State v. Delong

¶1dissenting.

¶2I respectfully disagree with the majority’s conclusion that defendant’s consent to the search of his vehicle attenuated the taint of the Miranda violation by the deputies in this case. In particular, I disagree with the majority’s conclusion that defendant’s consent — -given in response to police questioning while he was in custody and in handcuffs — was volunteered. I disagree with the reasoning that the majority has used to reach its conclusion, and I am concerned about how that reasoning may affect other cases in which citizens are interrogated by law enforcement officers under compelling circumstances in violation of their right to remain silent under Article I, section 12, of the Oregon Constitution.1

¶3I.

¶4I begin with a brief additional background. Sergeant Robeson stopped defendant because defendant was not wearing his seat belt while operating his automobile. After Robeson activated his overhead lights and pulled defendant over, defendant identified himself by name and date of birth but did not provide Robeson with a driver’s license. Robeson handcuffed defendant, searched him, and placed him in the backseat of his patrol car. Robeson testified that his purpose in taking defendant into custody was to establish defendant’s identity. While defendant was handcuffed in Robeson’s patrol car, Robeson questioned him about illegal activity unrelated to the stop without first warning him that *399he had a right to remain silent. Robeson asked defendant if there was anything in defendant’s car that the deputies should be concerned about. According to Robeson, defendant “told me ‘no,’ and that if we wanted to search the vehicle, we could.” Robeson testified that the above encounter lasted “maybe two or three minutes.” Deputy Poe searched defendant’s vehicle immediately after Robeson’s conversation with defendant and found a fanny pack that contained drug paraphernalia and drug residue. Poe did not give defendant Miranda warnings until after his search had disclosed the physical evidence of drugs.

¶5At the suppression hearing, defendant agreed that he had told Robeson that Robeson would not find anything in his car, but he denied that he had consented to the search of his car. The trial court found that

“during this period of time … is when Sergeant Robeson initiated the conversation with the defendant about does he have anything of concern in his vehicle, and the defendant responded no, and then ultimately gave consent to search. Deputy Robeson said he did not ask for consent. That it was volunteered in a way by the defendant and the Court finds that appears to be credible.

«H? Hí ❖ H« ❖

“ [I] t was after the search when Deputy Poe approached the defendant, he gave him his Miranda rights and after giving those Miranda rights did ask extensive questions about what he found in the car and statements — incriminating statements were made by the defendant. The Court is finding that the prior conversation about identification while he was detained to pursue that investigation, that at that point Miranda wasn’t needed.”

¶6On appeal, the state conceded that the trial court had erred and that Miranda warnings were required before Robeson could question defendant because defendant had been in custody and under compelling circumstances at the time. The state nevertheless argued that suppression of the physical evidence was not required, because, in its view, defendant had made a spontaneous offer of consent to search his car. Defendant argued that State v. Vondehn, 348 *400Or 462, 475-76, 236 P3d 691 (2010), supported his contention that his consent (as found by the trial court), other statements he had made, and the physical evidence all derived from the Article I, section 12, violation and should have been suppressed.

¶7The Court of Appeals concluded that this case “is largely governed by the principles and reasoning that the Supreme Court set forth in Vondehn” and reversed the trial court’s denial of defendant’s motion to suppress. State v. Delong, 260 Or App 718, 722-23, 320 P3d 653 (2014). In Vondehn, the defendant was likewise handcuffed, placed in a patrol car, and briefly interrogated without the benefit of Miranda warnings, in violation of his rights under Article I, section 12. In response to a deputy’s question, the defendant admitted that he owned the backpack in the car and that it contained marijuana. The defendant consented to a search of the backpack, marijuana was found, and the officers then gave the defendant his Miranda warnings. This court observed that it had long held “that the Oregon Constitution requires suppression of statements made without the benefit of Miranda warnings.” Vondehn, 348 Or at 472. The court held that “the state is precluded from using evidence derived from the violation to obtain a criminal conviction,” including “physical evidence that is derived from that constitutional violation.” Id. at 475-76.

¶8This court also explained in Vondehnthat the rationale for the constitutional requirement that police warn citizens of their right against self-incrimination is the level of coercion inherent in custodial interrogations. Article I, section 12, protects citizens against the use of compelled statements, because such statements do not provide an acceptable basis for proving guilt of a crime in a civilized society. See State v. Mendacino, 288 Or 231, 236, 603 P2d 1376 (1979) (so stating). To give effect to that constitutional right, this court has prohibited the state from using physical evidence derived from an Article I, section 12, violation to prosecute a suspect. The principles of law that this court applied in Vondehnare highly pertinent to this case:

“Since Magee, this court consistently has held that the Oregon Constitution requires suppression of statements *401made without the benefit of Miranda warnings. See, e.g., State v. Roble-Baker, 340 Or 631, 643-44, 136 P3d 22 (2006) (suppressing unwarned statements made during custodial interrogation); State v. Smith, 310 Or 1, 7, 791 P2d 836 (1990) (so stating). The full extent of the court’s discussion of the rationale for that rule has been to state that, when a suspect is subjected to custodial interrogation, warnings are necessary “‘because of the inherent level of coercion that exists in such interrogations.’” State v. Scott, 343 Or 195, 200, 166 P3d 528 (2007) (quoting State v. Joslin, 332 Or 373, 386, 29 P3d 1112 (2001)); see also State v. Meade, 327 Or 335, 339, 963 P2d 656 (1998). …

‡ ‡ *

“Article I, section 12, affords a constitutional right to remain silent. That right is, however, subject to waiver. Because a custodial interrogation is inherently compelling, and to ensure the validity of a waiver of the right against self-incrimination, Article I, section 12, requires that the police inform a person subjected to custodial interrogation that he or she has a right to remain silent and to consult with counsel and that any statements that the person makes may be used against the person in a criminal prosecution. Article I, section 12, requires those Miranda warnings to ensure that a person’s waiver is knowing as well as voluntary. If the police conduct a custodial interrogation without first obtaining a knowing and voluntary waiver of the suspect’s rights, then they violate the suspect’s Article I, section 12, rights. To give effect to those constitutional rights, the state is precluded from using, in a criminal prosecution, statements made in response to the interrogation.”

¶9348 Or at 472-74.

¶10In applying those principles to the facts in Vondehn, this court concluded:

“As noted, defendant was in custody, in the back seat of a patrol car and handcuffed, when the police subjected him to custodial interrogation. Defendant had the right to remain silent and to advice of counsel, but the police conducted their custodial interrogation without obtaining a valid waiver of those rights. When they did so, the police violated Article I, section 12. That constitutional violation *402requires suppression of both the answers that defendant gave in response to, and the marijuana that the police identified and seized as a result of, that interrogation.”

¶11Id. at 476.

¶12II.

¶13With that background in mind, I now explain my disagreement with the majority’s conclusion that defendant’s consent to the search of his vehicle attenuated the taint of Robeson’s illegal questioning. The majority begins its attenuation analysis by “not[ing] that the amount of time that passed between the Miranda violation and the discovery of the physical evidence was brief. Additionally, defendant remained in custody during the encounter. In those respects, this case is similar to Vondehn.” 357 Or at 374. The majority does not, however, discuss the compelling nature of the circumstances confronting defendant or give those circumstances any weight in its attenuation analysis. Not only did defendant “remain in custody” after he was stopped for a seat belt violation, he was searched, handcuffed, placed in the back of a patrol car, and then basically asked by Robeson if he had engaged in any illegal activity.2 The majority does not consider the effect of those circumstances on defendant at the time that he responded to Robeson’s unwarned question.

¶14The majority instead focuses on what it sees as “the primary factual difference” between Vondehnand this case — “defendant’s invitation to search his car.” 357 Or at 374. “When Sergeant Robeson asked defendant ‘if there was anything we should be concerned about’ in his car, defendant ‘told [him] “no,” and that if [the deputies] wanted to search *403the vehicle [they] could.’” Id. Although defendant’s response to the unwarned question was immediate and directly pertained to the deputy’s question, the majority sharply breaks the response into two parts, giving great weight to what it characterizes as defendant’s “invitation” to search his car. Notwithstanding the compelling circumstances confronting defendant, the majority views the “invitation” to search as spontaneous and in no way causally related to those circumstances or to the deputy’s illegal conduct. The majority concludes that defendant’s consent was “volunteered.” 357 Or at 375.3 Thus, the majority ultimately views defendant’s consent as representing a complete break in the causal chain between the taint of the Miranda violation and the physical evidence obtained from the consent search.

¶15In my view, the majority’s conclusion that defendant’s consent to search was volunteered is inconsistent with the weight of pertinent case law. As the majority notes, Professor LaFave identifies two types of statements that properly may be considered “volunteered” in the Miranda context: (1) statements that are not prompted by police questioning, and (2) statements that are nonresponsive to a police officer’s question. 357 Or at 375 (citing Wayne R. LaFave et al, 2 Criminal Procedure § 6.7(d) (3d ed 2007 & Supp 2014) (discussing volunteered statements)). The majority appears to implicitly concede that defendant’s statements did not fall into the category of unprompted statements. Rather, the majority posits that defendant’s statement that the deputies could search his car was nonresponsive because his answer to Robeson’s question “went beyond what Robeson had asked and included an offer for the officers to search his car.” 357 Or at 376.

¶16The cases that Professor LaFave cites for the proposition that a nonresponsive statement may be considered volunteered, however, do not support the majority’s position. *404Those cases primarily involved suspects who had made incriminating statements that were wholly unrelated to the questions asked by the police or that were not a product of unwarned interrogation. See, e.g., United States v. Crisolis-Gonzalez, 742 F3d 830, 836-37 (8th Cir 2014) (the defendant’s statement that he had a gun under his mattress was volunteered, because that statement was “wholly unrelated” to law enforcement agent’s inquiry into his immigration status); United States v. Woods, 711 F3d 737, 741 (6th Cir 2013) (the defendant volunteered the “unexpected and unresponsive reply” that he had a weapon in his car when asked by officer what object was in his pocket); United States v. Fleck, 413 F3d 883, 893 (8th Cir 2005) (when the officer asked the defendants “how they liked the food” in the county jail, the officer’s question was not “calculated to elicit an incriminating response” from the defendants); United States v. Castro, 723 F2d 1527, 1530-31 (11th Cir 1984) (when law enforcement agent asked the defendant, “What in the world is going on here?”, the defendant’s subsequent offer to bribe the agent was a “spontaneously volunteered” statement that was “totally unresponsive” to the agent’s question).

¶17In contrast to the above cases, the state has conceded that Robeson’s question whether “there was anything [the deputies] should be concerned about” in defendant’s car constituted a Miranda violation. Moreover, defendant’s statements that nothing in his car should concern the deputies and that they could search his car were made in direct response to that unwarned question. Accordingly, because defendant’s statements were both prompted by and made in response to Robeson’s question, I would conclude that defendant’s consent to search in this case may not properly be considered “volunteered.”

¶18In determining that defendant’s consent attenuated the taint of the Miranda violation, the majority relies on State v. Unger, 356 Or 59, 333 P3d 1009 (2014), a case recently decided by this court in which police officers trespassed into the defendant’s backyard in violation of his privacy rights protected by Article I, section 9, of the Oregon Constitution. In that case, the officers knocked on the back door of the defendant’s residence, and the defendant consented to the entry into and the search of his residence. The defendant *405was not physically restrained by handcuffs or otherwise subjected to compelling circumstances. Thus, no Miranda warnings were required or at issue in Unger. The majority cites Ungeras “reaffirm [ing] that, when a defendant’s consent to a search either was not affected or was only tenuously connected to a prior illegality, the defendant’s voluntary consent can be sufficient to break the causal chain [between the illegality and the evidence obtained in the consent search].” 357 Or at 378. The majority does not explain, however, why the attenuation analysis in Unger— an Article I, section 9, case — should apply in an Article I, section 12, case in which a suspect was not given his Miranda warnings before he consented to a search under compelling circumstances. Moreover, in applying Unger, the majority again disregards the effect of the compelling circumstances confronting defendant as a causal factor in producing defendant’s consent and the physical evidence obtained in the consent search. In my view, Ungerdoes not provide a readymade answer to how this court should conduct an attenuation analysis under the circumstances of this case.

¶19In relying on Unger, the majority applies three factors identified in that case that it thinks bear on when a consent to search will attenuate the illegality of a Miranda violation:

“That legal determination — whether, in the circumstances of a particular case, consent has so attenuated the connection between the prior illegal conduct and the evidence obtained in the consent search — requires a court to consider the illegal conduct that comprised the stop or search, the character of the consent, and the causal relationship between the two.”

¶20Unger, 356 Or at 78. First, the majority considers the illegal conduct of law enforcement. Again, no consideration is given to the compelling circumstances confronting defendant when he consented to the search. Instead, the majority observes that the violation of defendant’s Miranda rights “can hardly be characterized as egregious …. Robeson did not engage in repeated efforts to wear down [defendant’s] resistance. … [0]ther than the initial background questions he asked, Robeson asked defendant only one question[.]” 357 Or at 378-79 (internal quotation marks omitted). Of course, *406there was no reason for Robeson to ask defendant repeated or additional questions. Under compelling circumstances, defendant provided incriminating information to the deputies based on a single question. From Robeson’s question and the circumstances, it is apparent that the purpose of the question was to elicit potentially incriminating information from defendant. Here, that purpose was achieved by Robeson asking a single question.

¶21Second, the majority considers “the character of defendant’s consent.” 357 Or at 379. Again, without any reference to the compelling circumstances confronting defendant, the majority concludes that, in response to an accusation by a sheriffs deputy who had searched and handcuffed him, defendant simply “invited the deputies to search his car if they wanted to do so. Defendant’s invitation to search his car in this case is virtually identical to the invitations in Kennedy and Rodriguez, which this court held attenuated the taint of the unlawful seizures in those cases.” 357 Or at 379 (citing State v. Kennedy, 290 Or 493, 624 P2d 99 (1981), and State v. Rodriguez, 317 Or 27, 854 P2d 399 (1993)). However, as with its reliance on Unger, the majority fails to recognize critical differences between Kennedyand Rodriguezand this case. In Kennedy, the defendant was not confronting compelling circumstances when he consented to a search of his luggage after police officers had stopped him at an airport without reasonable suspicion. The defendant was not in custody or improperly questioned as part of any Miranda violation. In Rodriguez, the defendant consented to a search of his apartment after he “had been read his Miranda rights and stated that he understood them. He was under no compulsion to answer the agent’s question.” 317 Or at 41 n 15. Both cases involved alleged violations of the defendants’ privacy rights under Article I, section 9 — not a defendant’s right to remain silent protected by Article I, section 12.

¶22Finally, the majority purports to consider “the causal connection between the violation and defendant’s invitation.” 357 Or at 380. As previously noted, defendant’s response to Robeson’s question immediately followed that question. The conversation was brief, with no intervening circumstances occurring between the question and the response. Notwithstanding the direct and obvious causal *407connection between the question and the consent, the majority observes that “[t]his is not a case in which Robeson’s unwarned questioning left little, if anything, of incriminating potential … unsaid.’” 357 Or at 380 (quoting State v. Jarnagin, 351 Or 703, 722, 277 P3d 535 (2012) (internal quotation marks omitted)). However, the quoted portion of Jarnaginpertains to the efficacy of belated Miranda warnings. The quoted discussion from Jarnaginappears to have little bearing on the causal connection between the Miranda violation and defendant’s consent to search in this case.4

¶23In short, I do not find the majority’s attenuation analysis in this case persuasive. As noted, pertinent case law does not support the majority’s conclusion that defendant’s consent was volunteered. Moreover, the majority does not recognize that the compelling circumstances to which defendant was subjected bears on whether defendant’s consent attenuates the police illegality in this case. Although the majority purports to consider the causal connection between the illegality and defendant’s consent, it declines to actually look at those compelling circumstances as a causal factor. In my view, that oversight represents a major flaw in the majority’s attenuation analysis. The majority then compounds that error by giving inordinate weight to defendant’s consent as a factor that — by itself, and in a highly fictionalized manner5 — attenuates the taint of the police illegality.

¶24*408III.

¶25As noted in Vondehn, 348 Or at 476 n 8, other state courts have decided under their state constitutions that physical evidence obtained in violation of Miranda rights must be excluded at trial as “fruit of the poisonous tree.” See, e.g., State v. Peterson, 181 Vt 436, 446-47, 923 A2d 585 (2007) (holding, under Vermont Constitution, that “[p]hysical evidence gained from statements obtained under circumstances that violate Miranda is inadmissible in criminal proceedings as fruit of the poisonous tree”); Commonwealth v. Martin, 444 Mass 213, 215, 827 NE2d 198 (2005) (adopting common-law rule under Massachusetts Constitution that physical evidence, “if derived from unwarned statements where Miranda warnings would have been required by Federal law in order for them to be admissible, is presumptively excludable from evidence at trial as ‘fruit’ of the improper failure to provide such warnings”); State v. Knapp, 285 Wis 2d 86, 123, 700 NW2d 899 (2005) (noting that “the goals of the exclusionary rule and fruit of the poisonous tree doctrines are to curb ‘illegal governmental activity,”’ and concluding that “it is appropriate that the exclusionary rule bars physical fruits obtained from a deliberate Miranda violation under Article I, Section 8” of Wisconsin Constitution); but see State v. Sole, 185 Vt 504, 514, 974 A2d 587 (2009) (holding that physical evidence not tainted by prior Miranda violation where the defendant consented to search in response to officer’s request because “a consent request is not designed to elicit an incriminating response” (internal quotation marks omitted)).

¶26The physical evidence that defendant has challenged in this case is derivative in nature, because it was obtained *409as the result of an unwarned question directed by Robeson to defendant when defendant was in custody. Defendant is therefore entitled to a determination as to whether that derivative evidence is “tainted” by the constitutional violation or, as famously stated by Justice Frankfurter, whether the evidence is the “fruit of the poisonous tree.” Nardone v. United States, 308 US 338, 341, 60 S Ct 266, 84 L Ed 307 (1939). As pointed out by Professor LaFave, Nardone“established the doctrine of ‘attenuation’ by authoritatively recognizing that the challenged evidence might sometimes be admissible even if it did not have an ‘independent source’ because the ‘causal connection … may have become so attenuated as to dissipate the taint.” LaFave, 3 Criminal Procedure § 9.3(a) at 419-20; see also Wong Sun v. United States, 371 US 471, 83 S Ct 407, 9 L Ed 2d 441 (1963) (suspect’s confession to crime untainted by his illegal arrest the day before, because it was given after his release and his voluntary return to police station).

¶27In this case, there was an immediate, direct causal relationship between the compelling circumstances confronting defendant and the consent to search that defendant gave in response to Robeson’s unwarned question. As observed in Vondehn, this court has long recognized the inherent level of coercion that exists when suspects in custody are questioned by police officers. Vondehn, 348 Or at 472; see Joslin, 332 Or at 380 (noting that protection against compelled self-incrimination under Article I, section 12, “extends to custodial interrogations, because of the inherent level of coercion that exists in such interrogations”); Meade, 327 Or at 339 (court “has recognized that a level of coercion is inherent in any custodial setting”); State v. Brewton, 247 Or 241, 244, 422 P2d 581, cert den, 387 US 943 (1967) (recognizing “the inherently coercive character of police interrogation of a suspect in custody who has not been advised of his rights”). Indeed, the United States Supreme Court identified the compulsion inherent in custodial interrogations as the primary reason for requiring police officers to warn suspects of their rights. See Miranda v. Arizona, 384 US 436, 467, 86 S Ct 1602, 16 L Ed 2d 694 (1966) (concluding that, without providing such warnings, “the process of in-custody interrogation of *410persons suspected or accused of crime contains inherently compelling pressures which work to undermine the individual’s will to resist and to compel him to speak where he would not otherwise do so freely”).

¶28In recent years, several courts have recognized the compelling effect of a police officer’s presence on an individual’s consent and have considered empirical studies in fashioning appropriate tests for determining the validity of consent searches. For example, the Supreme Court of New Jersey cited various psychological studies regarding the compulsion inherent in police-citizen encounters in State v. Carty, 170 NJ 632, 790 A2d 903 (2002). The court noted, “In the context of motor vehicle stops, where the individual is at the side of the road and confronted by a uniformed officer seeking to search his or her vehicle, it is not a stretch of the imagination to assume that the individual feels compelled to consent.” Id.at 644 (citing psychological studies that have shown that “there is an almost reflexive impulse to obey an authority figure”). The court also cited data from the New Jersey State Police Independent Monitors’ reports that indicated that nearly 95 percent of detained motorists granted a law enforcement officer’s request for consent to search. Id. at 644-45. Based in part on that social science, the court altered its test for determining the validity of a motorist’s consent to search, holding that law enforcement personnel must have a reasonable and articulable suspicion of criminal wrongdoing, beyond the initial valid motor vehicle stop, before seeking consent to search. Id. at 647.

¶29Similarly, in Brown v. State, 182 P3d 624 (Alaska Ct App 2008), the Court of Appeals of Alaska relied on various studies on the inherently compelling nature of police-citizen interactions to conclude that federal law does not adequately protect motorists. The court noted that consent searches are nearly always held to be valid under the Fourth Amendment: “The federal law in this area is premised on the assumption that, all things being equal, a motorist who does not wish to be subjected to a search will refuse consent when the officer seeks permission to conduct a search. But experience has shown that this assumption is wrong.” Id. at 630. After citing various studies that have undercut the *411notion that a person truly consents to be searched, the court concluded:

“Whatever the exact reasons for motorists’ willingness to accede to the requests of law enforcement officers, it is clear that large numbers of motorists are consenting to be searched each year — indeed, each month, and each week. Motorists are giving consent in such large numbers that it is no longer reasonable to believe that they are making the kind of independent decision that lawyers and judges typically have in mind when they use the phrase ‘consent search.’”

¶30Id. at 630-31. Ultimately, the court held, under the search- and-seizure provision of the Alaska Constitution, that an officer who had stopped the defendant for a vehicle-equipment violation was prohibited from requesting the defendant’s consent to search her person and vehicle for drugs. Id. at 634.

¶31Several other courts have acknowledged the widespread academic criticism of case law pertaining to consent searches. The Supreme Court of Kansas, for example, observed, “Commentators, in addition to noting the difficulty in applying the case law relating to consensual searches to specific fact situations, argue that the [United States Supreme] Court’s analysis utilizes an ill-crafted paradigm [for interpreting and applying the Fourth Amendment].” State v. Thompson, 284 Kan 763, 777-79, 166 P3d 1015 (2007), as modified (Oct 17, 2007) (noting numerous scholars’ critiques that consent searches after routine traffic stops are inherently coercive). Similarly, the Supreme Court of Iowa described the abundant academic commentary on consent searches pursuant to traffic stops and acknowledged that a common criticism with the consensual search doctrine is that “a traffic stop gives rise to an element of compulsion.” State v. Pals, 805 NW2d 767, 780-82 (Iowa 2011).

¶32In particular, scholars have criticized the wide gap between the fiction that ordinary citizens consent to a police officer’s requests and the reality that police-citizen encounters involve such inherently compelling circumstances as to vitiate any true choice on the part of the citizen. According to a study performed in Maryland and Ohio, which examined *412motorists’ compliance with police requests for consent to search their vehicles, approximately 90 percent of the motorists studied consented to have their vehicles searched. Daniel J. Steinbock, The Wrong Line Between Freedom and Restraint: The Unreality, Obscurity, and Incivility of the Fourth Amendment Consensual Encounter Doctrine, 38 San Diego L Rev 507, 534-35 (2001). Similar studies on police-citizen encounters have shown that “people tend to underestimate the strength of situational constraints and overestimate the voluntariness of others [’] actions.” Josephine Ross, Can Social Science Defeat a Legal Fiction? Challenging Unlawful Stops Under the Fourth Amendment, 18 Wash & Lee J Civil Rts & Soc Just 315, 332 (2012) (advocating for defeat of legal fiction of consensual nature of police-citizen encounters through use of social science). Indeed, one commentator has concluded, “The truth is that people consent so often that it undermines both the meaningfulness of the consent and the believability that the police are really respecting the doctrine.” Oren Bar-Gill & Barry Friedman, Taking Warrants Seriously, 106 NW U L Rev 1609, 1662 (2012).

¶33In State v. Jenkins, 298 Conn 209, 3 A3d 806 (2010) (Palmer, J., dissenting), Justice Palmer recently surveyed much of the academic commentary and social science regarding the coercive effect that an officer’s request for consent is likely to have on a motorist who has been detained in connection with a traffic stop. Id. at 325-34. He summarized that literature as follows:

“[E]mpirical studies over the last several decades on the social psychology of compliance, conformity, social influence, and politeness have all converged on a single conclusion: the extent to which people feel free to refuse to comply is extremely limited under situationally induced pressures. … It therefore has been argued that the United States Supreme Court should incorporate the empirical findings on compliance and social influence into … consent [search] jurisprudence … to dispel the air of unreality that characterizes the current doctrine.”

¶34Id. at 326 (internal quotation marks omitted).

¶35*413The foregoing authorities have all recognized that compulsion is an inherent feature of a police encounter when a motorist is detained in connection with a routine traffic stop. In this case, we have the additional circumstances that defendant was personally searched, handcuffed, and placed in the back of a patrol car; that strong show of authority subjected defendant to more compulsion than is ordinarily inherent in a routine traffic stop. Based on its prior precedents, this court should give those compelling circumstances appropriate weight in considering the causal connection between Robeson’s Mirandaviolation and defendant’s consent to the search of his car. Vondehn, 348 Or at 472; Joslin, 332 Or at 380; Meade, 327 Or at 339; Brewton, 247 Or at 244.

¶36This court has generally looked to the totality of the circumstances and applied a fact-intensive inquiry to determine whether physical or testimonial evidence derives from or is a product of a Mirandaviolation. Jarnagin, 351 Or at 716-17; see Vondehn, 348 Or at 482 (considering “all relevant circumstances” in deciding whether belated Mirandawarnings were effective in ensuring valid waiver of rights). Here, Robeson illegally asked defendant an unwarned question in violation of Article I, section 12, for the purpose of eliciting incriminating information from defendant. Defendant’s direct and immediate response to that question, given under compelling circumstances, included defendant’s consent to search his car. Those circumstances involved a strong show of police authority, including personally searching defendant, handcuffing him, and placing him in the back of Robeson’s patrol car. There were no intervening events between Robeson’s unwarned question and defendant’s consent. Robeson exploited that unwarned question to obtain incriminating information from defendant. And that information, in turn, was offered by the state to convict defendant of possession of a controlled substance. Under those circumstances, defendant’s consent cannot properly be viewed as a complete break in the causal chain between the Mirandaviolation and the physical evidence obtained by the police. Put differently, defendant’s consent was more than tenuously related to the Mirandaviolation. Jarnagin, 351 Or at 716-17; see also State v. Ayles, 348 Or 622, 636-39, *414237 P3d 805 (2010) (.Miranda warnings alone not sufficient to “ensure that the unlawful police conduct did not affect, or had only a tenuous connection to, [the] defendant’s responses”); LaFave, 3 Criminal Procedure § 9.3(c) at 423-24 (discussing relevant criteria for determining “when there is only an ‘attenuated connection’ between a violation and certain derivative evidence”); Comment, Fruit of the Poisonous Tree — A Plea for Relevant Criteria, 115 U Pa L Rev 1136, 1148-49 (1967) (source of “relevant criteria” relied upon by Professor LaFave).

¶37For the foregoing reasons, I would hold that defendant’s consent to the search of his car did not attenuate the taint of the Mirandaviolation by Robeson. Where, as here, a suspect consents to the search of his car under compelling circumstances and in direct response to an unwarned question by a law enforcement officer seeking to elicit incriminating information from that suspect, the physical evidence obtained from that search must be excluded as the “fruit of the poisonous tree” to give effect to the constitutional protection against self-incrimination provided for by Article I, section 12, of the Oregon Constitution. In my view, the majority’s contrary conclusion unduly diminishes the vital constitutional protection against self-incrimination provided for by Article I, section 12.1 therefore respectfully dissent.