Blog
Miranda Rights & Federal Interview
The Interview That Is Not an Interrogation
Federal agents do not read Miranda warnings during most of the interviews that produce criminal convictions. The warning itself, perhaps the most recognized sentence in American criminal procedure, applies to a circumstance most people will never encounter in the way they imagine: the custodial interrogation, conducted in a room from which the subject cannot leave, by officers whose authority is both visible and constraining. The typical federal interview resembles none of this. Two agents arrive at a place of business on a Tuesday morning, or at a residence before the household has fully organized itself for the day. They present credentials. They ask whether the subject would be willing to answer a few questions. They mention, in tones that are meant to convey reasonableness, that the subject is not under arrest and is free to decline. The subject, who has never been contacted by federal law enforcement and who processes the encounter through whatever partial understanding of criminal procedure was acquired from television, agrees to speak. From that point forward, everything the subject says is admissible.
The encounter has been structured, with considerable precision, to avoid the condition that would require a warning. The agents know the doctrine. They have been trained in it. The choice to approach the subject at home or at work rather than at a field office, the verbal confirmation that the subject is free to leave, the absence of drawn weapons or physical restraint: these are not courtesies. They are procedural architecture. In Beckwith v. United States, the Supreme Court held that an IRS interview conducted at a private residence was non-custodial even though the investigation had focused on the taxpayer, and even though the agents who arrived at eight in the morning came precisely so the taxpayer would not be embarrassed at his place of employment. The Thompson v. Keohane test asks whether a reasonable person would have felt free to terminate the encounter and leave. Federal courts apply that test, and they apply it in a manner that tends to favor the government’s position.
What a person experiences and what the law recognizes as custody are different things. The two agents at the kitchen table, the credentials on the counter, the sense that declining to speak would signal guilt: none of this constitutes custody. A person can feel that the conversation is inescapable and still be classified, for constitutional purposes, as free. The doctrine does not measure what the subject felt. It measures whether physical restraint was present.
Whether this gap between experience and doctrine is an oversight or a feature is a question the case law has not resolved.
One thing is certain: the decision to speak or remain silent is made, by most people, in the first forty seconds of the encounter, before any understanding of the consequences has formed.
Custody and Its Absence
The legal definition of custody turns on restraint of movement, not on the character of the questions being asked or the severity of the investigation. Rhode Island v. Innis defined the functional equivalent of interrogation as words or actions that officers should know are reasonably likely to elicit an incriminating response. But the custody inquiry is separate.
Interviews in FBI vehicles have been found non-custodial where the door was unlocked. Interviews at field offices have been found non-custodial where the subject arrived voluntarily and was told the door was open. The analysis depends on details the subject often did not notice at the time, like whether the agent’s weapon was holstered or whether the door to the room was open or closed.
In the substantial majority of federal cases involving statements obtained from interviews, the government’s position is that the encounter was non-custodial. The subject arrived voluntarily. The subject was told, at least once, that participation was optional. The subject remained in a location from which departure was physically possible. By the time a defense attorney examines the circumstances, the record is composed almost entirely of the government’s account, because the subject had no counsel present and no recording exists.
The distinction between custody and non-custody is not a technicality.
18 U.S.C. § 1001 and the Form 302
The interview, conducted without counsel and without recording, is the environment in which a second and independent crime can be committed. Under 18 U.S.C. § 1001, any materially false statement made to a federal agent is a felony carrying a sentence of up to five years. The statement need not be made under oath. The interview need not be custodial. The subject need not have been warned that dishonesty carries criminal consequences. A casual denial on a front porch satisfies the statutory elements, provided the denial is material and the government can establish that the subject knew it to be false.
The mechanism by which such a statement is documented deserves particular attention. FBI interviews are, as a general matter, not recorded. One agent conducts the questioning; a second takes notes. Those notes are memorialized in a Form FD-302, prepared sometimes hours and sometimes days after the conversation. The 302 is the agent’s summary of what the subject said, rendered in the agent’s language, shaped by the agent’s understanding of the investigation and filtered through whatever the agent’s notes managed to capture of an hour-long conversation that the subject believed was informal. The 302 becomes the evidentiary record. If the subject’s recollection of what was said differs from the 302, the subject faces the practical reality that two federal agents, whose testimony carries the presumption of credibility in most courtrooms, will describe what was said, and the subject has no recording to contradict them.
The agent’s notes do not capture hesitation. The agent’s summary does not preserve the qualifier “I think” or “I am not sure.” What arrives at trial is a definitive statement attributed to the subject, authored by the government.
The Stewart prosecution remains the clearest illustration of the pattern, though it is far from the only one. The government could not establish the underlying securities violation. What it could establish was that the subject made statements during interviews that were inconsistent with other evidence. The false statements charge became the conviction. This is, if we are being precise, not an aberration. It is a prosecutorial methodology. When the original conduct cannot be proven to the satisfaction of a jury, the interview itself generates the offense.
I am less certain about the frequency of this pattern than the preceding paragraph might suggest, because the data on how many § 1001 charges originate from otherwise non-prosecutable investigations is not aggregated in any source I have found reliable. What I can say is that the cases I have reviewed follow the same architecture with sufficient regularity that the pattern warrants identification.
One does not need to have committed the underlying offense to be convicted of a false statement.
What Miranda Protects and What It Does Not
Vega v. Tekoh, decided in 2022, held that a failure to provide Miranda warnings does not, standing alone, give rise to a civil claim under 42 U.S.C. § 1983. The majority characterized the Miranda rules as prophylactic: required to safeguard the Fifth Amendment right, but not themselves constitutional rights. The practical consequence is that an officer who conducts a custodial interrogation without warnings and obtains statements that are used at trial cannot be sued for damages. The remedy remains exclusion of the un-Mirandized statement from the prosecution’s case in chief. There is no further remedy.
The dissent in Vega stated the objection plainly enough that it needs no elaboration here, except to note that the word “prophylactic,” which the majority employs, does considerable work in the opinion: it permits the Court to characterize Miranda as something less than a right while preserving it as something more than a recommendation. The contract was enforceable. It was also, in its application, insubstantial.
For the person who has already spoken to federal agents without counsel, the question of remedy is immediate and narrow. Exclusion operates only if the interview was custodial and the warnings were not provided. For the vast majority of federal interviews, which are structured as non-custodial, the question never arises.
Waiver, Invocation, and the Ambiguity Problem
Even when Miranda warnings are required and administered, the protections they offer are thinner than most people suppose. In Berghuis v. Thompkins, the Supreme Court held that a suspect who sat in an interrogation room for nearly three hours, largely silent, and then made incriminating statements had implicitly waived his Miranda rights. The standard for waiver is that it must be voluntary, knowing, and intelligent.
The invocation of Miranda rights must be unambiguous. Davis v. United States established that a suspect’s statement that he “might want a lawyer” was insufficient to require the cessation of interrogation. The rule demands clarity at a moment when clarity is precisely what the circumstances erode. A person seated across from federal agents, aware that the conversation concerns criminal conduct but uncertain of the scope of the investigation, is not in an optimal position to produce the kind of declarative sentence the case law requires. “I want an attorney” is sufficient. “Maybe I should talk to a lawyer” is not. The distance between those two sentences is the distance between the protection applying and the protection failing, and most people do not know where the line falls until they have already crossed it.
And our practice in this area (and it is, I acknowledge, a departure from what some practitioners advise) is to instruct clients that if there is any doubt about whether to invoke, the answer is to invoke. The equivocal statement is treated by the courts as no statement at all. The clear invocation cannot be ignored. The cost of precision at that moment is zero. The cost of imprecision can be measured in years.
There is a particular silence in a conference room after a client describes what they said to agents. It is the silence of recognizing that the words cannot be retrieved.
The Procedural Sequence
A person contacted by federal agents for an interview should understand the following sequence, which is not intuitive and which no agent will explain.
- The subject has no obligation to agree to an interview. Declining is not obstruction. It is the exercise of a right that federal agents are trained to discourage but cannot penalize.
- If the subject agrees to speak, every statement is admissible and every false statement is independently prosecutable under § 1001.
- The subject may end the interview at any time, even after answering questions for an hour.
- If the subject is placed under arrest, the encounter becomes custodial and Miranda warnings are required before further questioning.
- If warnings are provided, the subject may invoke the right to counsel. The invocation must be clear and unambiguous.
The firmest protection available is the one that requires the least from the person asserting it: silence. Not partial silence. Not qualified silence. The kind in which no information passes in either direction until an attorney is present and the terms of the conversation have been established.
Representation and Its Timing
Most people contact an attorney after the interview. This is understandable and it is also the wrong sequence. The value of representation is concentrated in the period before any statements are made, because the statements, once documented in a 302, are permanent. A 302 can be challenged at trial, but the challenge requires the defense to contradict the recollection of two trained agents with nothing more than the defendant’s memory. An attorney present during the interview alters the dynamic entirely, though even this is, in most cases, less protective than declining the interview outright.
We approach the initial contact from federal agents as the beginning of an adversarial process, not as its prelude, because what we have seen when clients treat the interview as preliminary is that the interview becomes the evidence that forecloses every subsequent option. The timing of representation is not a question of convenience. It is the single decision that determines what the government possesses when it evaluates whether to indict.
A consultation assumes nothing and costs nothing. It is the point at which the facts can be assessed, the scope of the investigation estimated, and the range of available responses identified before any of them have been foreclosed by a conversation in a kitchen conducted without the benefit of counsel.
The architecture of federal enforcement is designed to produce speech. Miranda was designed to protect silence. The two structures meet, when they meet at all, only in the narrow corridor of custodial interrogation. For everything that occurs outside that corridor, the protection is the one you carry with you: the willingness not to speak.