Errors Before the Grand Jury

Errors Before the Grand Jury

Harmless-error review has so far been described as a way of sorting the mistakes made at trial, which raises the question whether it stops at the verdict. It does not, because a federal prosecution begins earlier, with a grand jury that hears the government's evidence in secret and decides whether probable cause supports a charge. Rule 6 of the Federal Rules of Criminal Procedure regulates that stage. Rule 6(d), for example, limits who may be present while the grand jury is in session, in order to guard secrecy, prevent intimidation of the jurors, and leave them free to examine the evidence without undue prosecutorial influence.¹ When the government breaks such a rule, the defendant asks a court to dismiss the indictment before trial or to set aside the conviction afterward. The answer depends on whether Rule 52(a), which requires a court to disregard any error, defect, irregularity, or variance that does not affect substantial rights, reaches the grand jury at all.²

The text of Rule 52(a) draws no line between the grand jury and the trial, and the Supreme Court has read it the same way. In United States v. Mechanik the Court saw no reason to apply the Rule to grand jury errors any differently than to errors in the trial itself.³ The Court assumed for the sake of argument that two government agents who testified together before the grand jury had violated Rule 6(d). The defendants, who had been reasonably diligent, did not obtain the transcript revealing the joint appearance until the second week of trial.⁴ The petit jury's later guilty verdict made reversal inappropriate, and the reasoning is straightforward. Rule 6(d) protects a defendant from having to answer a charge unsupported by probable cause, and a trial jury's finding of guilt beyond a reasonable doubt establishes far more than probable cause, so any distortion of the charging decision was harmless beyond a reasonable doubt.⁵ The cost side of the ledger points the same way. Reversal forces jurors, witnesses, courts, and lawyers to repeat a trial, memories fade, and a retrial can become difficult or impossible, so reversal for an error that changed nothing may hand the accused complete freedom from prosecution.⁶ After a verdict, moreover, nothing can restore the defendant to the position of never having been tried on that indictment, because whatever inconvenience, expense, and public stigma a proper indictment might have spared has already been borne.⁷ Lower courts have disagreed about how broadly this verdict-cures-the-indictment reasoning extends, which is why the kind of defect matters.⁸

Mechanik arose from a challenge that surfaced only during trial, and the Court expressly declined to say what remedy fits a violation that affected the charging decision and was raised before trial began.¹⁰ Bank of Nova Scotia v. United States addressed that earlier setting. A court there was asked to dismiss an indictment before the trial had concluded, and the Court held that, as a general matter, a district court may not dismiss an indictment for errors in grand jury proceedings unless those errors prejudiced the defendants.¹¹ The opinion reaches that result in two steps, and they organize the rest of this section. The first concerns the source of the district court's power, since a federal court cannot invoke its supervisory authority over procedure to circumvent the inquiry Rule 52(a) prescribes.¹² The second supplies the prejudice standard for dismissing an indictment and identifies a narrow class of fundamental errors, such as racial discrimination in the selection of grand jurors, that bypass the inquiry because prejudice is presumed.¹³ The supervisory-power question comes first, because until a court knows whether its own authority can override Rule 52(a), the prejudice standard has nothing to govern.

Rule 52(a) Limits Supervisory Power

Once a defendant shows that the government broke a grand jury rule, the first question is whether the judge may dismiss the indictment simply because the conduct deserves a response, without asking whether it hurt anyone. Bank of Nova Scotia v. United States answers no, and the answer begins with what supervisory power is: the authority of a federal court, within limits, to formulate procedural rules that the Constitution and Congress do not specifically require.¹⁴ The power exists to fill gaps, since written law cannot anticipate every procedural problem a court must manage. The phrase "within limits" carries the doctrine, because even a sensible and efficient use of the power is invalid if it conflicts with a constitutional or statutory provision.¹⁵ The reason is structural: a court that could override conflicting law whenever the override seemed wise would hold a discretionary power to ignore the limits that the law it enforces has deliberately set.¹⁶

Rule 52(a) sits on the binding side of that line. It directs that any error, defect, irregularity, or variance that does not affect substantial rights must be disregarded, and the verb is mandatory.¹⁷ The Rule draws its force from its origin. It was promulgated under a statute (18 U.S.C. § 687 in the 1946 edition, now codified as amended at 18 U.S.C. § 3771) that authorized the Supreme Court to prescribe rules of practice and procedure for proceedings up to and including verdict, and that statute provided that once a rule took effect, laws in conflict with it would have no further force.¹⁸ Rule 52 is therefore, in every pertinent respect, as binding as a statute enacted by Congress, and a federal court has no more discretion to disregard its mandate than to disregard a constitutional or statutory provision.¹⁹ That settles the ranking between the two sources of authority. Supervisory power is the court's own gap-filling authority, and Rule 52(a) is a command the power must yield to, so a judge cannot reach a dismissal by labeling the ground supervisory and skipping the question the Rule asks, which is whether substantial rights were affected.

The facts of Bank of Nova Scotia show the temptation the ranking is meant to resist. The district court had dismissed the indictment before trial after finding several violations of the Rule 6 provisions governing grand jury proceedings, and it added a finding that the totality of the circumstances had undermined the integrity of the grand jury process.²⁰ The court of appeals held that the violations that remained did not justify dismissal, and a dissenting judge argued that egregious misconduct warranted dismissal as a prophylactic measure even without any showing of prejudice.²¹ The Supreme Court rejected the prophylactic dismissal and held that a federal court may not invoke supervisory power to circumvent the harmless-error inquiry that Rule 52(a) prescribes.²² A district court therefore exceeds its powers when it dismisses an indictment for prosecutorial misconduct that did not prejudice the defendant.²³

The Court supported that result with United States v. Hasting, which had held that supervisory power to reverse a conviction is unnecessary when the error is harmless, since by definition the conviction would have been obtained notwithstanding the error.²⁴ Hasting also closed off the rationales a judge is most likely to offer instead. Deterrence is an inappropriate basis for relief where more narrowly tailored means of deterring objectionable prosecutorial conduct exist, concern for the integrity of the judicial process carries less weight when the error is harmless, and a court may not set harmless-error doctrine aside to chastise what it views as prosecutorial overreaching.²⁵ The common logic is that a remedy should answer a wrong done to the defendant. When the violation changed nothing about the decision to indict, dismissal hands the defendant a benefit unconnected to any harm, and its real function is to punish the government, a function that more targeted measures can serve without voiding a charge the grand jury might have returned anyway. Rule 52(a) also reflects a balance between the costs to society and the rights of the accused, and that balance may not be casually overlooked because a court has chosen to frame the question as one of supervisory power.²⁶

There is a further reason the supervisory label cannot enlarge the remedy. Hasting involved constitutional error, while the Rule 6 violations in Bank of Nova Scotia were nonconstitutional, and the Court drew the consequence that it would be inappropriate to devise a rule letting federal courts deal more sternly with harmless nonconstitutional errors than with harmless constitutional ones.²⁷ The ordering of seriousness explains why. Constitutional violations are the weightier category, so if supervisory power could not be used to reverse a conviction for harmless constitutional misconduct, allowing it to dismiss an indictment for a harmless breach of a mere procedural rule would attach the heavier remedy to the lighter wrong. Applying the same harmlessness inquiry to both removes that inversion.

Bank of Nova Scotia itself marks the side where the rule applies: a district court that dismisses an indictment to respond to prosecutorial misconduct, with no finding that the misconduct prejudiced the defendants, exceeds its power, because Rule 52(a) cannot be bypassed through supervisory authority.²⁸ The rule does not shield every violation, however, because the Rule protects only errors that leave substantial rights unaffected.²⁹ Flipping a single fact shows the difference.

The second version works because the Court treats a violation that substantially influenced the decision to indict as outside the harmless category, which is the standard the next section develops along with the narrow class of fundamental errors where prejudice is presumed.³⁰ For now the pivot is the object of the inquiry. Students lose this issue when they treat the seriousness of the prosecutor's conduct as a substitute for a finding about its effect on the charging decision.³¹

Prejudice and Fundamental Exceptions

Rule 52(a) directs a court to disregard any error that does not affect substantial rights, and the preceding discussion showed that a judge cannot reach around that command by invoking supervisory power.³³ What remains is to learn what the Rule's own inquiry asks when the decision under review is a grand jury's vote to indict. Bank of Nova Scotia answers in two parts: a prejudice standard that governs the ordinary violation, and a narrow class of fundamental errors that escape the standard because prejudice is presumed. The standard comes first, since the exceptions are defined by contrast with it.

The Court adopted its test, at least where dismissal is sought for nonconstitutional error, from Justice O'Connor's concurring opinion in United States v. Mechanik.³⁴ Under that test, dismissal of an indictment is appropriate only if it is established that the violation substantially influenced the grand jury's decision to indict, or if there is grave doubt that the decision to indict was free from the substantial influence of such violations.³⁵ The two prongs work together. The first asks whether the violation's influence on the charging decision has been shown, and the second applies when the record leaves the reviewing court unable to say either way. A court left with real doubt therefore does not treat the violation as harmless.

The test is a transplant from trial-error review. It rests on Kotteakos v. United States, which construed the statute later incorporated into Rule 52(a) and held that a conviction should not be overturned unless, after examining the record as a whole, a court concludes that an error may have had substantial influence on the outcome.³⁶ Kotteakos added that the inquiry asks whether the error itself had substantial influence, not merely whether enough other evidence remained to support the result, and that a court left in grave doubt cannot let the judgment stand.³⁷ Borrowing that approach for the indictment stage makes sense because the question is the same at both stages: did the error change the decision under review? A violation that cannot be shown to have influenced the vote to indict leaves the defendant no worse off than a flawless proceeding would have, and dismissing for it would mostly send the government back to present the same case to a new grand jury. Where influence is shown or doubt remains, dismissal is available, because there the error may have changed the charging decision itself.

Bank of Nova Scotia shows the standard failing on a record that looks alarming. The district court had found that the government disclosed grand jury materials to Internal Revenue Service employees with civil tax enforcement duties, instructed two witnesses not to reveal the substance of their testimony, and allowed joint appearances by IRS agents to read transcripts to the jurors, among other findings.³⁸ The Supreme Court nonetheless reviewed the record only to assess the influence, if any, of the alleged errors on the decision to indict, and it agreed with the court of appeals that prejudice had not been established.³⁹ The disclosure and secrecy violations show why a long list can add up to little: misconduct that has nothing to do with how the case was presented to the jurors almost always qualifies as harmless, because only in unusual circumstances can it contribute to the decision to indict.⁴⁰ A violation that touches the jurors' deliberations directly stands on the other side of the line. A prosecutor who remains in the jury room while the grand jurors deliberate, contrary to Rule 6, creates a real possibility that their vote will be unduly influenced, and that is the kind of violation for which a defendant can hope to establish substantial influence.⁴¹

The Court's qualification that it adopted the test at least for nonconstitutional error also deserves attention. The Rule 6 violations in the case were not constitutional, so the Court did not fix the showing that governs a constitutional violation before the grand jury.⁴² It is arguable that the more demanding trial-stage rule would apply there, under which dismissal would follow unless the court is convinced beyond a reasonable doubt that the error did not contribute to the decision to indict.⁴³ Where the misconduct is constitutional, then, the substantial-influence test should not be assumed to be the whole inquiry.

The standard has a boundary. Bank of Nova Scotia set apart a class of cases in which indictments were dismissed without any particular assessment of prejudice, because the errors were deemed fundamental.⁴⁵ Those decisions can be read as isolated exceptions to the harmless-error rule, but the Court preferred a clearer explanation: they involve grand juries whose structural protections were so compromised that the proceedings were fundamentally unfair, which allows prejudice to be presumed.⁴⁶ In Vasquez v. Hillery racial discrimination in the selection of grand jurors compelled dismissal of the indictment, because the error was of constitutional magnitude, other remedies were impractical, and a discriminatorily selected grand jury could be presumed to treat defendants unfairly.⁴⁷ Ballard v. United States reached a like result where women had been excluded from the grand jury, since the nature of the violation allowed a presumption of prejudice and any harmless-error inquiry would have required unguided speculation.⁴⁸

The presumption exists because the prejudice inquiry needs a baseline, and structural error removes it. In an ordinary Rule 6 violation, a reviewing court can picture the same jurors hearing the same evidence without the violation and ask whether the vote would have differed. When the defect lies in who sat as grand jurors, the comparison would be to a different body whose vote exists nowhere in the record, so a court deciding harm or harmlessness would be guessing. Kotteakos is hostile to broad presumptions applied across all cases, yet it allows a particular one that arises from the nature of the error and its natural effect in the setting, and a discriminatorily composed grand jury fits that description.⁴⁹ The presumption is therefore narrow, attaching to errors that corrupt the grand jury as an institution rather than to every rule violation that sounds serious.

Grand jury discrimination also survives a later guilty verdict, which separates it from the Rule 6(d) violation in Mechanik. In Rose v. Mitchell the Court asked whether a claim of grand jury discrimination should be treated as harmless where the defendant had been found guilty beyond a reasonable doubt by a properly constituted petit jury at a trial free of other constitutional error, and it declined to depart from its long line of decisions requiring that the conviction be set aside and the indictment quashed.⁵⁰ In Mechanik, by contrast, the petit jury's verdict rendered harmless a violation of Rule 6(d), the rule protecting against a charge lacking probable cause.⁵¹ Timing cannot explain the difference, because both cases came after conviction. What differs is the error. The Rule 6(d) violation threatened only the reliability of the probable-cause charging decision, which a verdict of guilt beyond a reasonable doubt confirms, while a defect in grand jury selection compromises the institution itself, and a verdict on guilt says nothing about what a fairly composed body would have done. Students lose this issue when they treat any serious-sounding Rule 6 violation as fundamental, because the Court reserved the presumption for compromised structure and reviewed even the long list of alleged misconduct in Bank of Nova Scotia for its influence on the decision to indict.⁵²

The first step does the real work, since a structural label removes the case from the prejudice inquiry altogether, while every other label sends the claim into step four, where the defendant carries the burden of showing influence on the charging decision.