dissenting.
An officer lawfully stopped a car in which defendant was a passenger. The majority holds that the officer unconstitutionally seized defendant when the officer retained defendant’s identification card and that all the evidence that followed was a product of that seizure and must be suppressed. Although I agree that the officer stopped defendant without reasonable suspicion, the rest of the majority’s reasoning sweeps too broadly. As an initial matter, defendant would not have left the scene of the stop, even if the officer had not retained his identification. In that circumstance, defendant’s later, voluntary consent to a patdown search was sufficient to break the causal chain. Beyond that, the officer lawfully discovered evidence during a search incident to the arrest of the driver and defendant chose to answer the officer’s questions about that evidence after having been advised of his Miranda rights. Those events occurred independently of the retention of defendant’s license and were not the product of that act. This case differs in those respects from State v. Hall, 339 Or 7, 115 P3d 908 (2005), and the majority errs in treating this case as if Hall governed it. I respectfully dissent.
The question whether a voluntary consent that follows an unconstitutional stop is a product of that stop is a fact-intensive one, and it is worth recounting the facts in this case before turning to the legal issues that they raise. One morning, a police officer lawfully stopped a car in which defendant was a passenger. The car was missing a license plate and exceeding the speed limit. The stop occurred at a *643remote location in the Coast Range along Highway 26.1 There were five people in the car (including the driver) and only a single officer. On approaching the car, the officer noticed that the driver of the car appeared to be under the influence of methamphetamine. As the officer explained:
“[The driver’s] body was contorting and twisting. She couldn’t sit still. She was running her fingers through her hair at a very high and unusual rate of speed. I saw her jaw muscles * * * and her face muscles were spasming and relaxing. I asked her for her driver’s license. She was making just some random statements that I really didn’t hear and understand. She made a couple of comments that were basically irrelevant to anything.”
The driver found her license and gave it to the officer. After giving the officer her license, the driver continued to rummage through her wallet, even though the officer had asked only for her license.
Given the driver’s behavior, the officer asked her if she had any methamphetamine in the car. At that point, defendant, who was sitting in the front passenger seat, spoke up and said, “Sir, what do we have to do to rectify this so that we stop getting stopped. So that we don’t get stopped anymore.” The timing of defendant’s question and the overly polite way in which he asked it raised the officer’s suspicions. The officer asked defendant for identification, and defendant gave him a veteran’s identification card. The officer took both the driver’s license and defendant’s identification card and put them in his car. He then asked the driver to step out of the car, spoke to her briefly, and returned to his car where he “ran a computer check on both the defendant and [the driver].” That check did not reveal any outstanding warrants for either person.
*644The officer went back to the car where defendant was sitting.2 He asked defendant “if he would mind stepping out of the car and talking to [him] for a minute.” Defendant said “[s]ure” and stepped out. At that point, the officer asked defendant to step to the back of the car, where the driver was seated, so that he could keep an eye on defendant, the driver, and the three other passengers who were still in the back seat of the car. As defendant walked to the back of the car, the officer noticed that defendant walked with an exaggerated limp, but he did not see any obvious signs that defendant was under the influence of any substance.
Once they reached the back of the car, the officer explained to defendant that, when he talks to someone outside of a car, he “like[s] * * * to pat them down for weapons” and then he asked defendant “for consent to pat him down.” Defendant replied, “ Yeah.’ ” As the officer explained, “[h]e told me that I could pat him down.” As the officer started to pat defendant down, he could see from his vantage point into defendant’s shirt pocket. Inside the pocket, he “saw a prescription pill bottle with no label on it[,] and [he] could see inside the pill bottle was plastic wrapping and [he] immediately suspected it was drugs.” He took the bottle out of defendant’s pocket and asked, “Is that the meth[,]” and defendant said, ‘Yeah, I guess it must be.” The officer placed defendant under arrest and advised him of his Miranda rights.
At that point, the officer turned his attention back to the driver, whom he “ran * * * through a field sobriety test and subsequently arrested her for driving under the influence of methamphetamine.” After the officer arrested the driver, he searched the car for “additional means of intoxication.” During that search, the three passengers in the back seat got out of the car. One of them told the officer that there was a backpack in the car, which the passenger identified as defendant’s. The officer asked defendant if there were methamphetamine in the backpack. Defendant said that there was and described the contents of the backpack in detail. The *645officer then looked in the backpack and found more methamphetamine.
Before trial, defendant moved to suppress all the evidence found during the stop. Specifically, defendant moved to suppress: (1) the methamphetamine found in the pill bottle during the patdown search; (2) defendant’s statements to the officer regarding the backpack; and (3) the contents of the backpack. The trial court denied that motion, and the Court of Appeals reversed the resulting judgment of conviction.
Before turning to the issue that this case presents, it is important to note what is not at issue. Defendant does not contend that the officer lacked probable cause to stop the car for a traffic infraction, nor does he dispute that the officer acquired reasonable suspicion that the driver was under the influence of methamphetamine once he approached the car and spoke with her. Defendant has not challenged the voluntariness of his consent to the patdown search. He has not questioned the officer’s testimony that the prescription bottle in defendant’s shirt pocket was in plain view when the officer began to pat him down, and he has not argued that the officer acted unconstitutionally when he removed the prescription bottle from his shirt pocket and asked him about its contents. Relatedly, defendant does not argue that the officer lacked probable cause to arrest the driver for driving under the influence of methamphetamine, once she performed field sobriety tests, and he has not challenged the officer’s authority to search the car for additional evidence of intoxication, as a search incident to the driver’s lawful arrest. Defendant’s argument accordingly reduces to the proposition that every lawful act that followed the retention of his identification must fall because it was a product of that single, initial illegality.
The state, for its part, does not dispute that the officer’s retention of defendant’s identification card prevented him from leaving and thus stopped defendant without reasonable suspicion. The state argues, however, that there is no causal connection between that illegality and defendant’s voluntary consent to the patdown search because defendant would not have left the scene, even if the officer *646had not retained his identification. It follows, the state contends, that defendant’s voluntary consent to the patdown search was not the product of the prior illegality. Alternatively, the state argues that, even if the consent were the product of the retention of defendant’s identification, the Miranda warnings that defendant subsequently received were, in the totality of the circumstances, sufficient to break the causal chain, rendering defendant’s post-Miranda statements and the contents of the backpack admissible.
In my view, both of the state’s arguments are well taken. Under this court’s decisions, when the officer retained defendant’s identification card, he prevented him from leaving and accordingly stopped him without reasonable suspicion. See, e.g., State v. Thompkin, 341 Or 368, 379, 143 P3d 530 (2006) (so reasoning). Defendant, however, also was prevented from leaving the scene of the stop for an independent and completely lawful reason. The officer lawfully had stopped the car in which defendant was riding and, on stopping the car, reasonably suspected that the driver was under the influence of methamphetamine. The stop occurred in a remote wooded location, three miles from the nearest home, and the trial court reasonably could have found that defendant would not have left the scene of the lawful stop of the car even if the officer had not retained his identification.3 In this case, as long as the driver’s car was lawfully detained, so was defendant.
As a matter of causation, two independent causes prevented defendant from leaving. One was lawful; the other was not. Under those circumstances, any causal connection between the retention of defendant’s identification card and his voluntary consent to the patdown search was so faint that defendant’s voluntary consent was sufficient to break the causal chain. To be sure, this court held in Hall that the defendant’s voluntary consent that followed an unlawful stop did not break the causal chain. 339 Or at 36. In that case, however, the unlawful seizure was the only reason that the *647defendant in Hall remained at the scene. In this case, by contrast, the lawful retention of the car and its driver prevented defendant (the passenger) from leaving, without regard to whether the officer retained the defendant’s identification or gave it back.
Not every causal connection, however faint, will be sufficient to disable a defendant’s voluntary consent from breaking the causal chain, and the causal connection in this case is, if existent at all, far weaker than the causal connection in Hall. This case does not require us to decide whether Hall was correctly decided, and the state has not asked us to reexamine that decision. Rather, the state argues, and I would hold, that any causal connection in this case between the retention of defendant’s identification and defendant’s decision to consent is so much weaker than it was in Hall that we should give effect to defendant’s voluntary consent to the patdown search.4
As noted, defendant does not challenge the voluntariness of his consent to the patdown search, nor has he argued that the officer did not lawfully see, in plain view, what the officer reasonably suspected was contraband once he started to conduct the patdown search. Because everything that followed defendant’s voluntary consent to the patdown search was constitutionally obtained and because, in these circumstances, defendant’s voluntary consent was not the product of the unlawful retention of his identification, I would hold, on that ground alone, that all the evidence that the officer discovered was admissible.
*648The majority reaches a different conclusion. It holds that, under Hall, a minimum factual nexus will exist between an illegality and a defendant’s consent as long as the consent occurs while the unlawful stop is ongoing. The majority’s holding is at odds with the reasoning in Hall. The question under Hall is not whether there is a temporal connection between the illegality and the defendant’s decision to consent. Rather, the question is whether there is a causal connection. The majority’s decision either assumes incorrectly that all events that occur contemporaneously are causally connected, or it dispenses with the requirement that the illegality must be the cause of the decision to consent. Either way you cut it, the majority errs.
The majority, for its part, invokes the decision in Hall and two cases that follow it as support for its holding. The analysis in Hall, however, undercuts the majority’s rationale, and the two cases that follow Hall add nothing to the analysis. The question in Hall was whether evidence discovered as a result of the defendant’s voluntary consent to a search “derived from a preceding violation of the defendant’s rights under [Article I, section 9, of the Oregon Constitution].” 339 Or at 21. The question that the court posed in Hall was at base a causal one. See id. at 24 (“[T]he critical inquiry is whether the state obtained the evidence sought to be suppressed as a result of a violation of the defendant’s rights under Article I, section 9.”) (emphasis added). Consistently with that focus, the court described the minimum factual nexus that defendants must establish as a causal connection between the illegality and the defendant’s decision to consent. See id. at 25 (“[A] minimum factual nexus * * * is, at minimum, * * * a ‘but for’ relationship[.]”). The standard that the majority announces in this case — that the illegality and the decision to consent need share only a temporal link — is inconsistent with the court’s recognition in Hall that the connection must be causal.
To be sure, it often will be the case that consent obtained during an illegal stop also will be causally connected to the stop. But that is not always so, as this case illustrates. Any causal link in this case was so faint that defendant’s voluntary consent was sufficient to break the causal chain. Not only is the majority’s reasoning at odds with the analysis in *649Hall, but this case differs factually from Hall in that the illegal detention in that case was the only reason that the defendant in Hall remained at the scene. No independent lawful reason prevented the defendant in Hall from leaving, and this court presumed without further discussion that there was a sufficient causal connection between the illegality and the consent to shift the burden to the state to prove independent source, inevitable discovery, attenuation, or the like. 339 Or at 36.5
The two other cases that the majority cites add nothing to the analysis. One case, State v. Rodgers / Kirkeby, 347 Or 610, 227 P3d 695 (2010), is factually similar to Hall. In that case, the illegal detention was the only reason that each defendant remained at the scene. Because no independent, lawful reason prevented either defendant from leaving, that case did not present the causation question that this case does.
The other case, State v. Thompkin, 341 Or 368, 143 P3d 530 (2006), is factually closer. In Thompkin, the defendant was a passenger in a lawfully stopped car. The police stopped the passenger (the defendant) when they retained her identification without reasonable suspicion. Id. at 377-79. Although the trial court’s findings are not particularly clear, its findings can be read to suggest that the defendant in Thompkin would have remained at the scene of the lawful stop without regard to whether the police unlawfully retained her identification.6 In that respect, Thompkin is factually closer to this case.
*650Despite that possible similarity, Thompkin provides no precedential support for the majority’s decision because the state never argued and this court never considered in Thompkin whether the possible presence of an independent, lawful reason for remaining at the scene meant that there was no minimum factual nexus between the unlawful retention of the defendant’s identification and her voluntary decision to disclose evidence to the police. See 341 Or at 380-81. Indeed, the decision in Thomkin omits any discussion of whether the defendant had established a minimum factual nexus that shifted the burden to the state. See id. The opinion is silent on that point, and we should not treat Thompkin’s silence as if it were binding precedent. See Coast Range Conifers v. Board of Forestry, 339 Or 136, 148-49, 117 P3d 990 (2005) (reaching that conclusion regarding an issue that an earlier decision could have but did not address).
As the majority ultimately recognizes, this court has assumed without discussion, both in Hall and also in Rodgers/Kirkeby and Thompkin, that a minimum factual nexus existed, and those cases thus provide little or no analytical help in determining what a defendant must show to establish that nexus. The majority attempts to supply the analytical deficiency in those decisions with the following explanation:
“We think that the reason both that the court sometimes assumes without discussion that a defendant has shown the required nexus when consent occurs during an ongoing seizure and that no case exists holding that there is no minimum connection in such circumstances is that the existence of a minimal factual nexus is obvious in cases in which the defendant consents to a search (or takes other incriminating action) during an illegal seizure.”
348 Or at 633-34 (emphasis in original). In my view, the majority’s explanation still leaves something to be desired. However obvious the causal connection may be in some cases, it is not obvious in every case. The majority errs in sweeping away all factual distinctions with an uncritical assertion that a minimum factual nexus will always exist when consent *651occurs during an illegal detention; in so doing, it incorrectly substitutes a temporal connection for a causal one.
One final point requires discussion. The majority says, in a footnote, that the state has not argued the first issue that this opinion discusses. The state, however, argued in its brief to the Court of Appeals that, under Hall, a defendant must show “ ‘at minimum, the existence of a “but for” relationship’ ” between the evidence sought to be suppressed and the police illegality to establish a factual nexus. Respondent’s Brief at 8 (quoting Hall, 339 Or at 25). The state contended that the evidence in this case was not sufficient to establish even that minimal causal connection. It followed, the state reasoned, that “[tjhere is no reason to suppose * * * that taking and retaining defendant’s ID had any effect on his decision to consent to the patdown[.]” Id. at 9-10.
In this court, the state raised the same issue. It contended that, because the traffic stop was lawful and ongoing, any connection between the retention of defendant’s identification and the officer’s request for consent to the patdown search was not sufficient to establish even “but for” causation.7 In my view, the state’s arguments in both the Court of Appeals and this court properly raised the question whether the causal connection between the retention of defendant’s identification and his consent to the patdown search was so weak that defendant’s voluntary consent broke the causal chain. As explained above, on that issue, I would hold that, because the detention in this case resulted equally from a lawful and an unlawful cause, the illegality did not have a sufficient effect on defendant’s voluntary decision to consent to the patdown search to prevent that decision from breaking the causal chain, even under Hall.
A second, more limited basis for reversing the Court of Appeals decision exists. Even if one assumes that defendant’s consent to the patdown search was not sufficient to break the causal chain, four later events, viewed collectively, *652were. First, the officer lawfully arrested the driver for driving under the influence of methamphetamine. Second, the officer lawfully searched the car for more evidence of intoxicants incident to his lawful arrest of the driver. See State v. Caraher, 293 Or 741, 759, 653 P2d 942 (1982) (authorizing a search incident to arrest when “it is relevant to the crime for which defendant is being arrested and so long as it is reasonable in light of all the facts”).8 Third, during that search, one of the passengers got out the car, told the officer that there was a backpack in the car, and identified it as defendant’s.9 Fourth, the officer asked defendant whether there was methamphetamine in the backpack, and defendant, having previously received Miranda warnings, chose to answer the officer’s questions.
Even if defendant’s voluntary consent to the pat-down did not break the causal chain, his receipt of the Miranda warnings did. Admittedly, giving a defendant Miranda warnings does not always break the causal connection between an Article I, section 9, violation, and a defendant’s subsequent statements. See Hall, 339 Or at 35 (identifying several considerations, in addition to Miranda warnings, to account for in assessing a causal connection); see Brown v. Illinois, 422 US 590, 603, 95 S Ct 2254, 45 L Ed 2d 416 (1975) (‘Miranda warnings, alone and per se, cannot always make the act sufficiently a product of freewill to break, for Fourth amendment purposes, the causal connection[.]”). Rather, determining whether those warnings were sufficient to break “a causal connection requires examination of the specific facts at issue in a particular case.” Hall, 339 Or at 35. That is, the question whether the warnings purged the *653taint requires consideration of the nature and the extent of the taint.
Here, any taint was minimal. The only illegality that defendant has identified was the fact that the officer’s retention of defendant’s identification card prevented defendant from leaving and thus stopped him without reasonable suspicion. As explained above, however, defendant could not leave for an independent, completely lawful reason. In those circumstances, the effect of retaining his identification card was minimal, if not nonexistent. The subsequent, independent discovery of defendant’s backpack and the Miranda warnings that defendant received were sufficient to attenuate any taint deriving from the officer’s retention of his identification. If defendant’s voluntary statements were not the product of the officer’s retention of defendant’s identification, then the subsequent search of his backpack was lawful, based solely on those statements, either under the automobile exception or as a search incident to defendant’s arrest.
Relying on State v. Olson, 287 Or 157, 598 P2d 670 (1979), the majority reaches a different conclusion. However, this is not a case in which the officers broke into the defendant’s house without justification and roused the defendant and his companion from their bed, as they did in Olson. Cf. id. at 159. Nor is this a case in which the defendant, on entering his home, found an officer unlawfully inside his home pointing a gun at him and saying, “Don’t move, you are under arrest.” Cf. Brown, 422 US at 592. In that case, neither the officers’ entry into the defendant’s house nor their arrest of him at gun point was constitutionally justified. Id. at 596. In those cases, this court and the United States Supreme Court respectively held that the mere fact that the officer gave the defendants Miranda warnings before the defendants made incriminating statements was not sufficient to purge the taint of the officers’ unconstitutional entries into the defendants’ homes. Olson, 287 Or at 166; see also Brown, 422 US at 603. In light of the nature and severity of the Fourth Amendment and Article I, section 9, violations in those cases, something more than Miranda warnings was required. Olson, 287 Or at 166; Brown, 422 US at 604-05.
*654In this case, by contrast, the nature and severity of the violation was minimal. The officer retained defendant’s identification and prevented him from leaving in a situation where defendant had no ability to go anywhere until the officer concluded his lawful stop of the car. The degree of attenuation necessary to purge the taint varies with the extent of the taint, and where, as here, any taint is minimal, the required degree of attenuation is correspondingly reduced. The point has nothing to do with deterrence. Rather, under a rights-based suppression analysis, the degree of attenuation necessary to purge the taint (and thus restore the defendant to the position he or she would have been in had no constitutional violation occurred) varies with the extent, nature, and severity of any illegality.10 Any other rationale would give a constitutional violation that had only minimal effect far greater reach than either the constitution requires or good sense warrants. Because I would uphold the trial court’s judgment in this case, I respectfully dissent.
Balmer and Linder, JJ., join in this opinion.The officer testified that the closest home “would probably be three miles away[.]” Later, the officer explained that there was one house on Saddle Mountain Road about a quarter to a half mile up that road from where it “leaves [Highway] 26.” There was no testimony regarding how far Saddle Mountain Road was from the scene of the stop and thus nothing to call into question the officer’s earlier testimony that the nearest home was three miles away. As defense counsel put the issue to the officer, “So as far as people in the car * * * would be able to observe by driving through that area, they’re three miles from anything in one direction out there * * * [and they’re] — I don’t know- — six or seven miles I think in the other direction.” The officer answered, “Approximately. Yes, sir.”
It is not clear when the officer returned the license to the driver and the identification card to defendant. The state, however, does not dispute that the officer did not return the identification card before he began speaking with defendant.
The majority notes in passing that the facts “do not establish that it would have been impossible, or even extremely difficult, for defendant to leave the scene.” 348 Or at 631. The majority uses the wrong standard of review. Whether defendant could or would have left the scene if the officer had not retained his identification is a factual issue for the trial court, which ruled in the state’s favor.
In that connection, it is worth noting that the United States Supreme Court and almost every state supreme court that has considered the issue have held that stopping a car seizes both the driver and the passengers. See Brendlin v. California, 551 US 249, 257, 259 n 5, 127 S Ct 2400, 168 L Ed 2d 132 (2007) (so holding and noting the holdings of the state supreme courts). Reasonable suspicion to stop the car justifies the stop of both the driver and the passengers. See Arizona v. Johnson, _ US _ , 129 S Ct 781, 172 L Ed 2d 694 (2009) (same) (citing Brendlin, 551 US at 255, 263). Under those decisions, the officer’s retention of defendant’s identification in this case would not have constituted an unlawful stop. Rather, it simply would have been a permissible inquiry during the lawful stop of the car, the driver, and all four passengers. As a practical matter, this case is no different. The stop of this car stopped defendant as well and, in the circumstances of this case, the officer’s retention of defendant’s identification contributed nothing additional to the restraint effectively placed on defendant’s freedom to leave as a result of the lawful stop of the car in which defendant was a passenger.
It is worth noting that the procedural paradigm that the court announced in Hall is difficult to square with the reasoning in both Hall and other cases. The court explained in Hall, 339 Or at 25, and reaffirmed in State v. Crandall, 340 Or 645, 652, 136 P3d 30 (2006), that the fact that an illegality is the “hut for” cause of a defendant’s voluntary consent does not render the resulting evidence inadmissible. However, in describing how the Hall paradigm works, the court stated that proof of a “but for” causal connection will be sufficient to establish a minimum factual nexus and thus to presume that the consent was the product of the illegality, unless the state offers some evidence other than the defendant’s voluntary consent to prove independent source, inevitable discovery, attenuation, or the like. Hall, 339 Or at 25. In my view, there is a substantial question whether proof of “but for” causation is by itself sufficient to disable the effect of a defendant’s voluntary consent. In this case, it is especially troubling that the majority finds that something less than “but for” causation suffices to establish a minimum factual nexus.
The court noted in Thompkin that the trial court had found that “ ‘there’s no evidence that [the defendant] would have gotten out of the car or left the scene, that *650she didn’t feel free to leave because [the officer retained her identification].’ ” 341 Or at 373-74 (quoting the trial court’s findings).
In this court, the state focused on the absence of a “but for” causal connection between the retention of defendant’s identification and the officer’s request for consent, hut it did so presumably because that was what the Court of Appeals had held was the correct focus in determining the existence of a minimal causal connection under Hall. See State v. Ayles, 220 Or App 606, 612, 188 P3d 378 (2008).
This court recognized in State v. Owens, 302 Or 196, 729 P2d 524 (1986), that, under Article I, section 9, a search incident to arrest permits officers to look in the trunk of the defendant’s car for evidence of the crime for which they had arrested him. See id. at 203-04 (quoting, with approval, an earlier case for that proposition).
Under Article I, section 9, an officer may look for evidence of the crime for which he or she arrested a person to the extent that the evidence “reasonably could be concealed on the arrestee’s person or in the belongings in [the arrestee’s] immediate possession at the time of the arrest.” Owens, 302 Or at 200. In this case, the state has not argued that the driver was in possession of defendant’s backpack at the time of her arrest. Accordingly, although the officer lawfully discovered defendant’s backpack as part of a search incident to the driver’s arrest, he could not, incident to the driver’s arrest, go further and search the contents of defendant’s backpack.
This court has rejected, in a footnote and in dicta, any reliance on the purpose and flagrancy of the constitutional violation to the extent that it bears on deterrence. Hall, 339 Or at 35 n 21. That reasoning is consistent with the rights-based suppression analysis that this court has adopted. To the extent, however, that the dictum in Hall goes further and suggests that the nature of the violation is irrelevant to the degree of attenuation necessary to purge the taint flowing from that violation, the suggestion lacks persuasive value. Not only is the suggestion dictum and thus not binding, but the assumption that underlies it — that all constitutional violations have the same effect — is erroneous.