupled with the absence of warnings as to his rights, foreclosed to Chevoor his third option, the one which should have been afforded to him explicitly: his right to take the Fifth Amendment.

The court found that the totality of the circumstances made Chevoor a 'potential, if not probable' and a 'likely, if not inevitable' grand jury target. As such, the court reasoned, Chevoor should have been told of his right to remain silent. While recognizing that the failure to give such a warning is ordinarily no license for perjury, the court cited 'unique and compelling circumstances' which in this case warranted such a result: (1) the unknown interception; (2) the two statements inconsistent with that conversation; (3) the government assurances that Chevoor was not a target; and (4) the summons without warnings for the purpose of asking the same questions.4 These factors added up to a situation where the government 'turned the screw too tightly' and violated Chevoor's right not to be 'ensnared'; the resulting fundamental unfairness justified suppression of grand jury testimony under the due process clause of the Fifth Amendment and the court's supervisory powers; and the indictment was dismissed because without the grand jury testimony there was no evidence of perjury. Were we to accept all of the premises assumed by the district court--i.e., the government's placing of a witness in a situation where criminal prosecution would follow whether he lied or told the truth--we might well be of the same mind. Our difference stems from our non-acceptance of one of the premises.

The district court saw Chevoor in the position of a grand jury witness whose testimony was so likely to be incriminating that the jury could be said to have the objective of using him as a target for ultimate prosecution rather than as an instrument for investigating the criminality of another. The general rule, of course, is that a prospective witness such as was Chevoor when he was subpoenaed5 is entitled to no warnings of constitutional rights. United States v. DiMichele, 375 F.2d 959 (3d Cir.), cert. denied, 389 U.S. 838, 88 S. Ct. 54, 19 L. Ed. 2d 100 (1967). Even if the prosecution and the grand jury fully expect that the witness will perjure himself, there is no obligation to give protective warnings.6 Absent other pressures, a witness is expected to tell the truth. The privilege against self-incrimination bars compelled testimony as to past crimes; it does not shelter new perjury.7

When a grand jury witness is knowingly put in the position, if he testifies at all, of either perjuring himself or incriminating himself, a number of courts have recently held that he should be given Miranda warnings or at least the advice that he may remain silent as to incriminating matters--the position taken by the district court.8 We have considerable sympathy with this approach. While there is not the isolated and unobservable station house custody which underlies the holding in Miranda v. Arizona, 384 U.S. 436, 86 S. Ct. 1602, 16 L. Ed. 2d 694 (1966), the conjunction of an assembled grand jury, a vigorous prosecutor, and ex parte proceedings conducted in the absence of a lawyer counselling the witness gives rise to a kind of coerciveness suggesting the wisdom of giving at least notice that a witness need not testify if such would incriminate him.

We need not, however, presently face the questions whether, when, and what warnings should be given 'target' grand jury witnesses, for Chevoor was not placed in a situation where his only 'safe harbor', to use the phrase of United States v. Mandujano, 496 F.2d 1050, 1058 (5th Cir. 1974), cert. granted, 420 U.S. 989, 95 S. Ct. 1422, 43 L. Ed. 2d 669 (1975), was a recourse to silence of which he had not been told. Although he claims to have been thrust upon the horns of, as Wigmore put it, a triceratops, 8 J. Wigmore, Evidence, § 2251 at 316 (McNaughton rev. 1961)--facing perjury, self-incrimination, or contempt--we find no such beast here.

We have concluded that Chevoor faced only the alternatives of perjury or telling the truth. Self-incrimination was not a foreseeable possibility and, therefore, there was no right to remain silent as to which he should have been warned. For he was not liable for prosecution under 18 U.S.C. § 1001 when he entered the grand jury room. Instead, his responses to Vaules' and Friedman's questioning fall within the 'exculpatory no' category of responses which are outside the scope of 'statements' within the meaning of the statute.

The statute from which § 1001 evolved originally prohibited false claims against the government and the use of false statements in support of such fraudulent claims. Act of March 2, 1863, 12 Stat. 696. In 1934, an amendment broadened the statute by dropping the restriction to pecuniary or proprietary loss to the United States. Act of June 18, 1934, 48 Stat. 996. This was meant primarily to protect the newly created federal regulatory agencies against fraudulent practices. The Supreme Court described the added scope as follows: 'The amendment indicated congressional intent to protect the authorized functions of governmental departments and agencies from the perversion which might result from the deceptive practices described.' United States v. Gilliland, 312 U.S. 86, 93, 61 S. Ct. 518, 522, 85 L. Ed. 598 (1941). The false claims section has since been split from the fraudulent statements section, Act of June 25, 1948, 62 Stat. 749, but none of the minor changes since 1934 have affected the substance of the statute, United States v. Bramblett, 348 U.S. 503, 75 S. Ct. 504, 99 L. Ed. 594 (1955).

Under Gilliland and Bramblett, the provision has a very broad scope. But in Friedman v. United States, 374 F.2d 363 (8th Cir. 1967), the court found that most prosecutions fell into four categories: (1) giving false information with the purpose of receiving a monetary or proprietary benefit, e.g., Bramblett, supra; (2) resisting monetary claims by the United States through the presentation of false information, e.g., United States v. Knox, 396 U.S. 77, 90 S. Ct. 363, 24 L. Ed. 2d 275 (1969); (3) seeking some governmental privilege such as employment or security clearance on the basis of falsified information, e.g., United States v. Marzani, 168 F.2d 133 (D.C. Cir.), aff'd by an equally divided court, 335 U.S. 895, 69 S. Ct. 299, 93 L. Ed. 431 (1948); and (4) giving false information to frustrate lawful regulation, e.g., Gilliland, supra.

The courts have had difficulty in affirming coverage of F.B.I. investigations. Although differing rationales have been used, there is widespread agreement that mere negative responses to government initiated inquires should not be included in the statutory proscription. Such utterances fall within neither of the dual categories of false statement intended, according to Gilliland and Bramblett, to be covered by the statute: those in support of fraudulent claims, and those tending to pervert the authorized functions of departments and agencies. Moreover, to allow § 1001 to cover such statements would undermine the safeguards that are normally provided in perjury prosecutions, primarily the formality of the oath, while permitting sanctions as great as those under perjury statutes.

Cases involving the F.B.I. or U.S. Attorney initiating the questioning9 have uniformly held the statute not to cover the responses, although not always by the same rationale.10 The harder question, which is not involved in this case, arises when the defendant is indicted for going to the authorities to falsely initiate an investigation against a third party.11

This case involves a clear exculpatory denial situation. Chevoor had presented no false claim against the government, nor was this investigation relative to a claim the government might have had against him. The F.B.I. questioned Chevoor in the course of a criminal investigation; he denied involvement in, or knowledge of the criminal activity. The scenario was repeated at the Strike Force office. The defendant here did not initiate anything; he did not even go so far as to fabricate a misleading story in response to the inquiries. He merely gave negative, oral responses to the questioning.12 No oath was given; no transcript taken. The interviews were informal. Under all these circumstances, we hold that Chevoor's responses were not 'statements' within the meaning of 18 U.S.C. § 1001. The government quite properly did not charge him under that section; instead, they charged him under a statute carrying with it the safeguards of formality, normally including a transcript, as well as the average citizen's awareness that lying under oath is a crime. Moreover, they provided Chevoor with specific warnings of his possible criminal liability if he lied before the grand jury.

Additionally, there are no facts which would indicate that the government ever considered the possibility of prosecuting Chevoor under § 1001. None of the pressure on Chevoor concerned possible liability for having already lied; it was all directed to the commission of perjury in the grand jury room. Instead of telling Chevoor that he had lied and was therefore in trouble, they told him that they knew he was lying and that he would be in trouble if he continued to do so before the grand jury. The district court found as a fact that neither Friedman nor Vaules told Chevoor of the existence of § 1001 or of the possibility of prosecution for any statements made outside the grand jury room. Indeed, even after the January 8 responses, Friedman assured Chevoor on January 9 that he was not a target. While our analysis does not depend upon self-serving statements of the prosecutor, whether contemporaneous or subsequent, we find nothing in the record to cast doubt on the truthfulness of that assurance. The district court's finding that it was 'misleading' depended upon its premise that Chevoor was in fact liable, a premise which we have rejected. And the fact is that Chevoor was not subsequently indicted under that section. If he had been the focus of a § 1001 inquiry during his grand jury appearance, and if he had subsequently been indicted on such a charge, any incriminating testimony he had given to the grand jury might well have been suppressible. But the actual facts of this case do not present a situation where the prosecution called a witness before the grand jury to obtain either perjury or incriminating evidence on an already committed offense.

If the test depends entirely upon a perceived dilemma on the part of the grand jury witness, Chevoor fares no better. At the hearing on the motion to suppress, he testified under the protection of a ruling analogous to that provided for in Simmons v. United States, 390 U.S. 377, 88 S. Ct. 967, 19 L. Ed. 2d 1247 (1968), that 'his testimony will not be able to be used for any other purposes'. At that hearing Chevoor evidenced no understanding that lying to federal officers was a crime. More telling, he insisted that he had never given false statements to the grand jury, and it is agreed that his statements to the grand jury paralleled the earlier statements to the F.B.I. and the U.S. Attorney. When asked if he had ever told his lawyer that he had testified falsely and wanted to change his testimony, he responded: 'That I had testified falsely? No.' And later: 'And to my knowledge, I did not say anything that was false.'

Not only does this testimony undercut Chevoor's argument that he had violated § 1001; it demolishes his entire case. It was not a triceratops which threatened him. Since he had not lied to a federal officer, he was not guilty of a substantive offense. Any choice to commit perjury was of his own making, not the result of a dilemma created by the government. This course, and the third alternative of risking contempt through silence, confronted Chevoor no more than the average citizen called before the grand jury. Without liability under § 1001, Chevoor was a mere witness, not any sort of a defendant; he therefore had no right to warnings, and the agent's indication that he 'had to testify' was entirely accurate.

Chevoor's second line of defense is that the government's sole purpose in calling him before the grand jury was to extract perjury for which to indict him. The facts do not support this argument. Unlike Brown v. United States, 245 F.2d 549, 555 (8th Cir. 1957), where such a defense was successful, the grand jury here was conducting a legitimate investigation into crimes which had in fact taken place within its jurisdiction. See LaRocca v. United States, 337 F.2d 39, 42--43 (8th Cir. 1964); but cf. United States v. Thayer, 214 F. Supp. 929 (D.Colo.1963). It is true that the government did not entirely cooperate with Chevoor, but it is not required to do so. It did not ensnare Chevoor into committing perjury: both Vaules the night before and Friedman prior to the grand jury appearance warned Chevoor that they had reason to believe he was lying, and Friedman added that if he committed perjury in the grand jury he could expect to be prosecuted for it. Moreover, it was possible (even though unlikely) that when it came to the crunch of testifying under oath, with a transcript, Chevoor would succumb to the truth. We cannot say that calling Chevoor in these circumstances even in the anticipation that he would perjure himself is beyond the pale of permissible prosecutorial conduct. See United States v. Nickels, 502 F.2d 1173, 1176 (7th Cir. 1974). Nor does this reach the level of inexcusable instigation condemned in United States v. Remington, 208 F.2d 567, 573 (2d Cir. 1953) (L. Hand, J., dissenting), cert. denied, 347 U.S. 913, 74 S. Ct. 476, 98 L. Ed. 1069 (1954).

It is obvious that if Chevoor had admitted knowledge of Pellicci's loansharking activities, his testimony could have been a fruitful source of leads for the investigation. It would be anomalous to rule that his criminal intent should be permitted to thwart that investigation. We hold that it was not unfair for the grand jury to question him. Of course, we do not rule on whether Chevoor did in fact perjure himself when responding to those questions; that will be for the jury to decide.

Reversed.