Derivative Evidence

Derivative Evidence

The preceding sections dealt with the statement itself: whether the prosecution may offer it in its case in chief, and whether it may be used to contradict a defendant who testifies. Harris and the cases that followed it all ask what use may be made of a single excluded item.¹ Derivative evidence raises a separate question, because police sometimes treat an unlawfully obtained statement as a lead and find something new, such as a witness, a weapon, or a later confession. The new item is a different piece of evidence, and the prosecution may want to offer it to prove guilt. The rules about how the statement may be used therefore do not settle whether the new item may come in. The defense argues instead that the item is tainted by the way police found it, and the court must decide how far the original illegality reaches.

The question needs its own rules because both obvious answers fail. If exclusion stopped at the statement, police could extract it, follow it to the physical or testimonial evidence it reveals, and offer that evidence at trial, which would leave them with most of what the violation was meant to gain. If exclusion reached everything that could be traced back to the violation, the defendant would receive far more than is needed to remove that advantage. The decisions accordingly stop short of treating every causal link as enough, and the doctrine does not extend as far as a simple but-for test would carry it.² The sections ahead take the general fruits rule first, then the independent-source and purged-taint limits on it, then the way noncoercive Miranda violations fit into that framework, and finally inevitable discovery as a limit of its own. The first step is the baseline rule that evidence found by exploiting an illegality is generally inadmissible.

Fruit of the Poisonous Tree

The baseline rule for derivative evidence is that exclusion does not stop with the item police took unlawfully. Evidence discovered by exploiting an illegality is generally inadmissible as the fruit of the poisonous tree, the image being that a tainted source contaminates whatever grows from it.³ In Wong Sun v. United States the Court framed the controlling question as whether, granting that a primary illegality occurred, the evidence now challenged "has been come at by exploitation of that illegality or instead by means sufficiently distinguishable to be purged of the primary taint."⁴ That question looks to how police reached the evidence, so it applies to a witness, a weapon, or a later confession just as readily as to an object seized in the original search. The preceding sections asked what use may be made of one excluded statement; this section asks how far the exclusion reaches into the items that the statement or search produced.

The reason for extending exclusion to fruits comes from what the primary exclusion is meant to accomplish. The exclusionary rule rests on two purposes, deterring police misconduct and keeping courts from lending their authority to the products of a violation, and the limits on the fruits doctrine are drawn by asking whether either purpose still calls for exclusion.⁵ On deterrence, an officer who lacks a lawful basis to search or question usually wants the information in order to follow it somewhere, and the item itself may be of little value at trial. If the court suppressed only that item, the officer could still collect the lead, pursue it to a witness or a weapon, and offer the result, so the reward for the violation would survive one step downstream. Excluding the fruit removes that reward and the incentive to cut corners that comes with it. The integrity purpose points the same way, because a court that admitted the fruit would be accepting a conviction built on evidence reached through the very violation it had just condemned.

The narcotics sequence in Wong Sun shows the doctrine running through two links. Narcotics agents entered a dwelling without probable cause, chased down and arrested the first suspect, and heard that suspect almost immediately accuse a second person of selling narcotics. Agents then seized narcotics from the second person. The Court excluded both the first suspect's statement and the narcotics later seized from the second person as products of the illegal entry and arrest.⁶ The statement was the direct product of the illegal arrest. The narcotics sat one step further out, but agents reached the second person only because the statement pointed there, so the drugs were obtained by exploiting the illegality and not by any means distinguishable from it. Wong Sun arose under the Fourth Amendment, and the same causal analysis has been applied under the Sixth Amendment to an in-court identification challenged as the fruit of a lineup that violated that right.⁷ How the analysis fits a Miranda defect is taken up in a later section.

Fruit status does not follow from causation alone. Wong Sun also involved a third suspect, who was implicated by the second, arrested illegally, arraigned, and released on his own recognizance. Several days later he returned and voluntarily confessed to a narcotics agent. The Court refused to treat that confession as the fruit of the illegal arrest, because the connection between the arrest and the statement had become so attenuated that the taint dissipated.⁸ The suspect might never have confessed had he never been arrested, so but-for causation was present, yet the confession came in. His voluntary return after release was an independent act that broke the chain. Both statements followed an illegal arrest, and what separated them was how directly each flowed from the illegality: immediately in the first suspect's case, and across days and a free choice in the third suspect's. The doctrine therefore reaches evidence obtained by exploiting an illegality and stops short of everything that a but-for test would sweep in.⁹ Students lose this issue when they treat but-for causation as enough to make evidence fruit.

Fruit status is accordingly the starting point of the analysis. Once the challenged evidence is traced to a violation, the question becomes whether one of the limits on the doctrine applies, and those limits are the subject of the next sections, beginning with the independent source. Because the first job is tracing, the inquiry is best run in order.

Independent Source

The first limit on the fruits doctrine asks whether the prosecution reached the challenged evidence by a route that never ran through the violation. Evidence discovered by means wholly independent of the constitutional violation is admissible, a rule that Nix v. Williams attributes to Silverthorne Lumber Co. v. United States.¹¹ The independent source limit follows from what exclusion is meant to accomplish. Suppression is supposed to leave the government no better off than it would have been had the violation never happened. When the government actually obtained the same evidence through a separate lawful channel, the violation contributed nothing to the prosecution's case. Excluding the evidence anyway would leave the government worse off than no violation would have, and the defendant would receive a windfall that has no connection to the misconduct. The deterrence logic points the same way, because an officer who cut a corner gains nothing from it when a lawful channel independently delivered the item, so suppression would have no benefit of the violation to take away.

A decided case shows the same structure. In Murray v. United States officers illegally entered a warehouse they suspected of being the hub of a drug operation and saw bales of marijuana inside. They did not seize the bales. They withdrew and applied for a warrant, taking care not to cite anything they had learned from the entry, and when the warrant issued they reentered under its authority and found the marijuana again. The Court held that the warranted search, which in no way exploited the illegal entry, was an independent source that erased the taint of the violation.¹² The decision shows what "wholly independent" demands. The lawful channel must not draw on the violation, since a warrant application built on what the officers saw during the illegal entry would be a continuation of the first channel with a lawful label attached. The lawful channel must also have produced the evidence in fact. Students lose this issue when they credit a lawful route that could have reached the evidence but did not.

That last requirement separates independent source from the two limits that sit beside it. Inevitable discovery, which later sections develop, has been described as a "hypothesized independent source," because the prosecution there cannot point to a lawful channel that produced the evidence and instead proves that one would have.¹³ In Nix v. Williams the body was found because Williams led the police to it, so the Court observed that the independent source exception "would not justify admission of evidence in this case," and it reached the same result through the extension that its rationale supports.¹⁴ The purged-taint limit differs in a different way. It accepts that police reached the evidence through the illegality and asks whether the link had become too attenuated to count as exploitation, whereas independent source denies that the evidence came through the illegality at all.¹⁵ The next section takes up that attenuation limit.

Purged Taint

The independent source limit denies that the evidence came through the violation at all. The purged-taint limit starts from the opposite premise: police did reach the evidence because of the illegality, and the evidence would not have surfaced without it, yet exclusion may still be unwarranted. The decisions make clear that the poisonous tree doctrine does not extend as far as a but-for causation test would carry it.¹⁷ What takes the place of but-for causation is the question posed in Wong Sun v. United States: granting the primary illegality, did police come at the challenged evidence by exploiting that illegality, or by means sufficiently distinguishable to purge the primary taint?¹⁸ This limit, called purged taint and also described as the attenuated-connection limitation, arises most often under the Fourth Amendment but has been applied under the Sixth Amendment as well.¹⁹

The limit exists because both purposes of the exclusionary rule stop calling for exclusion once the means of acquiring the evidence are substantially removed and distinguishable from the initial illegality.²⁰ Deterrence is the easier purpose to follow. An officer is tempted to cut a corner when the violation reliably delivers evidence, and suppression takes that payoff away. When the evidence reaches the prosecution only because a free person makes a choice of their own, or because an unrelated fact turns up, the officer could not have counted on that result when committing the violation, so suppressing it adds little to the deterrence already supplied by excluding the direct products of the illegality. The judicial-integrity purpose follows the same path, because a court that admits attenuated evidence is not resting a conviction on the violation itself. The cost of the opposite rule is also real: if every item in a but-for chain were excluded, the defendant would receive protection far out of proportion to the violation, and probative evidence would leave the trial although the police gained no advantage from the misconduct.

Wong Sun supplies both poles out of one set of narcotics arrests. Agents entered a dwelling without probable cause and arrested the first suspect, who almost immediately accused a second person, and narcotics were then seized from that second person. Nothing intervened between the illegal arrest and either item, so the Court excluded the statement and the narcotics as products of exploitation.²¹ The third suspect had also been arrested illegally, but he was arraigned and released on his own recognizance, and several days later he returned and voluntarily confessed to an agent after being warned of his rights. The Court refused to treat that confession as fruit, because his voluntary action after release had made the connection to the arrest so attenuated that the taint dissipated.²² He might never have confessed had he never been arrested, so but-for causation was present, and the confession came in anyway. The fact that moved the answer was an intervening independent act: when the defendant or a third party does something of their own accord that breaks the causal chain, the evidence that follows is no longer viewed as obtained by exploiting the illegality.²³

The intervening event also need not be a free choice by the defendant. In Utah v. Strieff an officer made a stop that the defendant challenged as unlawful, learned during it that an arrest warrant for him was outstanding, arrested him on that warrant, and found incriminating evidence in the search incident to the arrest.²⁴ Without the stop the officer would never have learned of the warrant, so but-for causation was plainly present. The Court nonetheless held that the evidence was not subject to exclusion, concluding that "the discovery of a valid arrest warrant was a sufficient intervening event to break the causal chain between the unlawful stop and the discovery of drug-related evidence on Strieff's person."²⁵ The warrant was valid, it predated the stop, and it was entirely unconnected to the stop, which made it a circumstance that the police had not manufactured through their misconduct.²⁶

Strieff also shows that no single circumstance decides the question. The Court acknowledged that the close timing of the stop and the search supported suppression, treated the warrant as an intervening circumstance pointing the other way, and then considered "the purpose and flagrancy of the official misconduct," which it found to be at most negligent, a simple mistake about what the law permitted.²⁷ The third consideration tracks the deterrence purpose: a deliberate violation is the conduct exclusion is meant to discourage, while an honest mistake leaves suppression with less to deter. Students lose this issue when they let closeness in time settle the question, because temporal proximity favored suppression in Strieff and the evidence was admitted anyway.²⁸ The next section applies this framework to the unwarned statement with which the chapter began.

Miranda Violations and Fruits

The fruits doctrine traced in the preceding sections starts from a constitutional violation, because evidence can be the fruit of an illegality only if there was an illegality. That starting point is what changes in the Miranda setting. A custodial statement taken without warnings and a valid waiver stays out of the prosecution's case in chief under an irrebuttable presumption that the questioning compelled it, a presumption that sweeps more broadly than the Fifth Amendment privilege itself.³⁰ When the statement was in fact voluntary, the police departed from a prophylactic rule, meaning a protective rule built around a right to keep that right from being violated, and did not abridge the privilege.³¹ The Court has therefore declined to treat the downstream consequences of a noncoercive Miranda violation as a Wong Sun problem, and its decisions form a sequence that this section follows in order: a witness, a later warned confession, and the limit on that second step.³²

The refusal has a reason on each side of the line it draws. The Fourth Amendment version of the fruits doctrine aims to deter unreasonable searches however probative the resulting evidence, so taking the fruits away takes away what the illegal search yielded.³³ The Fifth Amendment aims to bar compelled statements, and the Miranda presumption already protects that aim by keeping the unwarned statement out of the case in chief. Carrying the presumption further would suppress evidence in cases where nobody compelled anything. In deciding how sweeping the consequences of a failure to warn should be, the Court noted that suppressing the derivative evidence would advance neither the deterrence of improper police conduct nor the Fifth Amendment's interest in trustworthy evidence.³⁴ The supporting theory is that excluding every non-attenuated fruit of a Miranda violation is unnecessary to deter police from skipping warnings.³⁵

The first application involves a witness. In Michigan v. Tucker, police learned the identity of a witness from the defendant's statement, which followed a good-faith failure to give warnings, and the Court held the witness's testimony admissible in the prosecution's case in chief.³⁶ The unwarned questioning had “departed only from the prophylactic standards” that Miranda laid down, so the rule requiring suppression of the fruits of a constitutional violation did not control.³⁷ Two assumptions support the result. Police are not expected to put a non-suspect witness through the coercive process of custodial interrogation, so the person who testifies is not someone the Miranda presumption exists to protect, and excluding the testimony would add little to the deterrence of failures to warn.³⁸ A witness found through a voluntary but unwarned statement therefore testifies for the prosecution like any other witness.

The second application involves the defendant's own later statement. In Oregon v. Elstad two officers questioned the defendant at his home without warnings, and when they said they believed he had been involved in a burglary he answered, “Yes, I was there.”³⁹ Everyone accepted that this remark was excludable. At the stationhouse he was then warned, waived his rights, and gave an extensive statement describing his part in the burglary, and the Court held that the fruits doctrine did not make that second statement inadmissible.⁴⁰ The inquiry shifts to “whether, in fact, the second statement was also voluntarily made,” judged by the surrounding circumstances and the entire course of police conduct.⁴¹ This extends the reasoning of Tucker: the prophylactic part of Miranda is carried no further than barring the unwarned statement from the case in chief, and a properly warned later statement stands on its own voluntariness, just as an unwarned voluntary statement may still be used to impeach.⁴²

The defendant's best argument against that result was psychological. The Oregon court had reasoned that a suspect who has blurted out an admission feels that he has “let the cat out of the bag” and sealed his fate, so that only the passage of time and a change of place could dissipate the pressure of that thought on the later confession.⁴³ The Supreme Court rejected the argument because the lingering pressure comes from the suspect's own voluntary admission, and no official coercion of the kind the Fifth Amendment forbids played a part in either statement. Giving constitutional weight to that psychological effect would, as a practical matter, disable police from obtaining a suspect's informed cooperation after any slip in the warnings, while a suspect who has been warned remains free to decide whether to speak.⁴⁴ Time, place, and a change of interrogator do matter when the earlier statement was actually coerced, because only then is there compulsion that could carry over into the second confession.⁴⁵

Elstad does not make every warned confession safe. In Missouri v. Seibert the Court held that when police intentionally interrogate without warnings to obtain a confession, then give warnings and obtain the confession a second time, the second confession is inadmissible.⁴⁶ The mechanism is the inference the sequence invites: the process leads to “an entirely reasonable inference” that what the suspect had already said will be used, “with subsequent silence being of no avail.”⁴⁷ Warnings exist to let a suspect decide whether to speak, and warnings that arrive only after the full confession cannot perform that function, which is why midstream warnings may be inadequate to convey the Miranda rights.⁴⁸ Seibert therefore does not apply the fruits doctrine. Its analysis asks whether the midstream warnings could do their job, and the fruits doctrine stays inapplicable to a simple Miranda violation.⁴⁹ The line between the two cases is police strategy, since absent deliberately coercive or improper tactics in obtaining the first statement, the mere fact of an unwarned admission does not justify presuming compulsion as to the second.⁵⁰

The relevant factors compare the two rounds of questioning on five points.⁵¹

The same factors explain why many two-statement sequences fall outside Seibert. In Bobby v. Dixon the defendant made an unwarned admission to forgery and later, after warnings, confessed to murder, and the Court found no Seibert violation because there was “no nexus” between the two statements and a “significant break in time and dramatic change in circumstances” separated them.⁵² Those findings correspond to the overlapping-content and timing-and-setting factors, and they show what a sufficient break looks like: the second statement addresses a different offense and follows a real interval in a changed setting.

The logic of Tucker has also been applied to physical evidence. In United States v. Patane the defendant was questioned in violation of Miranda and made incriminating statements that led officers to a gun, and the gun was held admissible.⁵³ The decision produced no majority. Three Justices reasoned that excluding the unwarned statement is a “complete and sufficient” remedy for a Miranda violation, and two others agreed that admitting nontestimonial physical fruits, meaning objects rather than statements, does not risk admitting an accused's coerced statements against him, while declining to decide whether anything remained to deter.⁵⁴ Because of the fractured opinions the point is mildly unsettled, but most lower courts hold that no fruits doctrine flows from a Miranda violation, at least for physical evidence.⁵⁵ The result depends on the absence of “actual coercion,” because Miranda requires exclusion of the physical fruit of an actually coerced statement, whereas a Miranda-defective statement is only presumed to have been coerced.⁵⁶

That contrast marks where the Miranda rule stops. Investigative fruits of statements found involuntary are excluded unless the prosecution establishes an independent source or inevitable discovery, while witnesses and physical evidence discovered through Miranda-defective statements are admissible.⁵⁷ A statement taken in violation of the Sixth Amendment right to counsel is treated differently again, because that violation occurs at the moment counsel is denied, so the statement can serve as a poisonous tree. In Brewer v. Williams the Court recognized as much by requiring that, on retrial, the question be resolved whether evidence of the body's location was fruit of a confession obtained in violation of that right.⁵⁸ The next section follows that case to its second trial. On an exam stem, the sorting question is whether the underlying statement was coerced, taken in violation of the Sixth Amendment, or merely unwarned, and for the merely unwarned statement, whether police used a deliberate two-step. Students lose this issue when they run Wong Sun against a voluntary but unwarned statement as though its witnesses, later confessions, and physical fruits were automatically tainted.⁵⁹