Invoking the Privilege

Invoking the Privilege

Once the privilege is understood to bind state proceedings, the working question is how a witness puts it to use when an examiner asks something dangerous. The witness in question here is an ordinary one, called before a grand jury, an investigating body, a civil case, or a criminal trial in which the witness is not the accused. Such a witness cannot refuse to appear on Fifth Amendment grounds. The witness must take the stand, hear each question, and claim the privilege as to that question, and the presiding judge decides whether the claim is good.¹ The accused at his own trial holds a broader position, since he may stay off the stand altogether, and the later treatment of Griffin takes that up. Malloy v. Hogan shows the ordinary-witness sequence in a state investigation. Malloy was subpoenaed and appeared, then refused to answer any question on the ground that the answers might tend to incriminate him. When the Superior Court ordered him to testify and he refused again, he was committed to prison for contempt until he answered.²

A lawyer advising a witness in that position asks two practical questions, and the two sections below take them in turn. The first is whether the witness may refuse to answer this particular question. The answer turns on whether the response could tend to incriminate, and the test is deliberately generous. A witness who had to prove the danger of an answer would have to reveal the very information the privilege protects, so the test asks only whether the answer could expose the witness to a criminal case.

The second question is whether, if the government offers immunity, the offer is enough to compel an answer. The privilege reaches only answers that could be used against the witness in a criminal case, so a grant of immunity that truly removes that use removes the ground for refusing.³ Everything then depends on how much protection the grant gives. Malloy itself describes the earlier cases as measuring immunity statutes against the privilege. One invalidated a federal statute whose protection was not coextensive with the privilege, and another held that a state statute falling short of that measure could not be used in a federal prosecution.⁴ Because the privilege extended to the states is the Fifth Amendment privilege itself, that measure now applies to state examinations as well.⁵ The first question comes first, since immunity matters only to a witness who would otherwise have a valid claim.

Questions That Tend to Incriminate

A witness who has taken the stand and heard a question must decide whether the answer is dangerous enough to withhold, and the standard for that decision is generous to the witness. The privilege protects an answer that would directly incriminate, and it protects equally an answer that would supply a link in the chain of evidence needed to prosecute the witness.⁶ In Malloy v. Hogan the Court took its standard for a valid claim from Hoffman v. United States: the claim is good if, from the implications of the question in the setting in which it is asked, a responsive answer, or an explanation of why the question cannot be answered, might be dangerous because injurious disclosure could result.⁷ A question with that tendency to incriminate may be refused, and the judge may overrule the refusal only when it is perfectly clear that the answer cannot possibly have that tendency.⁸ The presiding judge decides by looking at the question and all of the surrounding circumstances, and a reasonable possibility that the answer would form part of the circumstantial evidence against the witness, or lead to such evidence, is enough to sustain the claim.⁹

Each half of that standard has its own reason. The link-in-the-chain reach exists because prosecutions are seldom built on a single admission. They are assembled from small facts such as who employed a person, who paid a bill, or who rented an apartment, and a privilege that covered only confessions would let the government collect those pieces one question at a time. An answer that cannot itself be used in another proceeding can still be such a piece, which is why a party answering a request for admission under Rule 36 keeps the privilege even though the rule bars later use of the admission.¹⁰ The setting-and-perfect-clarity half solves a different problem. A witness who had to prove that an answer is dangerous would often have to give the answer to prove it, and that would destroy the protection the witness came to the stand to claim. So the test reads the question as the examiner framed it, in the inquiry where it was put, and lets the judge reject the claim only on perfect clarity. The same logic lets an innocent witness claim the privilege, since truthful answers from someone who has done nothing wrong can still hand the government incriminating material in the witness's own words.¹¹ The common error is to test the answer by asking whether it amounts to a confession, when a link in the chain is enough.

The setting carried the claim in Malloy. Harold Malloy had been arrested in a 1959 gambling raid and had pleaded guilty to a misdemeanor, receiving a one-year jail term suspended after ninety days and followed by two years of probation.¹² About sixteen months later he was subpoenaed in an inquiry into gambling and other criminal activity in the county, where he was asked about the circumstances of that arrest: who employed him, who paid his fine and his bondsman, who was the tenant of the apartment where he was arrested, and whether he knew a man named John Bergoti.¹³ Taken alone, questions about a closed arrest look harmless, because a conviction ordinarily ends further exposure for the offense it covers.¹⁴ They were not asked alone. They were put in a broad investigation of gambling that was still going on, and answers naming his employer, the person who paid his fine and bondsman, and the apartment's tenant could tie him to a continuing operation and so to a new charge. He was also still on probation, so such answers could matter to the sentence hanging over him as well as to a fresh prosecution. The Court held that his refusal to answer was justified.¹⁵

The Connecticut courts had reached the opposite result. The state's highest court affirmed the contempt order on the footing that the Fifth Amendment privilege was unavailable in a state proceeding and that Malloy had not properly invoked the privilege under the state constitution.¹⁶ The second ruling reflected a narrower state-law view of when a claim is good, and Malloy treated it as error, because once the privilege binds the states the federal standard decides whether the danger is real enough to justify silence.¹⁷ A state therefore cannot jail a witness whose claim satisfies the federal test simply because it fails a stricter local one.

Two limits keep the standard from reaching every uncomfortable question. The danger must be a danger of criminal prosecution, so a witness asked about drug use decades earlier, after the limitation period for that offense has long expired, has no reasonable fear of prosecution and no valid claim.¹⁸ The danger may arise under the criminal law of any government within the United States, so a witness questioned by a state body may rely on a fear of federal prosecution.¹⁹ Fear of prosecution by a foreign nation stands differently, because United States v. Balsys held that it lies outside the Self-Incrimination Clause.²⁰ The Court left open a case in which the United States and foreign governments cooperate so closely that a foreign prosecution is in substance brought on the United States' behalf, but the Third Circuit in In re Impounded treated that language as conditional rather than as a test, so the dependable exam answer is that a foreign-prosecution claim fails.²¹

Because the test attaches to the question rather than to the kind of proceeding, it works the same way in civil cases and on cross-examination. A witness in a lawsuit over a car crash who is asked whether the witness was robbing a store in another city on the day of the accident may claim the privilege even though no robbery charge has ever been filed, since the answer might be used in a later prosecution. The same is true of cross-examination about uncharged misconduct offered to attack credibility, and testifying on another matter does not waive the privilege for questions that go only to the witness's character for truthfulness.²² The claim must actually be made, because a witness who answers without claiming it gives up the privilege as to that answer, which may then be used freely to assess credibility.²³ A defendant who takes the stand is generally treated as waiving the privilege on the charged matters but keeps it for other crimes that have not yet produced a conviction.²⁴

Immunity as a Substitute

A valid claim of privilege stops the examiner only until the government offers something in its place. The privilege protects against answers that could be used in a criminal case, so a grant of immunity that truly removes that danger also removes the ground for refusing, and the witness must then answer or face contempt.²⁶ Everything turns on how much danger the grant must remove. Immunity statutes strike a rational accommodation between the privilege and the government's need for testimony, and the accommodation is fair only if the witness receives protection that matches what the witness gives up.²⁷ The measure is protection coextensive with the privilege, and it applies to state grants because the privilege that binds the states is the Fifth Amendment privilege itself.²⁸

Coextensive protection has two parts, and each closes a different gap. The first part bars direct use, so the compelled answers themselves cannot be offered against the witness in a criminal case. The second part bars derivative use, so evidence found by following those answers cannot be offered either. The privilege covers both because it protects against disclosures the witness reasonably believes could be used in a prosecution or could lead to other evidence that might be so used.²⁹ A grant limited to the first part leaves the witness exposed. The government can read the testimony for leads, locate the same facts through other witnesses or documents, and prosecute with evidence that exists only because the witness spoke, all without offering a word of the transcript. A bar on direct use alone therefore leaves the witness no safer than before, since evidence found by following a compelled answer is as much the product of the compulsion as the answer is.

Malloy v. Hogan applied this measure to a state statute. The Court drew on Counselman v. Hitchcock, which invalidated a federal immunity statute whose protection fell short of the privilege, and on Adams v. Maryland, which refused to let a state immunity statute of the same kind support a federal prosecution.³⁰ Connecticut's statute provided only that a witness's answers before the investigating body could not be used in criminal proceedings against the witness, except in a prosecution for perjury, and it did not bar using the answers to discover other evidence.³¹ That protection stops at direct use, so it was not coextensive with the privilege, and Malloy could still refuse to answer. In this setting a result for the witness means the immunity is inadequate: the privilege survives, and the state cannot keep the witness in jail for silence.

Changing a single feature of Connecticut's statute reverses the result.

Kastigar v. United States, decided in 1972, tested the federal immunity statute against the same measure.³² Section 6002 provides that when a witness is compelled by court order to testify over a privilege claim, no compelled testimony or information, and nothing directly or indirectly derived from it, may be used against the witness in any criminal case, except in a prosecution for perjury, giving a false statement, or failing to comply with the order. The Court held that this protection against use and derivative use is coextensive with the privilege and suffices to supplant it.³³ The same perjury carve-out appears in Connecticut's statute, so the exception is compatible with coextensive protection, and the missing derivative-use bar was the defect in the state's version. The witnesses in Kastigar had demanded transactional immunity, which shields the witness from prosecution for any transaction the testimony touches, and the Court refused because the privilege has never meant that a witness who invokes it can never be prosecuted; its concern is protection against being forced to give testimony that leads to penalties.³⁴ Justice Douglas, dissenting, would have required the broader protection, reasoning that immunity is adequate only if it operates as a complete pardon, since use immunity permits the compulsion without removing the criminality.³⁵ The majority view is the law. Either kind of immunity suffices to compel testimony, and legislatures remain free to grant transactional immunity although the Constitution does not demand it.³⁶

The derivative-use bar would mean little if the witness had to prove that the government used the testimony. The government alone knows where its evidence came from, so the Court placed the burden on it: once a witness shows testimony under a grant of immunity, the government must show that its evidence comes from a legitimate source wholly independent of the compelled testimony, and the Court called that burden heavy.³⁷

Once adequate immunity is in place, refusal is punishable. In United States v. Cintolo, a grand jury witness named LaFreniere received use immunity in a loan-sharking investigation but, on his lawyer's advice, kept refusing to testify, and he served eighteen months in prison for contempt, because a witness granted use and derivative-use immunity has had the privilege fully satisfied.³⁸ Immunity does not give the witness a pass from prosecution. A criminal case built on independent evidence remains possible.³⁹ The privilege speaks only to use in a criminal case, so coextensive immunity must bar criminal use and need not bar use in a noncriminal administrative proceeding.⁴⁰

The protection must also travel across governments. If a state's grant bound only state prosecutors, a witness compelled under it would still face a federal prosecution built on the answers, and the danger the privilege guards against would remain. For that reason a grant by one government within the United States must be honored by every other, which may not use the immunized statements or their fruits, and the leading decision is Murphy v. Waterfront Commission.⁴¹ The other government stays free to prosecute on independent evidence, so State B may prosecute a witness immunized by State A as long as it does not use the compelled testimony or its fruits.⁴²

The same measure governs public employees who are threatened with discharge for refusing to answer. An employee ordered to answer narrow questions directly related to official duties may be discharged for refusing if the employee has been given immunity from use of the answers and their fruits in a criminal prosecution, as in Gardner v. Broderick and Lefkowitz v. Turley. Where the employer demands answers without that immunity, the answers cannot be used in a later criminal prosecution by state or federal authorities, as Garrity v. New Jersey and Murphy held.⁴³