dissenting.

Defendant’s conviction rests on the discovery of a handgun after a warrantless and unconsented search of a closed bag located in the locked trunk of defendant’s automobile after defendant was stopped and arrested on unrelated charges, removed from the automobile, and handcuffed. From prior information the officers had reason to expect to find a handgun in a bag somewhere in the vehicle. Assuming that the information gave the “probable cause” required by Article I, section 9, of the Oregon Constitution for a search of one’s person, house, papers, or “effects,” the question is whether the search could be conducted without the warrant ordinarily required by that section.^[1]

The state argues primarily that the search can be[*280] brought within the exception of searches incident to a person’s arrest. Alternatively, the state asks this court to recognize an “automobile exception,” as that phrase has come to be used for cases decided under the federal fourth amendment, contrary to the view of the Court of Appeals that “[t]here is no ‘automobile exception’ under the Oregon Constitution.”^[2]

The majority today affirms defendant’s conviction neither on the theory that the search was “incident to” defendant’s arrest nor on the United States Supreme Court’s current version of an “automobile exception” from the warrant requirement of the fourth amendment, but upon a more narrowly defined exception limited to reasonable searches on probable cause of vehicles lawfully stopped in transit, which is based on a theory that searches in such a setting are always “exigent.” The majority’s discussion of modernizing the warrant process suggests that today’s decision is a temporary accommodation that remains open for future reconsideration.

Because the court does not accept the broader arguments for dispensing with search warrants for the closed compartments of automobiles generally, my disagreement with the majority’s narrow holding is correspondingly narrow. I do not agree that in every case, the search of the trunk of a mobile vehicle, once it has been stopped, is “exigent per se” regardless of individual circumstances. A rule authorizing warrantless police searches of automobiles on probable cause alone, without actual circumstances of necessity, should be enacted by politically accountable elected policymakers rather than by this court, before we decide its constitutionality. Because I disagree with the majority’s contrary ruling, I also must explain why the state’s other theory does not justify this warrantless search.

I.

This court last undertook to “clarify the law relating to the search and seizure of vehicles” in four decisions in 1979.^[3] I wrote at the time that the court’s stated aim was “a[*281] worthy ambition but an incautious one,” because “the goal of a clear and comprehensive statement of the law of Oregon governing police searches of vehicles or of anything else will not be achieved by the means by which the Court pursues it.” State v. Greene, 285 Or 337, 345, 591 P2d 1362 (1979) (Linde, J., specially concurring.)^[4]

Seven years have passed, but that statement is as true today as it was seven years ago. In today’s opinions in State v. Bennett, 301 Or 299, 721 P2d 1375 (1986), and in this case, three judges articulate a rule for searching automobiles and their contents in one kind of situation and the Chief Justice, although “troubled,” joins to make a majority with the aim of “putting the question to rest.” 301 Or at 308.

The rest will not be a good night’s sleep. Today’s opinions, too, remain transitory not only because the Court itself is in transition — the “committee attempting to draw a horse [while] the membership of the committee constantly changes,” to which Professor Anthony Amsterdam ascribed the “camel” of fourth Amendment Law^[5] — but also because the opinion rests upon premises of “exigency” and the limitations of the warrant process that can be, and therefore should be, overcome.

Search and seizure law has not stood still in the years since this court’s 1979 decisions, nor has the United States Supreme Court been more successful than this court in finding a coherent rationale for searches of vehicles and containers in vehicles. The Supreme Court held that the federal fourth amendment required search warrants to open luggage and packages. Then, with Justice Stewart writing for a plurality just before his retirement, the Court in 1981 limited the area of an automobile that could be searched incident to an arrest to the passenger area and containers in that area, excluding the car’s trunk. Within a year the Court abandoned[*282] that limitation both for trunks and for containers found within them.^[6]

There were independent developments in this court’s decisions. The 1979 majority opinions still followed a view stated in 1974 that tied Oregon’s law to federal decisions unless there were “persuasive reasons” to the contrary.^[7] That view, which treated each latest formulation of the United States Supreme Court as a benchmark for the Oregon Constitution, had not been followed in other contexts,^[8] and it was overruled with respect to search and seizure law in State v. Caraher, 293 Or 741, 653 P2d 942 (1982). Caraher and other decisions reasserted the responsibility of Oregon courts to make independent determinations of Oregon law in criminal procedure as much as in other areas which might or might not correspond to federal decisions in method or in result.^[9]

[*283] As to searches and seizures, the court has reiterated that under Article I, section 9 of the Oregon Constitution, the state may not prosecute a person with evidence obtained in violation of law that is designed to protect the person against the kind of intrusion by which the evidence is seized,^[10] that the requirement of a search warrant is the norm and warrantless searches are unlawful with limited exceptions,^[11] that an arrest may be accompanied by a search for evidence only of the crime for which the arrest was made,^[12] that a warrantless seizure and a warrantless search are distinct events requiring separate justification,^[13] and that officers may not search the belongings of a person held for reasons other than a criminal charge and may inventory them only under validly promulgated and uniformly applied administrative procedures.^[14]

Accordingly, the majority faithfully recites that “the warrant requirement * * * may be dispensed with in only a few specifically established and well-delineated circumstances” and that “we are deciding this case independent of federal law.” 301 Or at 273.

The rule that searches must be judicially authorized whenever possible and that warrantless searches are extraordinary departures from the rule deserves to be more than a comforting fable. Routinely recited in opinions sustaining[*284] another warrantless search, it becomes an empty ritual. As for independent analysis of Oregon law, the majority states: “We cite the United States Supreme Court decisions only because we believe they are persuasive, not because they are precedent for this court in interpreting the Oregon Constitution.” 301 Or at 274.

It may be tempting to adopt another court’s reasoning by reference rather than to spell out one’s own, but in areas such as search and seizure law, quotations only beg the question why the quoted opinion is more persuasive than other opinions or academic critiques that are not quoted. In a 1985 decision applying California’s search and seizure clause, the California Supreme Court, like today’s majority, wrote: “[W]e find the reasoning in [federal cases] persuasive.” People v. Ruggles, 39 Cal 3d 1, 11, 216 Cal Rptr 88, 702 P2d 170 (1985). The difference is that what the California court found persuasive were the United States Supreme Court’s decisions in Chadwick, Sanders, and Robbins, supra, n. 6. The California court rejected the position taken in United States v. Ross, 456 US 798, 102 S Ct 2157, 72 L Ed 2d 572 (1982), and invalidated the warrantless search of bags found in a car trunk despite abundant probable cause for the search.

If the present cases had come to this court early in 1982, after the Supreme Court’s decision in Robbins and before Ross, a majority of this court similarly might have found the then latest word from the Supreme Court “persuasive” on the meaning of Oregon’s Article I, section 9. People v. Ruggles, supra, strikingly demonstrates the costs of reversing logic and deciding cases on federal grounds before reaching state issues. When an intermediate California court originally sustained the search, the United States Supreme Court remanded the decision for reconsideration in light of its decision in Robbins. The California court reconsidered and reversed itself, only to face another grant of certiorari and a further mandate to reconsider once more in light of the Supreme Court’s change of direction in Ross. On the second remand the California Supreme Court finally decided the case under the California constitution. 39 Cal 3d at 7, n. 3. Cf. State v. Kennedy, 295 Or 260, 666 P2d 1316 (1983). Fortunately the majority does not resurrect this approach.

An independent analysis under a state constitution is[*285] exemplified by the Washington Supreme Court’s opinion in State v. Ringer, 100 Wash 2d 686, 674 P2d 1240 (1983), which also involved searches of vehicles after their drivers had been arrested and handcuffed outside the vehicles. The decision rejected the same arguments made by the state here. Justice Dolliver’s opinion for the court noted that, beginning in the 1960s,^[15] the Washington decisions began to follow the United States Supreme Court’s fourth amendment doctrines in applying Washington’s constitution, disregarding the state’s own law. Returning to the common-law origins of searches incident to arrest, the Ringer court stated that the exception “began as a narrow rule intended solely to protect against frustration of the arrest itself or destruction of evidence by the arrestee,” and that this was its scope when the Washington constitution was adopted. 100 Wash 2d at 698. The court also observed that it early had refused to recognize a “so-called ‘automobile exception’ for nonexigent searches.”^[16] I do not suggest, and I do not cite State v. Ringer to suggest, that old cases or understandings contemporaneous with the adoption of a constitutional text place its interpretation and application beyond further development. But the development must respect the logic of its premises.

II.

The conditions of law enforcement have changed since the adoption of the fourth amendment and its state counterparts in the 18th and 19th centuries. The notion of crime as an exceptional and abnormal event calling for a judge to authorize someone to seize a known offender or to search[*286] for him or for stolen goods in a specified place may never have been accurate. In this century, those simple assumptions have undergone radical changes. The growth of cities and the use of automobiles and airplanes gave offenders new mobility and anonymity. Law enforcement was delegated to and administered by large police organizations, and governments chose to rely on criminal law in attempts to control first alcohol and then other drugs by prohibitions on production, trade, possession, and use that depend on finding, seizing, and identifying evidence without the aid of a victim. The entire process still is legally required to be supervised by judges, but the task is no more popular with courts than it is with law enforcement officers.

Three devices can be tried to evade the warrant requirement. One is to deny that it exists, on the theory that under the constitutional text warrantless searches need only be “reasonable,” and that only warrants must be based on probable cause and particularly describe the place to be searched and the person or thing to be seized. This theory would obliterate the use of warrants, because officers could search anywhere and seize anything without a warrant and without probable cause as long as their conduct could be defended as “reasonable.” Neither the United States Supreme Court nor this court has accepted that reading of the guarantee.^[17]

A second device for evading the warrant requirement is to create or expand exceptions for searches or seizures that of necessity must be made without awaiting a warrant (specifically searches of an arrested person or for items that an officer cannot keep under control) beyond their original and necessary scope. The majority’s “per se exigency” exception for vehicles stopped in transit, irrespective of actual exigency in the individual case, is such a device.

A third device is to manipulate what constitutes a “search” requiring a warrant. In ordinary speech, “search” connotes a purpose or intention of the searcher; one searches whenever one acts with the aim of finding something. But it has long been held that purposeful looking or listening alone[*287] does not make unlawful a warrantless search of what can be seen in plain view or overheard without the aid of technical enhancement. See, e.g., State v. Louis, 296 Or 57, 672 P2d 708 (1983). In recent years, the United States Supreme Court has taken a concept first developed to carry the protections of the fourth amendment beyond trespass law, the concept of “privacy,” and has used it instead to limit the requirement of warrants for what otherwise are conventional “searches” by denigrating “expectations of privacy” in containers or other private property in various settings, such as automobile trunks.

The “privacy” device need not be dissected in detail here, since the majority does not adopt it. In essence, however, “privacy” relativism only reinvents the relativism of “reasonableness” under another label. Under the constitutional warrant clauses, downgrading “privacy” in automobiles fails as a limiting device; for if one’s privacy interests, though described as “diminished,” suffice to recognize the investigatory conduct as a “search,” the warrant clauses require probable cause and a warrant. If the conduct is not a “search” in the legal sense at all, then not only is no warrant required but neither is probable cause or even “reasonableness” under the first clause of the constitutional texts. See Gardner, Searches And Seizures Of Automobiles And Their Contents: Fourth Amendment Considerations In A Post-Ross World, 62 Neb L Rev 1, 13 (1983). Sliding-scale “privacy” cannot logically rationalize when a search may and when it may not be made without a warrant, nor can it in practice guide police officers, prosecutors, and trial courts in a new factual setting until its degree of privacy has been calibrated by an appellate opinion. The court is right not to start down that road.

Cases on warrantless searches and seizures of vehicles and of various containers involve enough variables to allow almost any precedent to be distinguished in a later case, variables that have been thought decisive in some opinions. (1) The vehicle to be searched may be stopped by the police, or it may already be stationary. (2) The vehicle may be movable at will or it may be physically immobilized. (3) The purpose of the stop may be related or unrelated to the search. (4) The object of the officer’s intervention may be the driver or a passenger. (5) The driver, passenger, owner, or other possessor of a vehicle may be subjected to arrest, or only cited, or[*288] not at all related to the probable cause for the search. The same is true of persons found in possession of suitcases, shipping boxes, or other containers. (6) Such containers are variously found in private premises, in private vehicles, in the possession of common carriers or in warehouses. (7) The vehicle may itself be the object, not only the site, of the officer’s investigation. (8) The vehicle may be on residential property, on privately owned parking premises, or on a public road. (9) An officer may have probable cause to believe either that a vehicle or that a container holds items subject to seizure, and the belief may result from his own observation or from other information of the contents. (10) The officer may or may not have used instruments beyond his own natural senses. (11) The objects of the search or the seizure may or may not be intrinsically dangerous or ephemeral. Finally, (12) the jurisdiction may have enacted one or another law governing the officer’s conduct.

Faced with the myriad combinations and permutations of these variables, one may retreat into matching the facts of each new case to the accumulating mass of prior holdings in the faith that some rational system will emerge from the empirical experience even if not from the opinions. The experience shows that faith to be misplaced. Opinions of the United States Supreme Court or individual justices have stated about the Court’s vehicle search cases, in 1971 that “no trick of logic will make them all perfectly consistent”;^[18] in 1973, with similar understatement, that the cases formed “something less than a seamless web;”^[19] and with more exasperation in 1982: “The law of search and seizure with respect to automobiles is intolerably confusing. The Court apparently cannot agree even on what it has held previously, let alone how these cases should be decided.”^[20] One scholar put his finger on the problem: “After more than fifty years of search and seizure litigation, neither federal nor state courts have established a satisfactory conceptual framework for analyzing the rather basic issue of when prior judicial approval is required to validate a search.” Grano, Rethinking [*289] the Fourth Amendment Warrant Requirement, 19 Am Crim L Rev 603 (1982). What the life of the law of search and seizure needs is more logic, not more experience.

Ill

The state offers two theories to defend the war-rantless searches of the closed automobile trunks in these cases and, in State v. Bennett, 301 Or 299, 721 P2d 1375 (1986), of the closed containers found in the trunk. One theory is that these searches required no warrant because they were “incident” to an arrest of the drivers. The other theory is that there should be an “automobile exception” to the warrant requirement. The Court accepts a version of the second theory. The holdings cannot be justified on either theory.

“Search incident to arrest. ” Because the Court does not rest its holdings on a theory of warrantless searches of car trunks “incident to arrest,” the inadequacy of that theory requires only a brief statement.

“Search-incident-to-arrest” has become one of those legal talismans that, although it is in English rather than Latin, has left its rationale behind. What justifies a war-rantless search of a person under arrest is not that there is probable cause to believe that the person has committed a crime. It is not that an officer who has valid reason to make an arrest also has valid reason to search for evidence of the crime for which he makes the arrest. If reasons sufficient for an arrest by themselves sufficed to dispense with a search warrant, the actual arrest would be immaterial and unnecessary for the search. That obviously is not the law.

To the contrary, an arrest is the crux of a “search-incident-to-arrest.” By definition a search “incident to” an arrest is a warrantless search that is justified by the fact that a suspect is arrested. The justification arises from the practical consequences of taking a person into custody, and it cannot extend beyond those practical consequences if the fact of the arrest is its premise. This rather self-evident proposition is demonstrated by assuming two variations from the typical case.

First, a person need not be in custody in order to be prosecuted. ORS 133.045 to 133.100 allow citations in lieu of arrest for most violations, for all misdemeanors, and for many[*290] felonies. Probable cause to believe that the suspect has committed a crime is required for a citation: as much as for an arrest. But no one maintains that Oregon law authorizes an officer to conduct a warrantless search of a person or a person’s house, papers, or effects by virtue of or “incident to” issuing a citation, although the crime, the officer’s knowledge, and the relevance of potential evidence are identical as in an arrest. The statutes assume that citation and arrest are equivalent means to initiate a prosecution, differing only according to the need to restrain the suspect. A doctrine that illogically extends a “search-incident-to-arrest” beyond the practical needs of custody to allow investigatory searches without a warrant contradicts that policy. Instead, it puts a premium on arresting the suspect, and if possible doing so in a place that officers wish to search for evidence.

Second, the statutes allow a “private person” as well as a “peace officer” to effect an “arrest,” ORS 133.220, which means “to place a person under actual or constructive restraint or to take a person into custody for the purpose of charging that person with an offense.” ORS 133.005(1). No doubt a private person lawfully taking another person into custody may safeguard himself and items in the suspect’s immediate possession that relate to the crime for which he makes the arrest. But would the state argue or a court hold that a “citizen’s arrest” authorizes a “citizen’s search” of an automobile trunk for proof of a crime, beyond any needs attendant on the arrest itself?

Possibly for these reasons, the majority does not justify the searches of the automobile trunks in this case and Bennett as incidents of the arrests. In a given situation some other necessity, generically labeled “exigent circumstances,” may permit a search or a seizure on probable cause without first obtaining a search warrant, regardless whether anyone is arrested. But if the fact of an arrest is to make the difference between a permissible and an impermissible warrantless search, the difference must arise from the necessary consequences of taking a person into custody. The fact of an arrest alone cannot logically bootstrap officers into any wider war-rantless search than they could conduct without making the arrest.

Exigency. Seeking the less untenable between two[*291] unsatisfactory theories, the majority embarks on rationalizing warrantless searches of automobiles as a new “exigent per se” exception to Oregon’s constitutional requirement of search warrants. The premise is that the mobility of a motor vehicle does not allow time to obtain a warrant to search it. Dating from Carroll v. United States, 267 US 132, 45 S Ct 280, 69 L Ed 543 (1925), it begins as an application of the wider exception that “exigent circumstances” can make a search or seizure conducted without awaiting a warrant, but with probable cause, not “unreasonable.”

The majority opinion leaves important questions unaddressed. In order to narrow the new rule, the exception applies only when officers have lawfully stopped the vehicle while in motion. This seems to reintroduce the idea that the search relates to the officers’ authority to detain the occupants traveling in a moving vehicle, an idea similar to a search “incident” to an arrest, although apparently no “arrest” in the statutory sense of “taking into custody” is required. It certainly reintroduces the problem of “pretext” stops and of the relationship between the grounds for the stop and the scope of the subsequent search.^[21]

The present case does not hold that probable cause to search a car for evidence of a crime unrelated to motor vehicle laws is itself sufficient reason to make a “lawful stop.” But officers apparently may be instructed that if they have probable cause to believe that a stationary vehicle contains items subject to seizure but have not obtained a search warrant, they should keep the vehicle under surveillance until someone begins to drive it away and then find an opportunity to make a lawful stop. If in some future case it appears that officers waited for that to happen longer than it would have taken to obtain a warrant, courts will once again have to narrow the “exigentper se” exception for automobile searches.

A well-advised felon who does not want his car searched, in turn, should learn to pull to the side of the road when he sees a police vehicle before the officers have a chance to stop him. If stopped, apparently he should remove any briefcase or other bag some distance from the car before the[*292] officers begin their search of the car, as long as the search of a car is what allows searching a closed container in the car, as the majority holds. The sense of these rules will not be apparent to the ordinary citizen or officer, and they have little to do with “exigency.” They do not strike me as clear and certain rules for which to sacrifice the constitutional requirement of a search warrant.

The fatal flaw in the majority’s position is its statement that “exigencies should not be determined on a case-by-case basis.” Exigencies are emergencies, circumstances that require urgent action; of course they arise case by case. The majority’s desire to rest its disregard of the warrant requirement on the exception for exigent circumstances and also to give police officers general permission for warrantless searches of automobiles irrespective of actual exigency leads only to an unresolved contradiction.

It is possible to state “guidelines,” in the majority’s word, for the kinds of circumstances that preclude waiting for a warrant before conducting a search of an automobile or of anything else so as to make it reasonable to proceed without a warrant. Such guidelines, however, must describe conditions of actual exigency, not whole classes.of objects such as vehicles that have been stopped. If anyone is to legislate for automobiles generally, it should be persons politically elected for that task (or the voters themselves), not this court. As a statement about “exigency,” the proposition that it always, or generally, is impossible to obtain a warrant to search a vehicle after it has been stopped in transit is simply contrary to fact, especially in cases where the occupants have been placed in custody outside the vehicle.

IV.

The principle that prior judicial authorization for a search must be the rule and warrantless searches a narrowly confined exception should be more than a comforting fable in appellate opinions, but in practice it is no more than that. According to a recent study by the National Center for State Courts, this appears from a number of examinations of search warrant practices over the past quarter century. Van Duizend, Sutton, & Carter, The Search Warrant Process x (1984). The authors of the study reported:

[*293] “The vast majority of searches are conducted without a warrant, usually with the consent of the suspect (or someone in legal control of the area to be searched) or incident to the arrest of the suspect. Delay and inconvenience were widely cited as the principal basis for officers’ reluctance to seek a search warrant.

<<* * * * *

“We also were told of a host of other strategies that police use to conduct a search without having to go to the trouble of obtaining a warrant, such as by timing an arrest so as to maximize the possibility of being in a position to conduct a search legally and seize any contraband that might be discovered as a result (e.g., by arresting suspected drug dealers in their cars rather than at home).” Id. at 19 (footnote omitted).

For Portland, Oregon, the study reported 41,303 “index crimes” in 1980 and only 319 search warrants. Id. at 18.^[22]

Facilitating warrants. The majority’s “per se exigency” exception for warrantless searches of vehicles stopped in transit cannot rest on a theory that holding the vehicle until a warrant is obtained (in the absence of consent to search) always is impossible, when in fact it is only burdensome, slow, expensive, and administratively inconvenient, given the large number of such vehicle stops and the limited resources of police and courts. At the end of its opinion, the majority notes that modern technology offers the means to overcome practical obstacles to compliance with the constitutional warrant requirements and sets out a plan for doing so. 301 Or at 278 n 6. I welcome this initiative toward restoring reality to the recitations that the constitution requires warrants for searches and seizures to be the rule and warrantless searches true “exceptions.” As I observed at the outset of this opinion, the suggestion serves as a reminder that the majority’s newly defined “Oregon automobile exception” is a transitory accommodation and is open to future reconsideration.

[*294] Nevertheless, it does not persuade me that under current conditions the new exception is justified by true “exigency.” The technology for communicating and recording affidavits and judicial warrants issued thereon exists today.^[23] If lack of official personnel and equipment in fact precludes obtaining a speedy warrant, this is the government’s (including the Judicial Department’s) responsibility, not that of the individual citizen. We are dealing with the constitutional rights of all drivers, passengers, or owners of vehicles, not only with searches in which evidence of unlawful conduct in fact is found. A person whose automobile is stopped by police officers and who does not wish to consent to a search of its trunk (or of any other “effect”) should not have to dispense with the protection of a judicial warrant because the state does not provide enough readily accessible officials and processes to seek such warrants. Officials, equipment, and procedures will be provided more expeditiously when this court enforces the warrant requirement than when it holds that a warrant may be omitted whenever the procedures for getting a warrant are slow and inconvenient for everyone concerned.

Consent or seizure. Although of the devices for evading the warrant requirement, the “exigency” theory of the majority opinion is the least untenable, its extension to all stopped automobiles rests on a faulty assumption. The majority writes that to conduct an immediate on-the-scene search of a vehicle stopped by the police is no greater intrusion than to hold the vehicle until a warrant is obtained. The same argument could be made for searching other movable private effects, effectively reducing the practical use of search warrants to houses and other immovable objects.

The faulty assumption is that the court must choose between the “intrusiveness” of an immediate search and of a temporary seizure to await a warrant and must make that[*295] choice as a categorical matter of law. That is not so. The obvious, and correct, alternative is that the choice belongs to the person whose constitutional interests are at stake. An officer reasonably believing that he has probable cause to search an automobile trunk in the presence of the owner or driver can offer the person an informed choice between consenting to an immediate search or having the automobile held for the time necessary to obtain a warrant.

This is even more obviously true of bags or other closed containers. The person, not the officer, is the one to decide whether to insist on the right to have the supposed probable cause tested by a magistrate and to accept the inconvenience of the necessary seizure. There simply is no basis for this court or any court to make such a categorical choice for all owners of automobile trunks or closed containers found in automobiles as a class.^[24] If a person insists on the required warrant, there well may be exigent circumstances for a seizure when there are none for searching a container after it has been seized. Because I believe that there is no “exigency per se” for searching the trunks of all automobiles stopped by-police officers, a fortiori there is no such justification for opening closed bags or packages found therein.

V.

I return to what I wrote in State v. Greene, supra. The uncertainty of police authority to stop, to question, and to search persons and their belongings — the confusion about which the Chief Justice complains — represents a failure of lawmaking. It is not, in the first instance, a failure of the courts, because lawmaking for the exercise of executive power should not be left to courts in the first place. Investigatory and enforcement powers of state agencies other than criminal law enforcement rest on and are governed by laws, and there is no good reason why investigatory and enforcement powers over[*296] persons for potential criminal prosecution, of all things, should be less dependent on laws than, for instance, authority to inspect a restaurant for cockroaches. The laws may be statutes, home rule charters, ordinances, or rules made by delegation from any of these, as long as they can be traced back to politically elected lawmakers or to the voters themselves. Many executive officials also are elected and politically accountable, particularly in Oregon, but they are not elected to define and authorize their own authority over private persons and property.

Yet we continue to see fallacious arguments that police have this or that authority because a court has held that the conduct at issue does not violate the state or the federal constitution. Courts do not authorize executive actions; laws do. A court on its own can authorize law enforcement actions (aside from carrying out orders of the court) only by the warrants authorized by the constitution in Article I, section 9.

Of course any law authorizing executive action, such as a search or a seizure, must comply with the constitution. Because courts, with more or less enthusiasm and success, have sought to apply the constitutional limits on government actions in individual cases, both lawyers and lay citizens have come to see questions of police practices as if they were a controversy between law enforcement officials and courts. Proposals for systematic lawmaking have met with arguments, often from lawyers, to “leave it to the courts,” and legislators, with enough other controversies clamoring for their attention, have understandably been pleased enough to do just that.^[25]

The majority traces its “automobile exception” from Carroll v. United States, supra. The Chief Justice asserts that Carroll “was and is good law.” Neither opinion notes or discusses the fact that Carroll sustained a search authorized by a statute, a section of the National Prohibition Act. The policy[*297] to authorize seizures of vehicles transporting intoxicating liquors was debated and enacted by Congress, not by the Supreme Court. What the Carroll court did was to sustain the statutory authority to search on the ground that similar statutes were enacted contemporaneously with the fourth amendment.^[26]

Clear and reliable rules for police conduct are desirable, as long as they do not violate the constitutional right of the people to be secure in their persons, houses, papers, or effects, against unreasonable, warrantless searches or seizures. Perhaps a law authorizing such acts can be sustained if its reach stays within constitutional bounds in the great majority of situations, and cases in which its application crosses the boundary can be dealt with by statutory interpretation or individualized decisions. Cf. State v. Moyle, 299 Or 691, 705 P2d 740 (1985), for a similar analysis under the free speech guarantee of Article I, section 8. But what should a court do when lawmakers make no such laws?

In his concurring opinion to State v. Bennett, 301 Or 299, 721 P2d 1375 (1986), the Chief Justice lists six possible rules to govern warrantless searches of automobiles and their contents. 301 Or at 306. Any of these is as clear and workable as any other. The test of a court is whether, in the absence of legislative direction, the court chooses individual liberty or executive power.

Regrettably the majority chooses to subordinate individual liberty. It states a rule in the name of “exigency” that goes beyond any actual need to dispense with a warrant in many circumstances. In pursuit of clear rules for the most basic confrontation between the government and the citizen, the majority is prepared to err on the side of the government and leave to legislators the burden to redress the balance if they so choose.

My view is the opposite. Like the Chief Justice and[*298] the majority, I also favor clear, practical rules for law enforcement actions. The simpler a rule is, however, the more it is likely to include some actions that the constitution forbids or to exclude some actions that constitutionally could be authorized. It is one thing for a court to give operational effect to a constitutional guarantee by “bright line” standards that are more protective in some instances than the guarantee strictly requires and let the legislature enact a different rule if it disagrees. But it is wrong for a court, as this court does today, to push “bright line” rules into a domain that includes many cases not defensible by constitutional premises, inviting legislators to take away governmental powers that the court has granted. It is the government that must ask lawmakers for authority against the citizen, not the citizen that must ask lawmakers to enact laws against “inherent” official power.

As I have said, today’s decision contemplates eventual change by the modernization of warrant procedures of which the majority writes. A rule that rests on a fiction, like today’s rule of “per se exigency,” has a limited life expectancy. The source of the governing premises remains the constitutional guarantee of a judicial warrant stated in Article I, section 9, enacted by the people in 1859, not what judges write about it. The guarantee will remain for future judges to apply, as long as the people do not choose to amend the constitution to sacrifice that guarantee. I would not sacrifice it on their behalf in the name of efficiency. I therefore dissent.