As Justice Werdegar observes, “defendant failed in the trial court to challenge the applicability of the passenger’s parole search condition or the permissible scope of the warrant-less search . . . .” (Cone. & dis. opn. of Werdegar, J., ante, at p. 934.) As a result, the record in this case is very limited. We know that the police officer conducted a search of the car based on the front seat passenger’s parole status. But we do not know what the officer asked, learned, or believed in the course of the search, even though such facts bear critically on whether the officer reasonably believed the areas and items searched were under the parolee’s control. Today’s opinion effectively deems such facts irrelevant to the lawfulness of the search and, in so doing, adopts a novel Fourth Amendment rule that may be broader than necessary to resolve the legality of what actually happened in this case. Because judicial restraint counsels against deciding constitutional questions when it is unnecessary to do so (see Santa Clara County Local Transportation Authority v. Guardino (1995) 11 Cal.4th 220, 230 [45 Cal.Rptr.2d 207, 902 P.2d 225]), I agree with Justice Werdegar that the prudent course here is to reverse the Court of Appeal based on defendant’s apparent forfeiture in the trial court. However, because the court has reached out to announce a far-reaching constitutional rule, I write further to explain why today’s decision is unpersuasive on the merits.
The court holds that a police officer who discovers that the passenger in the front seat of a car is on parole may search “those areas of the passenger *936compartment where the officer reasonably expects that the parolee could have stowed personal belongings or discarded items when aware of police activity.” (Maj. opn., ante, at pp. 913, 926.) The court describes a search pursuant to this standard as “confined” and “not without limits.” (Id. at p. 926.) But the logic of today’s holding appears to authorize a police officer, simply upon learning that the front seat passenger is on parole, to search all open areas of the passenger compartment of a standard five-passenger car. These areas include the backseat, the area behind the backseat headrests, the backseat foot areas, any door pockets in the front or back on both sides of the car, and the floor areas under both front seats—for these are all places where an officer may reasonably expect “the parolee could have stowed personal belongings or discarded items when aware of police activity.” (Id. at pp. 913, 926.)
This holding is unduly broad. After today, a commuter who picks up a rider in a casual carpool on the way to work and is stopped for speeding may be subject to a search of all open areas in the car’s passenger compartment if the police officer learns that the rider is on parole. The same goes for a driver who volunteers to drive a group of parent chaperones on an elementary school field trip. And the same goes for a person who agrees to pick up a friend of a friend on the way to the movies. I am not sure what societal baseline the court deems legally relevant when it suggests these arrangements are “ ‘atypical’ ” or not “ ‘usual.’ ” (Maj. opn., ante, at p. 925, fn. 15.) In any event, it seems an unduly cramped reading of the Fourth Amendment to say that the only sure way a driver in these everyday situations can protect himself or herself from the possibility of a warrantless, suspicionless search of all open areas of the passenger compartment is to ask, before letting a rider into the car: “By the way, are you on parole?” Yet that is apparently now the rule in California.
Further, the court holds that if an officer encounters personal property in the course of searching the car’s passenger compartment, the property also may be searched “if the officer reasonably believes that the parolee owns those items or has the ability to exert control over them.” (Maj. opn., ante, at p. 913.) This too is exceedingly broad. To say that an officer may search any property he or she reasonably believes to be under the parolee’s control is to adhere faithfully to the parole search condition. (See Cal. Code Regs., tit. 15, § 2511, subd. (b)4 [parolee upon release shall be notified that “[y]ou and your residence and any property under your control may be searched without a warrant at any time . . . .”].) To say that an officer may search any property that he or she reasonably believes to be within the parolee’s “ability to exert control” is to authorize the officer to search virtually any property found in the passenger compartment.
I would not hold, as the Court of Appeal seemed to imply, that a lawful parole search may never go beyond the front passenger seat when a police *937officer knows the passenger in that seat is a parolee. But nor would I hold, as the court does today, that a lawful parole search may always go beyond the front passenger seat. Instead, I would hold, as precedent dictates, that the reasonableness of a search beyond the front passenger seat depends on the “ ‘totality of the circumstances’ ” in each case. (Samson v. California (2006) 547 U.S. 843, 848 [165 L.Ed.2d 250, 126 S.Ct. 2193] (Samson).) Accordingly, I respectfully disagree with the broad rule adopted by the court in derogation of the Fourth Amendment’s protections.
I.
It is well established that California parolees “have severely diminished expectations of privacy by virtue of their status alone.” (Samson, supra, 547 U.S. at p. 852.) As today’s opinion notes: “Under California statutory law, every inmate eligible for release on parole ‘is subject to search or seizure by a . . . parole officer or other peace officer at any time of the day or night, with or without a search warrant or with or without cause.’ (Pen. Code, § 3067, subd. (b)(3).) Upon release, the parolee is notified that ‘[y]ou and your residence and any property under your control may be searched without a warrant at any time by any agent of the Department of Corrections [and Rehabilitation] or any law enforcement officer.’ (Cal. Code Regs., tit. 15, § 2511, subd. (b)4; see also Cal. Code Regs., tit. 15, § 2356 [requiring the department staff to notify the prisoner of the conditions of parole prior to release].)” (Maj. opn., ante, at p. 916.) The parole search condition furthers the state’s “ ‘ “overwhelming interest” ’ in supervising parolees because ‘parolees ... are more likely to commit future criminal offenses’ ” as well as the state’s “interests in reducing recidivism and thereby promoting reintegration and positive citizenship among ... parolees.” (Samson, supra, 547 U.S. at p. 853.)
Although parolees have “substantially diminished expectation^] of privacy” {Samson, supra, 547 U.S. at p. 855), our cases have been careful to delimit the proper scope of a warrantless, suspicionless parole search where such a search implicates the privacy interests of third parties. In the context of a residential search, we have said that “the expectation of privacy of cohabitants is the same whether the search condition is a condition of probation or parole.” (People v. Sanders (2003) 31 Cal.4th 318, 330 [2 Cal.Rptr.3d 630, 73 P.3d 496].) The cases involving persons who are not on probation or parole living with a person who is (see maj. opn., ante, at pp. 917-918) make clear that “common or shared areas of their residence may be searched by officers aware of an applicable search condition.” (People v. Robles (2000) 23 Cal.4th 789, 798 [97 Cal.Rptr.2d 914, 3 P.3d 311] (Robles).) At the same time, we have “emphasized” that “a search pursuant to a probation search clause may not exceed the scope of the particular clause *938relied upon. [Citation.] Nor may such a search be undertaken in a harassing or unreasonable manner. [Citations.] Moreover, officers generally may only search those portions of the residence they reasonably believe the probationer has complete or joint control over. [Citation.]” (People v. Woods (1999) 21 Cal.4th 668, 681-682 [88 Cal.Rptr.2d 88, 981 P.2d 1019]; see Robles, at p. 798 [nonprobationers “retain valid privacy expectations in residential areas subject to their exclusive access or control, so long as there is no basis for officers to reasonably believe the probationer has authority over those areas”].)
The search in the present case implicates the privacy interests of a driver whose car carried a passenger who was on parole. As the court notes, the legality of a warrantless parole search is not based on consent. (Maj. opn., ante, at p. 921.) In Cahfornia, the parole search condition is specified by regulation. It provides that a parole search may extend to the parolee’s person and to “any property under [the parolee’s] control.” (Cal. Code Regs., tit. 15, §2511, subd. (b)4.) In upholding the parole search at issue here, today’s opinion defines the scope of a valid search in terms that exceed the scope of the parole search condition. Instead of focusing on what property is under the parolee’s control, the court expands the scope of a parole search, first, by authorizing a police officer to search any “areas of the passenger compartment where the officer reasonably expects that the parolee could have stowed personal belongings or discarded items when aware of police activity” and, second, by authorizing the officer to search items of “personal property located in those areas if the officer reasonably believes that the parolee . . . has the ability to exert control over them.” (Maj. opn., ante, at p. 913.) The upshot is that an officer is permitted to search all open areas of the car’s passenger compartment and virtually any property located in those areas, simply upon ascertaining that the front seat passenger is a parolee.
The court says its holding reflects “commonly held social conventions” (maj. opn., ante, at p. 926) concerning the places in a car where “[a] front seat passenger, even if only a casual acquaintance of the driver, will likely feel free to stow personal items . . . .” {Id. at p. 925.) I suspect many law-abiding citizens who drive or ride in cars will be surprised to learn that the areas under the “control” of a front seat passenger invariably extend to anywhere the passenger “could have stowed personal belongings or discarded items when aware of police activity” {id. at p. 913), including the backseat, the area behind the backseat headrests, the backseat foot areas, any door pockets in the front or back on both sides of the car, and the floor areas under both front seats. For example, the etiquette of a causal carpool for commuters may call for riders in the front passenger seat to place belongings on the floor at their feet or in their laps, but not in the backseat or anywhere else unless permission is asked and given. Five coworkers driving to lunch in a five-passenger car may understand that they must keep their possessions next to *939them and not elsewhere in the vehicle. But a driver who picks up a friend at the airport may expect that the passenger will place a suitcase on the backseat or in the trunk of the automobile. As these examples suggest, it is questionable to posit—as the court does without empirical or other authority—what social conventions in an automobile are “commonly held,” “usual,” or “typical.” (See id. at pp. 926, 925, fn. 15, 927.)
Because social conventions vary depending on the situation, the issue of whether a car’s backseat or items located there are “property under [the parolee’s] control” (Cal. Code Regs., tit. 15, § 2511, subd. (b)4) must be decided on the basis of the totality of the circumstances in each case. That is a straightforward application of settled law. As the high court stated in Ohio v. Robinette (1996) 519 U.S. 33, 39 [136 L.Ed.2d 347, 117 S.Ct. 417]: “We have long held that the ‘touchstone of the Fourth Amendment is reasonableness.’ [Citation.] Reasonableness, in turn, is measured in objective terms by examining the totality of the circumstances. [f] In applying this test we have consistently eschewed bright-line rules, instead emphasizing the fact-specific nature of the reasonableness inquiry.” (Accord, Samson, supra, 547 U.S. at p. 848 [“ ‘[U]nder our general Fourth Amendment approach’ we ‘examin[e] the totality of the circumstances’ to determine whether a search is reasonable within the meaning of the Fourth Amendment.”]; United States v. Knights (2001) 534 U.S. 112, 118 [151 L.Ed.2d 497, 122 S.Ct. 587] [“we conclude that the search of Knights was reasonable under our general Fourth Amendment approach of ‘examining the totality of the circumstances’ . . .”]; People v. Robinson (2010) 47 Cal.4th 1104, 1120 [104 Cal.Rptr.3d 727, 224 P.3d 55] [“ ‘As the text of the Fourth Amendment indicates, the ultimate measure of the constitutionality of a governmental search is “reasonableness.” ’ [Citation.] [f] ‘Reasonableness . . . is . . . measured in objective terms by examining the totality of the circumstances’ . . . .”].)
Although today’s opinion purports to examine the totality of the circumstances (maj. opn., ante, at pp. 921-922), it is clear that the court does so at a wholesale level rather than on the facts of this particular case. The court notes that Officer Mihai “observed that [defendant’s] arms were covered with abscesses, which she associated with drug use.” (Maj. opn., ante, at p. 914; see id. at p. 925, fn. 15.) But the court does not explain how this fact is relevant to the scope of a lawful parole search where the front seat passenger, not defendant, was the parolee. The court also notes that “[testimony at the suppression hearing established that defendant and the parolee had known each other for approximately three years at the time of the search.” (Id. at p. 925, fn. 15.) But nothing in the record indicates that this fact was known to Officer Mihai at the time of the search. Indeed, the record before us contains no facts indicating what areas or property in the car Officer Mihai reasonably believed to be under the parolee’s control.
*940Given the deficiencies in the record, the court upholds Officer Mihai’s search of the backseat and the property found there on the basis of general considerations rather than facts specific to this case. Those considerations include the “general[]” characteristics and layout of a standard, noncommercial, five-passenger car (maj. opn., ante, at pp. 925, 926-927), the “[t]ypical[]” behavior of occupants in a standard five-passenger car according to “modem social conventions” (id. at pp. 924—925), a general statement of “the state’s interest in supervising parolees” (id. at p. 923), and a general statement of a driver’s “reduced expectation of privacy with regard to an automobile” especially “when he allows others to ride in his car” (id. at p. 924.). Because this is what the court understands the “totality of the circumstances” to mean, there is no real significance to the limiting phrases in the court’s statement of its holding: “Considering the layout of a standard five-passenger car, it was objectively reasonable for the officer to expect that this parolee could have stowed his personal property in the backseat, tossed items behind him, or reached back to place them in accessible areas upon encountering the police. Accordingly, under these circumstances, the parolee status of the front seat passenger justified a warrantless search of the backseat area where the chips bag and shoes were located.” (Id. at pp. 926-927, italics added.) How does it matter that the present case involved “this parolee” “under these circumstances” when today’s opinion authorizes a search of the same breadth with respect to any parolee riding in the front passenger seat of any standard five-passenger car?
Moreover, the court authorizes a parole search not just of “property under [the parolee’s] control” (Cal. Code Regs., tit. 15, § 2511, subd. (b)4), but of property within the parolee’s “ability to exert control” (maj. opn., ante, at pp. 913, 930). This is a subtle but important difference that significantly expands the permissible scope of a parole search. It may be true, as a matter of sheer physicality, that a parolee has the “ability to exert control” over any unlocked container within the passenger compartment. But it is an odd definition of “control” to say that any item of property in which a parolee might potentially discard or conceal contraband is property under the parolee’s “control.” Apart from a single case that glancingly constmed property under a parolee’s control to encompass portions of a house “to which [the parolee] had access” (People v. LaJocies (1981) 119 Cal.App.3d 947, 955 [174 Cal.Rptr. 100]), I am aware of no authority—and the court cites none—-that supports today’s expansive and unusual interpretation of “control” as that term is used in the parole search condition.
The available authority interprets “control” more naturally and sensibly to mean not mere physical access but rather ownership, possession, or authority over the property searched. For example, in People v. Baker (2008) 164 Cal.App.4th 1152 [79 Cal.Rptr.3d 858], the male driver of a car stopped for speeding was on parole. The female defendant, Baker, was the front seat *941passenger, and her purse was on the floor at her feet. Upon discovering that the driver was on parole, the officer ordered the occupants out of the car. “Baker did so without taking her purse and without asserting ownership of the purse.” (Id. at p. 1156.) The officer searched the vehicle and found drugs in Baker’s purse. The Court of Appeal held the search unlawful on the ground that “there was no reasonable basis to believe the purse belonged to anyone other than the sole female passenger.” (Id. at p. 1160.) The court reached this conclusion even though Baker did not claim ownership of the purse and even though the parolee who drove the car could have easily stowed contraband in the purse.
In People v. Veronica (1980) 107 Cal.App.3d 906 [166 Cal.Rptr. 109], the court similarly suppressed contraband found in a purse during a parole search of a male parolee’s residence. The court explained: “We do not, of course, suggest that simply because a garment or container is clearly designed for a person other than the parolee, it may never be searched under the parolee’s prerelease consent. The particular circumstances may indicate that the object is, in fact, one of the parolee’s own effects or, at least, jointly possessed by him and another. In this case, however, there was simply nothing to overcome the obvious presumption that the purse was hers, not his.” (Id. at p. 909; cf. People v. Boyd (1990) 224 Cal.App.3d 736, 749-751 [274 Cal.Rptr. 100] (Boyd) [upholding search of a handbag where “ ‘articulable facts’ ” supported a rational inference that the handbag was owned or controlled by a parolee].)
In People v. Montoya (1981) 114 Cal.App.3d 556 [170 Cal.Rptr. 624], the court suppressed drugs found in a pair of jeans during a parole search of a residence. The circumstances indicated that the jeans belonged to one of two female guests, one of whom was on parole. The court held the search unlawful on the ground that the officer failed to determine “who owned the jeans before searching them” when “he had no more reason to believe [the jeans] belonged to [the parolee] rather than [to] appellant.” (Id. at p. 562.) A number of probation search cases similarly focus on the reasonableness of an officer’s belief as to who owned or possessed the area or item searched. (See, e.g., People v. Tidalgo (1981) 123 Cal.App.3d 301, 308 [176 Cal.Rptr. 463] (Tidalgo) [finding substantial evidence that “it was unreasonable for officers to believe that the [searched] residence was occupied or owned by respondent”]; People v. Alders (1978) 87 Cal.App.3d 313, 317-318 [151 Cal.Rptr. 77] [invalidating search of “a distinctly female coat” because “there was no reason to suppose [the coat] was jointly shared by [the female defendant] and [the male probationer]”.)
The cases above indicate that “search conditions should be literally construed in order to protect the rights of both probationers/parolees and nonprobationers/nonparolees who associate with the individual subject to the *942particular search condition.” (Tidalgo, supra, 123 Cal.App.3d at p. 306.) The cases also demonstrate that whether an officer reasonably believed an item of property is under a parolee’s “control” within the meaning of the parole search condition must be determined case by case based on indicia of ownership, authority, or possession, not mere physical access. By focusing the inquiry in this way, the cases give due regard to the fact that property that does not belong to the parolee likely belongs to someone else. To equate “control” with “ability to exert control,” as the court does today, eviscerates any limit on the scope of a parole search as applied to the open areas of a car’s passenger compartment, notwithstanding the privacy interests of third parties.
Because this rule cannot be correct when followed to its logical conclusion, the court drops a footnote that purports to leave undecided whether a search of “closed-off areas” such as “the glove box, center console, or trunk” can “be based solely on a passenger’s parole status.” (Maj. opn., ante, at p. 926, fn. 16.) But why should such closed-off areas be exempt (at least for now) from the rule announced by the court today? After all, it seems entirely plausible that a front seat passenger who seeks to hide contraband from an approaching officer would put it in the glove box or center console, both of which are areas within the parolee’s “ability to exert control” and areas where the parolee “could have stowed personal belongings or discarded items when aware of police activity.” (Id. at p. 913; see, e.g., People v. Chavers (1983) 33 Cal.3d 462, 466 [189 Cal.Rptr. 169, 658 P.2d 96] [firearm found in glove compartment]; People v. Walker (1969) 273 Cal.App.2d 720,123 [78 Cal.Rptr. 439] [two firearms found in glove compartment]; People v. Prochnau (1967) 251 Cal.App.2d 22, 25-26 [59 Cal.Rptr. 265] [two firearms found in glove compartment of car driven by parolee]; People v. Allen (1967) 254 Cal.App.2d 597, 602 [62 Cal.Rptr. 235] [drugs found in glove compartment].)
The court declines to follow its own reasoning to this obvious conclusion on the ground that “[t]he reasonableness of such a search must necessarily take into account all the attendant circumstances, including the driver’s legitimate expectation of privacy in those closed compartments, the passenger’s proximity to them, and whether they were locked or otherwise secured.” (Maj. opn., ante, at p. 926, fn. 16.) That is exactly right. But why shouldn’t the reasonableness of searching the backseat or other parts of a car’s passenger compartment likewise turn on such contextual factors? There is no reason , to evaluate the reasonableness of searches of closed compartments through an analysis that differs from the analysis used to evaluate searches of other parts of a car. The same analysis applies: “we ‘examin[e] the totality of the circumstances’ to determine whether a search is reasonable within the meaning of the Fourth Amendment.” (Samson, supra, 547 U.S. at p. 848.)
*943The court notes that “a driver has a reduced expectation of privacy with regard to an automobile.” (Maj. opn., ante, at p. 924.) That is true, but the automobile search cases serve only to highlight how (literally) unprecedented today’s opinion is. This is not a case involving an inventory search of an impounded vehicle (Arizona v. Gant (2009) 556 U.S. 332 [173 L.Ed.2d 485, 129 S.Ct. 1710]; South Dakota v. Opperman (1976) 428 U.S. 364 [49 L.Ed.2d 1000, 96 S.Ct. 3092]), a search incident to arrest (New York v. Belton (1981) 453 U.S. 454 [69 L.Ed.2d 768, 101 S.Ct. 2860]), a protective search for weapons based on specific, articulable facts indicating that a suspect is dangerous (Michigan v. Long (1983) 463 U.S. 1032 [77 L.Ed.2d 1201, 103 S.Ct. 3469]), or a search for a vehicle identification number, license, registration, or other information that a driver is legally obligated to disclose (New York v. Class (1986) 475 U.S. 106 [89 L.Ed.2d 81, 106 S.Ct. 960]; In re Arturo D. (2002) 27 Cal.4th 60 [115 Cal.Rptr.2d 581, 38 P.3d 433]). This is not a case where the police found contraband in plain view. (Maj. opn., ante, at pp. 929-930.) And this is not a case where the police had probable cause to believe that the car or the belongings or containers within it contained evidence or contraband. (Wyoming v. Houghton (1999) 526 U.S. 295 [143 L.Ed.2d 408, 119 S.Ct. 1297]; California v. Acevedo (1991) 500 U.S. 565 [114 L.Ed.2d 619, 111 S.Ct. 1982].) This case involves a suspicionless, warrantless search of the passenger compartment based solely on the police officer’s knowledge that the front seat passenger was on parole. The automobile search cases have never hinted, much less held, that this kind of search is valid under the Fourth Amendment. Instead, the automobile search cases, while premised on a reduced expectation of privacy, recognize that “[a] citizen does not surrender all the protections of the Fourth Amendment by entering an automobile.” (Class, supra, 475 U.S. at p. 112.)
In sum, because the driver and other passengers in a car retain a legitimate if diminished privacy interest, it is important to properly delimit the scope of a lawful parole search when the front seat passenger is a parolee. The court cites no authority for construing the term “control” in the parole search condition (Cal. Code Regs., tit. 15, § 2511, subd. (b)4) to mean the mere “ability to exert control.” By contrast, there is ample precedent suggesting that “property under [the parolee’s] control” (ibid.) is most sensibly read to mean property over which the parolee appears to have authority, possession, or ownership from the perspective of a reasonable police officer. This latter reading, applied to the circumstances of each case, is what distinguishes the proper scope of a parole search from an unlawful intrusion on someone else’s privacy.
II.
The court cites Maryland v. Pringle (2003) 540 U.S. 366 [157 L.Ed.2d 769, 124 S.Ct. 795] (Pringle) and People v. Vermouth (1971) 20 Cal.App.3d 746 *944[98 Cal.Rptr. 65] (Vermouth) for the proposition that “the law does not presume that a front seat passenger has nothing to do with items located elsewhere in the passenger compartment of a car.” (Maj. opn., ante, at p. 924.) That is true, but equally important—and contrary to today’s holding—Pringle and Vermouth show that whether a police officer reasonably believes there is a relationship between a front seat passenger and items elsewhere in the passenger compartment depends on the totality of the circumstances in each case.
In Pringle, a police officer stopped a car for speeding and obtained the driver’s consent to search the car. The officer found a roll of cash amounting to $763 in the glove compartment and five plastic baggies of cocaine behind the backseat armrest. The officer asked the three occupants of the car about the ownership of the drugs and money. When none of the car’s occupants claimed ownership, the officer arrested all three. Pringle, who was the front seat passenger, argued that the officer lacked probable cause to arrest him. In upholding the arrest, the high court noted that the cocaine behind the backseat armrest was “accessible to all three men.” (Pringle, supra, 540 U.S. at p. 372.) But the high court further observed that the rolled-up cash was “in the glove compartment directly in front of Pringle” {ibid.), that “[u]pon questioning, the three men failed to offer any information with respect to the ownership of the cocaine or the money” {ibid.), and that “[t]he quantity of drugs and cash in the car indicated the likelihood of drug dealing, an enterprise to which a dealer would be unlikely to admit an innocent person with the potential to furnish evidence against him” {id. at p. 373). Based on these facts, and not the mere accessibility of the cocaine, the high court found it “an entirely reasonable inference . . . that any or all three of the occupants had knowledge of, and exercised dominion and control over, the cocaine.” {Id. at p. 372.)
In Vermouth, the police stopped a vehicle at 3:30 a.m. for having a defective rear light. During the stop, the police questioned the two male occupants of the car and obtained consent to search the car and trunk. The search turned up, among other things, a five-inch pocket knife sticking out between the instrument panel and tape-playing machine, a billy club resting against the driver’s door, and a pair of wire strippers and a small paring knife from the backseat area. (Vermouth, supra, 20 Cal.App.3d at pp. 750-751.) Inside the trunk, the police found “an electronic unit or part of a stereo, a small speaker, and a tool which at one end had a tire lug wrench and at the other a prying edge. The prying section had fresh, long scrape marks on it.” (Id. at p. 751.) The police arrested both men for burglary and unlawful possession of a billy club. Although the driver had claimed ownership of the billy club during the stop, the Court of Appeal held that the police had probable cause to also arrest the passenger for possession of the billy club because the facts gave rise to an inference that “there was an alliance *945between [the passenger and the driver], either defensive or offensive.” (Id. at p. 756.) The reasonableness of arresting the passenger for possession of the billy club turned on the “alliance” between the two men as inferred from the totality of the circumstances, not on the mere accessibility of the billy club to the passenger.
Pringle and Vermouth thus follow the settled principle that the reasonableness of police conduct under the Fourth Amendment must be determined based on the facts of each case. We should follow the same approach. As those cases demonstrate, a police officer in the ordinary conduct of a lawful automobile stop has many ways to determine what areas or items of property in a car are within a passenger’s control. Requiring each case to be decided on its own facts would not place an undue burden on law enforcement.
In the present case, Officer Mihai had lawful authority to search the parolee’s person and the area immediately adjacent to the parolee. Absent unusual circumstances, a further search of the passenger compartment would have required Officer Mihai to make a reasonable determination of what areas or property in the car were under the parolee’s control. Officer Mihai could have done this by asking a few simple questions of the driver, the parolee, or the other passengers. For example, “does that chips bag belong to any of you?” If the parolee had said yes, then it would have been subject to search. If no one had claimed the chips bag—a scenario with some likelihood, since the chips bag contained contraband—then Officer Mihai could have lawfully searched it because she would have had a good faith belief that the search would invade no one’s privacy. Upon discovering the contraband, she would have been justified in arresting the occupants of the car (see Pringle, supra, 540 U.S. at p. 372) and searching the rest of the car for additional contraband. Even if a passenger other than the parolee had claimed the chips bag, Officer Mihai could have asked whether the chips had been shared with the parolee. As to any such question, the officer need “not be bound by the reply in the face of overwhelming evidence of its falsity.” (Boyd, supra, 224 Cal.App.3d at p. 749.) The officer could have taken into account the relationship among the car’s occupants, the demeanor of the individuals questioned, and any other verbal or observable indicators of what areas or property in the car were under the parolee’s control.
In sum, there are many ways that Officer Mihai could have developed a reasonable belief that the chips bag was under the parolee’s control and thus within the scope of a parole search. Contrary to what the court suggests (maj. opn., ante, at p. 929), my approach would not require Officer Mihai to have witnessed the front seat passenger tossing contraband into the backseat area. It requires only that an officer have a reasonable belief that the area or item *946searched is owned, controlled, or possessed by the parolee. And the reasonableness of an officer’s belief will depend on the totality of the circumstances in each case.
In adopting a novel categorical rule instead of the case-by-case approach dictated by precedent, the court appears motivated by “law enforcement’s need for a workable rule to monitor parolees.” (Maj. opn., ante, at p. 927.) But the court places artificial limits on its own theory. For example, the court limits its ruling to a noncommercial, standard five-passenger car, apparently leaving open the permissible scope of a parole search involving sport utility vehicles or passenger vans, even though in those vehicles no less than in a five-passenger car, “[t]he driver is not necessarily in a position to supervise his passengers at every moment, nor is he in a position to control their every move once they are in the car” and “an occupant of an automobile may hide contraband without the other occupants’ knowledge or permission.” (Id. at p. 925.) Moreover, as noted, the court declines to apply its ruling to “closed compartments” (id. at p. 926, fn. 16), even though such compartments are places where a parolee may readily stow or discard evidence or contraband.
Further, if there is any logic to the court’s exception for closed compartments like the glove box or center console, it would seem also to exempt closed containers found in the car. But whether a container is open or closed can itself be problematic. In the present case, the limited record states only that two syringes were found in “a bag of chips” in the backseat of the car. It is not clear whether this bag of chips should be treated as an open or closed container. The record does not reveal whether the bag was wide open, partially open, flattened or clipped at the top, or otherwise sealed. If “it was objectively reasonable for the officer to believe that the parolee was able to reach back and conceal contraband inside the chips bag” (maj. opn., ante, at p. 931), then "what about a shoebox whose cover is partially off? Or a backpack whose zipper is partially open?
The stock response is to say that these concerns are not presented on the facts of this case and that we can decide such issues when they arise. (Maj. opn., ante, at p. 926, fn. 16, p. 930, fn. 22.) Of course, there is nothing wrong with that response insofar as precedent dictates that reasonableness under the Fourth Amendment requires case-by-case analysis. But if that is the response, then it is all the more inexplicable why the court arbitrarily cordons off some cases, but apparently not others, from the usual fact-specific inquiry.
Finally, it cannot be said that today’s opinion has the virtue of bringing certainty to situations where there previously was none. Before today, there was no uncertainty for many law-abiding citizens who would, for any number of innocuous or even virtuous reasons, occasionally or regularly give a ride *947to, or ride with, a stranger or mere acquaintance. Under “commonly held social conventions” (maj. opn., ante, at p. 926), those citizens had no reason to think that a front seat passenger invariably would have “control” over the open areas of a car’s passenger compartment. They had no reason to think that the passenger compartment or items of property located in it would be subject to a suspicionless, warrantless search simply because the front seat passenger turned out to be a parolee.
After today, we must all be more wary of the company we keep when we drive or ride in a car, lest we surrender our legitimate expectations of privacy. Although Fourth Amendment doctrine is built on cases involving guilty people, it is important to remember that “the ‘reasonable person’ test presupposes an innocent person.” (Florida v. Bostick (1991) 501 U.S. 429, 438 [115 L.Ed.2d 389, 111 S.Ct. 2382].) Just as “[m]any law-abiding citizens might choose not to open their homes to probationers if doing so were to result in the validation of arbitrary police action” (Robles, supra, 23 Cal.4th at p. 799), so too many law-abiding citizens might decline to spare the air, to help a person in need, or to otherwise be a good Samaritan because doing so may result in a suspicionless invasion of their privacy.
For the reasons above, I join the court in reversing the judgment of the Court of Appeal but respectfully disagree with the erosion of Fourth Amendment protections worked by today’s opinion.