concurring.

The Court decides, and I concur, that the City of Roseburg cannot legally refuse to bargain with its firefighters under the state’s Public Employe Collective Bargaining Act on the grounds that PECBA alters the city’s charter or denies the right of referendum contrary to the “home rule” provisions of the Oregon Constitution.1 Nothing that has happened so far in this proceeding has that effect.

*289I write separately because the issues are difficult and statements made in explaining the present decision could be read to have unintended implications for future cases. For this reason it seems appropriate to state what I understand to be the grounds of this decision and what is not decided in this case.

Like the parties themselves, the Court takes its starting point of analysis in the principles stated in LaGrande/Astoria v. PERB, 281 Or 137, 576 P2d 1204 aff’d on rehearing 284 Or 173, 586 P2d 765 (1978). Those principles afford room for the state either to set substantive terms and conditions of employment for municipal employees or to assure such employees of procedural protections in the process by which these terms and conditions are determined.

First. Roseburg happens to have enacted its own procedures for bargaining collectively with its employees, which conflict with those prescribed by PECBA. Because we sustain the validity of PECBA, there is no occasion to consider whether the presence or absence of a conflict affects the restraint of article XI, section 2, upon the state’s authority. A statute which would “amend or repeal” one city’s conflicting charter provision might also be said to “enact” a charter provision for another city that has chosen not to legislate on the subject. Arguably any given city’s constitutional immunity from a state law should not depend on its formal political decision to enact a conflicting law, although a contrary argument also is possible. That question can arise when a state law is invalidated under article XI, section 2, or is sustained for lack of a conflicting *290municipal provision; it does not arise in this case. Cf. LaGrande/Astoria v. PERB, 281 Or at 151 n. 21, citing Boyle v. City of Bend, 234 Or 91, 98 and n. 6, 380 P2d 625 (1963).

Second. One basis for the Court’s decision sustaining PECBA is that it enacts a substantive state policy rather than a state-imposed procedure for local decisions. This is a complex issue that deserves careful scrutiny.

A state law plainly would be a substantive enactment if it were simply to specify pay scales, working hours, job security, insurance, or other fringe benefits for municipal employees. This is the same in constitutional principle as the state-mandated retirement benefits sustained in LaGr ande/Astoria and the other kinds of labor standards mentioned there, 281 Or at 153, even though it would represent a farreaching extension in political practice. A constitutional attack would need another basis than that the law does not lay down substantive standards. Nor would it matter that such a substantive state law displaces a local procedure for deciding the same issues. The limit is reached when such a law contains provisions incompatible with the city’s freedom to choose its political form. 281 Or at 156.

This remains true if the same kinds of terms and conditions of employment were set by a state agency under statutory authority delegated by the Legislative Assembly rather than by the statute itself. The fact that rules are made by an agency, and by administrative procedures in which the views of the interested parties are heard and considered, does not deny the resulting rules the same character of substantive state law that they would have if the legislature had enacted them. Contrary to the assumption of the dissent, such a state process for state-mandated local employment standards therefore need not be justified as a law imposing a state-mandated local process must be.

The more difficult question is which description fits the terms and conditions of employment that emerge from the process mandated by PECBA.

*291Unquestionably the prohibition of strikes by the firemen and certain other public employees, ORS 243.736,2 is a substantive state law, and its arbitration provisions are designed to balance that prohibition. But that alone does not make the arbitration provisions into a law fixing substantive labor standards rather than a process for arriving at such standards. Unquestionably, also, PECBA as a whole has social and economic objectives, as the opinion says. But such objectives also could be said to motivate many laws that would impose organizational forms and procedures on cities contrary to article XI, section 2. To escape that section on a “substantive state law” theory, it is not enough that a state law has a substantive “goal” or “purpose” to be promoted by prescribed procedural “means”; the state must itself define the substance of the standards it intends to achieve. The state did so in the law challenged in LaGrande/Astoria, which required cities to place police officers and firefighters under the state’s Public Employees’ Retirement System or provide equal or better benefits under another retirement system. One question in the present case, therefore, is whether PECBA similarly results in state-mandated substantive standards rather than only state-mandated procedures.

This is a close question. The fact that the results may not be uniform throughout the state is immaterial; state agencies often have authority to set different rates or standards of service to meet different local conditions. The majority opinion identifies two reasons why PECBA can be said to result in state-mandated substantive labor standards. One is that unless the parties agree otherwise, the arbitrator who resolves the impasse ultimately is selected by the Employment Relations Board, a state agency. The second reason is that the statute includes a list of “factors” to guide the arbitrators selected by the parties or by the board. Therefore the ultimate terms of employment in a city which does not reach agreement with its own firefighters are characterized as a rule imposed by a state agency according to statutory criteria.

*292I am not persuaded that PECBA as a whole fits this single characterization as a state procedure for arriving at state-mandated labor standards. Rather, the statutory scheme is at best a hybrid between state-administered labor standards and conventional, bilateral labor relations and dispute settlement. A city and its firefighters are entirely free, even encouraged, to reach their own agreement without reference to any public interest or public policy of the state other than the policy of collective bargaining itself. The “factors” listed in ORS 243.746(4) only govern arbitration when the parties do not agree. And they are stated only as “factors,” a word often used to avoid commitment to any ultimate test, value, or policy that the decisionmaker is directed to realize. Nor is it clear that the arbitrators under this act are expected to see themselves as agents for the state and its public interest in the same sense as a state administrative agency, rather than as an agency of the local community or as conventional, nongovernmental labor arbitrators.3 Either of the first two views, at least, obviates any issue whether a “nongovernmental” view of the arbitrators’ role would make the procedure an unlawful delegation of governmental authority to a nongovernmental decision-maker, see Or Const, art I, § 21, Hillman v. Northern Wasco County PUD, 213 Or 264, 323 P2d 664 (1958). See generally City of Detroit v. Detroit Police Officers Ass’n., 408 Mich 410, 294 NW2d 68, 83-95 (1980); Salt Lake City v. Firefighters Local 1645, 563 P2d 786 (Utah 1977).

Delegation issues have proved troublesome to other state courts that have considered similar statutory provisions for interest arbitration in public sector employment. The reported decisions do not always distinguish clearly between three different attacks: (1) on delegating authority to set terms of employment at all; (2) on the nongovernmental status of the arbitrators; or (3) on removing authority from the politically responsible local officials. *293The statutory “factors,” when treated as indicators of some coherent public policy channeling the arbitrators’ discretion, are generally held adequate to satisfy any modern nondelegation doctrine. The same statutory factors, however, coupled with provisions for judicial review, are sometimes also cited to show that the arbitrators are agents of the public. See, e.g., Town of Arlington v. Board of Concil. & Arbit., 370 Mass 769, 352 NE2d 914, 920-921 (1976); City of Amsterdam v. Helsby, 37 NY2d 19, 332 NE2d 290, 299-302 (1975) (Fuchsberg, J., concurring); City of Warwick v. Warwick Regular Firemen’s Ass’n., 106 RI 109, 256 A2d 206 (1969); contra, Salt Lake City v. Firefighters Local 1645, 563 P2d 786, 789-90 (Utah 1977); Town of Berlin v. Santaguida, 98 LRRM 3259 (Conn Super Ct 1978). The use of politically unaccountable, independent arbitrators chosen for the individual case, as an issue distinct from delegation as such, was given prominence in an opinion by Justice Levin in Dearborn Fire Fighters Local 412 v. City of Dearborn, 394 Mich 229, 231 NW2d 226 (1975). Following an amendment of the statute so as to provide a permanent panel of official arbitrators tied more strictly to statutory standards, the Michigan Supreme Court upheld the amended scheme in City of Detroit v. Detroit Police Officers, Ass’n. supra. Similarly, the Washington statute makes the arbitration panel a state agency. RCWA 41.56.450, sustained in City of Spokane v. Spokane Police Guild, 87 Wash2d 457, 553 P2d 1316 (1976) .

It should be understood that the three challenges to delegation are not the same. If an obligation to decide according to statutory standards were enough to make one a delegate of the state, many individuals and entities that administer regulated private institutions might be surprised to find themselves public officials. The ambiguity of the arbitrators’ role is no accident; it results from the deliberate wish to use an “independent” or “neutral” decisionmaker for decisions which, however, are more policymaking than judicial in the sense of applying preexisting norms to facts. Professor, now Judge, Grodin, put it this way:

“The problem of political responsibility arises initially from the fact that arbitrators may be called upon to determine policy issues otherwise subject to the local legislative *294process. But if the arbitrator is made politically responsible to the local electorate, which is in effect a party to the dispute, then arbitration loses its neutral character; and to the extent that the arbitrator’s constituency is the same as that of the legislative body that would otherwise exercise authority over the policy questions posed, the process becomes redundant.

“If, on the other hand, the arbitrator is made politically responsible to a larger electorate — for example, the electorate of the state as a whole — problems of neutrality and redundancy give way to problems concerning local autonomy and the role of collective bargaining. What was once a local issue, determinable ultimately through the local political process, becomes a state issue, determinable by state officials. ...”

Grodin, Political Aspects of Public Sector Interest Arbitration, 64 Cal L Rev 678, 693-94 (1976). It is this third attack on the delegation, the replacement of locally accountable officials by arbitrators, that is at issue in the present case.

Grodin notes that the arbitrators’ role is left ambiguous because collective bargaining is undercut precisely to the extent that the law assumes the existence of a substantive state policy for terms and conditions of employment, a policy which will take form in an eventual arbitration award. He finds the “heart of the dilemma” in the tension between the perspective that sees the arbitrator as “an essentially private person who happens to be involved in resolving a dispute concerning a public entity” and the legal necessity to legitimize the arbitrator as a public decisionmaker.4 Justice Tanzer would resolve the tension for purposes of the third, “home rule,” attack on the arbitration provision by characterizing the arbitrator as a state official making a substantive state decision pursuant to state criteria. This may well be the correct choice when the “neutral” arbitrator is chosen by the Employment Relations Board. It is tenuous when the parties choose an arbitrator without reference to any list of official arbitrators designated by ERB.

*295The present dispute has not reached the stage when it is necessary to characterize such an arbitrator. For this case, therefore, we may best describe PECBA as a hybrid under which some steps are conducted by the state itself through ERB (and arguably through an ERB-appointed arbitrator in case of impasse) and other steps (the actual bargaining and arbitration by an agreed, independent arbitrator) are state-mandated local procedure, which must have the justification stated in LaGrande/ Astoria.

There is an additional reason to be cautious in characterizing the eventual result of arbitration under PECBA as a state-imposed substantive rule, because this may have implications for the referendum issue under article IV, section 1(5), to which I return below.

Third. In this proceeding, it is premature to argue that PECBA leads to substantive results that are incompatible with the city’s freedom to choose and maintain its own political form. Possibly this could become an issue in the designation of bargaining units, see LaGrande/Astoria v. PERB, 281 Or at 156 n. 31, or in disputes over mandatory issues of bargaining, or over the permissible reach of an arbitration award. No such issues have arisen here. This does not mean that the present controversy is not ripe for adjudication. For the reasons stated by the Court, the validity of PECBA is properly at issue on review of the board’s order. However, the possibility that one or another hypothetical future arbitration awards may contain provisions impinging on a city’s reserved home rule powers does not invalidate the statute. If a specific provision of an award is attacked on this ground, it can be dealt with as an issue of statutory interpretation. The mere apprehension is *296no defense to the present charge that the city refused entirely to bargain under the statute.

Fourth. An alternative basis for the Court’s decision is that insofar as PECBA imposes state-mandated local procedures rather than state-generated substantive results, the act nevertheless is valid under the precedents recognizing the state’s authority to safeguard the interests of persons affected by governmental decisions in the procedures by which these decisions are reached. See LaGrande/Astoria v. PERB, and cases cited in 281 Or at 146 & note 15. This refers to “procedural protections,” see 284 Or at 182; the state cannot justify altering the processes of city government as a means to obtain the state’s preference of one rather than another substantive city action.

While the majority opinion does not make the distinction as clear as it might, I agree that PECBA can be sustained as a procedural safeguard for city employees. At least with respect to the conditions under which they work and the pay and other benefits they receive in return, the interest of city employees in the process by which these matters are decided is as appropriate for legislative protection as the interests of property taxpayers in assessments and appeals, ORS 309.090-.100, Boyle v. City of Bend, supra, interests of landowners and others in urban services and planning, see, e.g. ORS 215.402-.513, the interests of bond purchasers, creditors, and taxpayers in local bonding procedures, City of Cascade Locks v. Carlson, 161 Or 557, 90 P2d 787 (1939), see generally ORS ch 287; or the interests of suppliers and users in contracting and procurement procedures, ORS 279.015-.045. As mentioned in the preceding paragraph (“Third”), it is conceivable that improper decisions of the Employment Relations Board or of arbitrators appointed under the act may trench on reserved home rule authority, but this possibility does not make the act itself invalid or justify the city’s refusal to bargain under it until such an issue arises.

Fifth. There remains the referendum issue under article IV, section 1(5). That subsection provides:

“The initiative and referendum powers reserved to the people by subsections (2) and (3) of this section are further reserved to the qualified voters of each municipality and *297district as to all local, special and municipal legislation of every character in or for their municipality or district. The manner of exercising those powers shall be provided by general laws, but cities may provide the manner of exercising those powers as to their municipal legislation. In a city, not more than 15 percent of the qualified voters may be required to propose legislation by the initiative, and not more than 10 percent of the qualified voters may be required to order a referendum on legislation.”

Roseburg contends that, because its ordinance provides for referring the parties’ final offers to the city’s voters, the absence of such a provision in PECBA deprives the people of Roseburg of the powers reserved in section 1(5) and invalidates the statute.

This confuses two provisions of section 1(5). The subsection itself reserves to the people the referendum “as to all local, special and municipal legislation of every character in or for their municipality or district.” This is to be exercised in a manner provided by “general laws” and does not depend on inclusion in a charter or ordinance. What cities may provide for themselves is the “manner of exercising those powers as to their municipal legislation.” But if a city’s own “municipal legislation” is superseded by an otherwise valid state law, even a special or local law, this clause has no further application.

Of course it can be argued that section 1(5) as a whole shows that local voters were to be able to vote on all legislation that is special to their own locality. But the question whether the terms and conditions arrived at under PECBA are “local [or] special. . . legislation” subject to referendum under general laws is not now before us. As previously mentioned, this question depends in part on the characterization of those terms and conditions as a state-imposed substantive rule. If it were to arise, it would be directly under article IV, section 1(5), independent of any provision in a city charter or ordinance.5

*298Summary. It is not surprising that state legislation meets the strongest “home rule” objections and causes the most difficult legal issues when it deals with municipal employment practices, as was the case in LaGrande/ Astoria and earlier in Heinig6 and in Branch v. Albee.7

The reason is not only that state-mandated labor standards commit local tax money; that can be true of state standards for other municipal functions. More to the point, state regulation of municipal employment practices deals with the city government’s personnel, that is to say, with the individuals and the institutional lines of authority by which “home rule” is carried out. Nonetheless, state-mandated collective bargaining laws for public employees, including interest arbitration at least for those employees who are forbidden to strike, have been enacted in many *299states since 1969. See McAvoy, Binding Arbitration of Contract Terms: A New Approach to the Resolution of Disputes in the Public Sector, 72 Col L Rev 1192 (1972). There are many social interests of city employees, such as wage and hour standards, safe and healthy working conditions, insurance and retirement benefits which state law may protect without threatening home rule with respect to the city policies and programs that city personnel are employed to carry out.

Accordingly, state laws governing municipal personnel must be designed and administered with sensitivity toward such inexact distinctions as those between employees and policymaking officials, and between working conditions and program content. As stated above, such distinctions could become “home rule” issues in the designation of bargaining units, in the specification of mandatory subjects of bargaining, or in an arbitration award. But they are not before us in the present proceeding, and we decide nothing about them.

All that was before the Court of Appeals for review was the board’s determination that the city committed an unfair labor practice in refusing to bargain under the state’s act on the ground that the act contravened the “home rule” amendments. As stated above, PECBA encompasses some state substantive policies (prohibition of strikes), some procedures administered by the state itself, some local procedures mandated to provide procedural protection for employees who are denied the right to strike (collective bargaining), and only in the last resort the settlement by arbitration that may be characterized as a state-mandated local procedure (when the parties agree on their own arbitrators) or arguably as a state-mandated substantive result (when ERB selects an arbitrator). The act therefore cannot be circumvented altogether by refusing to bargain under it, in the absence of some legal challenge to the designation of the bargaining unit or to some other specific order. The Court of Appeals did not err in affirming ERB’s order.

This is judicial review of an order of the Employment Relations Board (ERB) finding the City of Roseburg and its city manager (Roseburg), a public employer, to have engaged in an unfair labor practice contrary to ORS 243.672(l)(e) by its refusal to bargain collectively with the exclusive representative of its employee firemen (Local 1489) and ordering Roseburg to cease and desist from its refusal. Roseburg sought judicial review and the Court of Appeals affirmed, citing our decision in LaGrande/Astoria v. PERB, 281 Or 137, 576 P2d 1204, aff’d on rehearing 284 Or 173, 586 P2d 765 (1978), and its own decision in Medford Firefighters Assn. v. City of Medford, 40 Or App 519, 595 P2d 1268, rev den 287 Or 507 (1979). We allowed review to examine the applicability of Article XI, section 2, and Article IV, section 1(5), of the Oregon Constitution, the so-called “home rule” amendments, to conflicting state and municipal legislation regulating public employment collective bargaining. Those amendments provide:

“* * * The Legislative Assembly shall not enact, amend or repeal any charter or act of incorporation for any municipality, city or town. The legal voters of every city and town are hereby granted power to enact and amend their municipal charter, subject to the Constitution and criminal laws of the State of Oregon * * Or Const, Art XI, § 2.

“The initiative and referendum powers reserved to the people by subsections (2) and (3) of this section are further reserved to the qualified voters of each municipality and district as to all local, special and municipal legislation of every character in or for their municipality or district. The manner of exercising those powers shall be provided by general laws, but cities may provide the manner of exercising those powers as to their municipal legislation. * * *” Or Const, Art IV, § 1(5).

By enactment of ORS 243.650 through 243.782, commonly referred to as the Public Employe Collective Bargaining Act (PECBA), the legislature has provided a comprehensive statutory scheme authorizing and regulating collective bargaining between municipal and other public employers and employees, administered by ERB. Of particular significance in this case, firemen and certain other public safety employees are forbidden from striking, *269ORS 243.736, and issues which are unresolved by bargaining and mediation are subject to compulsory binding arbitration administered by ERB. ORS 243.742.

Roseburg adopted its Ordinance No. 2074 which also purports to regulate its municipal employment relations. It is generally similar to the state’s PECBA, but there are differences both before and after impasse. Rather than set out both PECBA and the ordinance entirely, we quote the summary description of the differences as stated in ERB’s order:

“* * * This ordinance purports to ‘incorporate and follow’ the provisions of the PECBA, ‘with the exception of those areas which are believed to be strictly of local concern for the City of Roseburg.’

“In its pre-impasse collective bargaining provisions, the ordinance does bear a general resemblance to corresponding pre-impasse provisions of the PECBA. In fact, many of the ordinance’s pre-impasse provisions are the same as their PECBA counterparts, e.g.: the definitions of ‘collective bargaining,’ ‘appropriate unit,’ ‘confidential employee,’ ‘supervisory employee,’ ‘labor organization,’ ‘mediation,’ ‘public employee,’ and ‘strike,’ the section ‘Rights of City Employees,’ and the section ‘Certification and Recognition.’ There are, however, some notable differences, pre-impasse, between the ordinance and the PECBA, including a retained management rights section, a provision that provides the City with the discretion ‘to voluntarily confer with City employees in the process of developing policies to effectuate or implement’ retained management rights, a partially different unfair labor practice section, and a section which sets some specific times for the commencement and conclusion of negotiations.

“With the exception of mediation, the post-impasse provisions of the ordinance differ significantly from their PECBA counterparts. The ordinance’s factfinding procedure dictates the use of a three-member Board of Factfinders to select ‘the most reasonable’ of the final offers presented by the parties. In making its ‘final offer’ recommendation, the Board of Factfinders must consider factors which differ from those to be considered by factfinders operating under the Rules of this Board. Although the Board of Factfinders’ recommendation is not binding on the ‘loser,’ the ‘loser’ will be deemed to have accepted the recommendation if it has not formally rejected it within five days after its issuance.

*270“If the recommendation of the Board of Factfinders is rejected, the two (2) ‘final offers’ are submitted to the City residents whose vote is final and binding upon the parties. Firefighters are prohibited from striking. Enforcement of the ordinance is obtained by resort to suits in equity brought in Circuit Court or actions filed in the Municipal Court of the City of Roseburg.”

From the inception of the bargaining process and consistently throughout it, Roseburg insisted that it was bargaining “under” its ordinance rather than under PECBA. The parties did not reach agreement and mediation was unsuccessful. Local 1489 requested initiation of factfinding by ERB under ORS 243.722. Roseburg instead initiated factfinding under its ordinances and requested ERB to defer to the municipal procedures. At that point, Local 1489 initiated this proceeding, complaining that Roseburg’s refusal to bargain in good faith under PECBA was an unfair labor practice.

ERB reasoned that state law was controlling, that Roseburg was subject to its requirements, that the ordinance affected the bargaining process, and that Roseburg’s insistence on bargaining under the ordinance instead of the statute was a refusal to bargain collectively in good faith.

I. JUDICIAL REVIEW

The most substantial difference between PECBA and the ordinance relates to post-impasse procedures: PECBA requires arbitration; the ordinance requires submission of the last offers of each party to the voters of Roseburg. This proceeding was brought before post-impasse procedures were activated. Therefore, the threshold question is whether the post-impasse difference in the state and municipal legislation has yet ripened into a material issue.

Roseburg contended before ERB that the post-impasse difference was not material because the collective bargaining obligations of the city under the pre-impasse procedures of the ordinance and of the statute were essentially similar. ERB rejected that contention. It found that the differences as to both pre- and post-impasse procedures affected Roseburg’s conduct during bargaining. That determination is subject to review under ORS chapter 183, the *271Administrative Procedure Act. ORS 183.482(7) requires that

«* * * court shall not substitute its judgment for that of the agency as to any issue of fact or agency discretion. * *

Thus we review for the legal correctness of any factfinding, not for whether we agree with it. This is usually expressed in terms of substantial evidence, i.e., we uphold any finding supported by substantial evidence. ORS 183.482(8)(c) provides:

“The court shall set aside or remand the order if it finds that the order is not supported by substantial evidence in the record.”

The substantial evidence rule is not entirely dis-positive in reviewing findings which embody inferences. An inference has two parts: a primary fact plus a deduction. The evidence directly establishes only the truth of the primary fact or facts from which an inference may be derived therefrom. Rational bases may exist for more than one inference to be drawn from the same primary fact, and the factfinder (i.e., the agency) has the task to decide which one to draw. The court does not substitute its judgment as to which inference should be drawn, but it must review for the existence of a rationale. The rationale is reviewed for soundness, not for conformity to judicial preference. Judicial review of an inference is thus in two stages: (1) whether the basic fact or facts are supported by substantial evidence, and (2) whether there is a basis in reason connecting the inference to the facts from which it is derived. It is a twofold review for substantial evidence and, in a sense, for “substantial reason,” cf. Springfield Education Assn. v. School Dist., 290 Or 217, 228, 621 P2d 547 (1980), and McCann v. OLCC, 27 Or App 487, 495, 556 P2d 973 (1976), rev den 277 Or 99 (1977).

On judicial review, the court will not substitute its judgment for that of the agency in drawing an inference, but the court must be satisfied that agency judgment has actually been exercised. Sometimes a rational nexus between an evidenced fact and an inference drawn from it is obvious from common experience (e.g., we may infer from the fact of a wet street that it recently rained). In other *272cases, however, and particularly in cases involving expertise, the reasoning is not obvious (e.g., we may infer from present meteorological conditions that it will snow tomorrow). In such an inference, we will not assume the existence of a rationale. Rather, we look to the order to state the rational basis of the agency’s inference.1 The explanation need not be complex, but it should be sufficient to demonstrate the existence of a rational basis and to allow for judicial review. Cf. Home Plate, Inc. v. OLCC, 20 Or App 188, 530 P2d 862 (1975).

Here, the agency inferred that the differences between the statute and the ordinance affected Roseburg’s pre-impasse performance of its obligation to bargain in good faith under the statute. The findings and reasoning which underlie that inference are expressed in the order:

“The City’s second contention falls short on two (2) accounts. In the first place, even good faith bargaining is directly affected by the procedures and steps which are superimposed on the negotiations process. For example, the parties’ knowledge that, should there be a rejection of the Board of Factfinders’ final offer recommendation, the contract dispute would be settled by the local electorate, would have a direct impact on the give-and-take process of negotiations. A different impact on negotiations would be felt if the parties were aware that they would have all outstanding impasse issues resolved by an impartial arbitrator. Any labor or management negotiator would readily perceive the impact of the governing procedure on the bargaining process and would develop his or her bargaining strategy accordingly. The bargaining process, pre-impasse, is directly and legitimately affected by the entire set of governing procedures.

“Secondly, the City’s ripeness contention is faulty in view of the fact that there are points of conflict between the pre-impasse ordinance provisions and corresponding PECBA pre-impasse provisions. For example, the City’s ordinance contains a retained management rights section; there is no such provision in the PECBA. The ordinance *273gives the City the discretion to voluntarily confer with City employes in the process of developing policies to effectuate or implement the retained management rights. There is no similar provision in the PECBA. The unfair labor practice section of the ordinance differs to some degree from corresponding provisions of the PECBA, thus arguably establishing different standards of lawful conduct under the ordinance than under the PECBA. Additionally, the ordinance establishes different time line obligations than the PECBA.”

The reasoning as to post-impasse procedures in the first paragraph is obvious. The difference between the statute and the ordinance in the method of final resolution is almost polar. The PECBA is intended to provide disinterested arbitration by arbitrators chosen by the parties or by “qualified, disinterested persons” nominated by ERB and not stricken by the parties. ORS 243.746(1) and (2). The ordinance provides for selection of the last offer of one of the parties by persons with a direct financial and service-benefit interest in the decision, i.e., the voters of Roseburg. The analogy in the private sector would be the difference between an employer submitting an unresolved wage dispute to neutral arbitration or referring its last offer to its own stockholders. It is obvious that the difference in final resolution procedures would affect bargaining because the employer would be less reluctant to disagree in the latter case. ERB’s reasoning in this regard is sound and we uphold it.

The primary facts found in the second paragraph regarding pre-impasse differences are supported by evidence of the ordinance itself. However, in the absence of a finding that the subjects upon which the statute and ordinance differed were subjects of bargaining or some other finding of similar effect, the inferential connection between those differences and Roseburg’s bargaining is not evident. Nor does the order express a reason by which ERB derived that inference from the evidenced facts. Because there is neither an evident nor a stated rational basis for the inferential finding that the pre-impasse differences in the ordinance and the statute affected Roseburg’s bargaining, we do not uphold it.

*274 Where two bases for an inference are expressed, either of which is sufficient, and it is expressed or implied in the order that the agency would so infer on either basis, the inference is supported and we will uphold it on review. Here, the first basis, that dealing with the difference in post-impasse conflict resolution by arbitration or plebiscite, is sufficient standing alone to support the inference that the existence of the ordinance affected Roseburg’s bargaining. It is implicit in the order that ERB regarded either basis to be sufficient to support its inference. Thus we uphold the inferential finding and we turn to the merits.

II. STATE PRE-EMINENCE IN SUBSTANTIVE LEGISLATION

The parties have properly framed their arguments under the principles expressed by this court in LaGrande/Astoria v. PERB, supra. There, this court held in essence that the home rule amendments grant pre-eminence to local governments in matters of local political organization and that the legislature remains pre-eminent in matters of substantive law. Accordingly, we sustained the constitutionality of a statute requiring municipal firemen and police officers to be brought within the statewide, state-administered Public Employees Retirement System unless the municipal employer provided them with an equal or better retirement program.

It is undisputed that it is within both the plenary authority of the state and the charter or statutory authority of Roseburg to enact legislation regulating public employment relations of local government. Roseburg agrees that PECBA, standing alone, would be constitutional even under the home rule amendments. The problem in this case is not posed by the existence of overlapping authority. Rather, as Roseburg frames the argument, this case involves a conflict in the exercise of overlapping authority and the issue is one of predominance. Roseburg acknowledges, as we held in LaGrande/Astoria, that state legislation controls unless it is limited by the constitution, but contends that under the home rule amendments, the municipal legislation predominates. But see, LaGrande/Astoria v. PERB, 281 Or at 151, n 21.

*275We summarized the purpose of the home rule amendments in LaGrande/Astoria v. PERB:

«* * * prjhese constitutional provisions are concerned with the structural and organizational arrangements for the exercise of local self-government, with the power of local voters to enact and amend their own municipal charters and to employ the initiative and referendum for ‘local, special [or] municipal legislation.’ They address the manner in which governmental power is granted and exercised, not the concrete uses to which it is put. * * *” 281 Or at 142-43.

We concluded that for purposes of home rule analysis, there are two types of statutes: those which regulate the organization of local government and those which deal with substantive state policy and affect local government. Each type has certain constitutional prerequisites. We described the first type of statute and the applicable constitutional limitations as follows:

“When a statute is addressed to a concern of the state with the structure and procedures of local agencies, the statute impinges on the powers reserved by the amendments to the citizens of local communities. Such a state concern must be justified by a need to safeguard the interests of persons or entities affected by the procedures of local government.” 281 Or at 156.

This statement is not applicable to this case because PECBA is not addressed to “the structure and procedures of local agencies.” The dissent contends that these are the dispositive principles from LaGrande/Astoria because PECBA requires procedures of local government. We will discuss that contention below.

We also stated the principles by which the validity of state legislation regarding substantive objectives is to be determined when it conflicts with local legislation addressed to similar objectives:

“Conversely, a general law addressed primarily to substantive social, economic, or other regulatory objectives of the state prevails over contrary policies preferred by some local governments if it is clearly intended to do so, unless the law is shown to be irreconcilable with the local community’s freedom to choose its own political form. In that case, such a state law must yield in those particulars *276necessary to preserve that freedom of local organization.” (Footnote omitted.) 281 Or at 156.

The ideas in this paragraph control this case. The dominant character of PECBA is that of “a general law addressed primarily to social, economic and other regulatory objectives of the state.” As the public employment retirement statutes which we upheld in LaGrande/Astoria v. PERB, supra, sought to protect the economic welfare of public employees after retirement, PECBA is intended to protect their economic welfare during their employment and to provide a means for them to affect certain negotiable working conditions. Another policy served by PECBA is to protect public safety by the substantive device of prohibiting strikes of public safety employees. The substantive goal of that ban is prevention of interruption in the provision of essential government services. The class of persons to be benefited by this policy extends beyond the population of any city. PECBA is expressly premised upon a legislative determination that the people of the state have an interest in public employment relations at both the state and local levels of government. ORS 243.656, a policy statement for the Act, states:

“The Legislative Assembly finds and declares that:

“(1) The people of this state have a fundamental interest in the development of harmonious and cooperative relationships between government and its employes;

U* * * * *

“(4) The state has a basic obligation to protect the public by attempting to assure the orderly and uninterrupted operations and functions of government * * *.”

As in LaGrande/Astoria v. PERB, supra, the people and interests affected by the maintenance of municipal public safety services are not necessarily defined by reference to city boundaries:

“* * * [C]ity police officers and firemen are sometimes assigned duties beyond their cities, but this is hardly needed to demonstrate a state concern. Large complexes of state buildings and state personnel such as college campuses, and indeed the state Capitol, executive offices, and this court, depend on the quality of police and fire protection within city limits, and thousands of persons who frequent city streets and business districts every day are not city residents. * * *” 281 Or at 154.

*277The method chosen for achieving these policies of protecting the interests of both the employees and of the public is the promotion of harmonious relations between labor and management in the public sector by imposition of a state-administered system of collective bargaining resembling in part that which regulates private employment. One difference between public and private employment regulation is in the treatment of employees engaged in providing public safety services. Instead of allowing the economics of a free labor market or of labor-managment confrontation to control, as it might have, the legislature has provided arbitration as a quid pro quo for the prohibition of strikes by firemen and other public safety employees. Indeed, a city may not choose to allow a strike to occur as a labor-management tactic. Together, these statutes are intended to assure the public within and without each municipality of a continuity of certain essential governmental services while also recognizing the obligation of government and its employees to engage in collective bargaining. Clearly, these social and economic objectives are the primary purpose of PECBA. Local governments may prefer other policies or may prefer other means of achieving these policies, but the legislature, by enacting PECBA, has made that substantive decision as a matter of state law. As substantive legislation, state law prevails unless it unlawfully interferes with the structure of local government.

PECBA does not deal with the structure of local government. Understanding of the distinction between legislated organization of local government and legislated organization of state government which affects local government may be aided by consideration of these hypothetical statutes:

1. A statute requires that cities provide workers’ compensation for all municipal employees and establish municipal workers’ compensation boards to administer the programs, adopt compensation schedules and adjudicate unsettled claims.

2. A statute requires that cities and city employees are covered by the state workers’ compensation program and all unsettled claims are to be adjudicated by the state Workers’ Compensation Board under compensation schedules adopted by it.

*278Both examples involve governmental structure. Both involve decisional procedures in which the city must participate. Both involve decisions which affect the conduct of municipal business. They are, however, different. The first involves structure (a mandated municipal agency) and procedures (adoption of compensation schedules and adjudication of claims) of local government. The latter involves structure and procedures of state government. A decision under the first statute would be an exercise of municipal authority by a municipal agency. Under LaGrande/Astoria, the statute would therefore be reviewed as one which is “addressed to a concern of the state with the structure and procedures of local agencies.” A decision under the second statute would be an exercise of state power by a state agency. That statute would be examined under LaGrande/Astoria as a substantive statute for irreconcilability “with the local community’s freedom to choose its own political form.” The dissent, although purporting to argue consistently with LaGrande/Astoria, fails to grasp that distinction, as if all procedural statutes were the same. Hence, much of the reasoning of the dissent simply is not in point.

PECBA was enacted for social and economic objectives. Its resort to arbitration as a device to achieve those objectives is not “irreconcilable with the local community’s freedom to choose its own political form.” 281 Or at 156. If the state legislation had required local government to resolve its labor disputes by means of specified local governmental agencies or procedures such as a Municipal Employment Relations Board, then PECBA would be open to challenge under the precedent of State ex rel Heinig v. Milwaukie, et al, 231 Or 473, 373 P2d 680 (1962). There, a statute required local governments to establish civil service commissions to control municipal employment relations of firemen. We held that because municipal interest in the organization of fire suppression was greater than the state’s, the statute offended the principles of home rule. Like the Heinig legislation, PECBA establishes state policy. Unlike the Heinig statute, however, PECBA requires no local agencies. Rather, it assigns political responsibility for administration and decision making to an agency of the state.

*279An ultimate decision under PECBA is made by a process of arbitration provided and conducted by ERB, an agency of the state. Arbitration is afforded by the state as part of a comprehensive statutory scheme by which ERB oversees and administers the entire process of public employment collective bargaining from initial certification through arbitration to contract compliance and unfair labor practice resolution. The statute does not specify a line of authority for the arbitrator, but assumes that the arbitrator works on behalf of the state in the implementation of state policy. Indeed, that assumption is so clearly implicit in the statutes that Roseburg, in its petition, complains of the statutory allocation of “decisional power to a state arbitrator to decide employee questions in a manner contrary to the power allocated in the charter to the city manager and council.” (Our emphasis.) Roseburg has phrased it correctly: the conflict is between a statute assigning ultimate authority to a state agency and a charter which assigns it to local government.

According to the policy statement of ORS 243.742(1), the state is the governmental entity which affords arbitration:

“It is the public policy of the State of Oregon that where the right of employes to strike is by law prohibited, it is requisite to the high morale of such employes and the efficient operation of such departments to afford an alternate, expeditious, effective and binding procedure for the resolution of labor disputes * * *.”

The parties may contract for private arbitration of disputes, ORS 243.706 and 243.762, but if they do not, then after mediation and factfinding (also state-administered, ORS 243.712 and 243.722), the parties “shall petition the board to initiate binding arbitration” or the board may do so on its own motion. ORS 243.742(2). In either event, ERB is the initiating party. The parties may agree on an arbitrator. If not, arbitrators are nominated by ERB and the parties may strike certain of them. The remaining ERB nominee is designated as the arbitrator. ORS 243.746.

Arguably, the arbitrator is an independent decision maker in the sense that a coin tossed to decide a dispute is neutral or unaccountable. Clearly, however, it is *280the state’s coin and the state’s toss. State law specifies the factors upon which the decision must be based. ORS 243.746(4) provides:

“Where there is no agreement between the parties, or where there is an agreement but the parties have begun negotiations or discussions looking to a new agreement or amendment of the existing agreement, and wage rates or other conditions of employment under the proposed new or amended agreement are in dispute, the arbitration panel shall base its findings, opinions and order upon the following factors, as applicable:

“(a) The lawful authority of the employer.

“(b) Stipulations of the parties.

“(c) The interest and welfare of the public and the financial ability of the unit of government to meet those costs.

“(d) Comparison of the wages, hours and conditions of employment of other employes performing similar services and with other employes generally:

“(A) In public employment in comparable communities.

“(B) In private employment in comparable communities.

“(e) The average consumer prices for goods and services commonly known as the cost of living.

“(f) The overall compensation presently received by the employes, including direct wage compensation, vacations, holidays and other excused time, insurance and pensions, medical and hospitalization benefits, the continuity and stability of employment, and all other benefits received.

“(g) Changes in any of the foregoing circumstances during the pendency of the arbitration proceedings.

“(h) Such other factors, not confined to the foregoing, which are normally or traditionally taken into consideration in the determination of wages, hours and conditions of employment through voluntary collective bargaining, mediation, factfinding, arbitration or otherwise between the parties, in the public service or in private service.”

The arbitrator’s decision is subject to judicial review for conformity to the requirements of state law on the petition of ERB or the parties, ORS 243.752. The members of ERB are appointed by the governor, ORS 240.065. If ERB and *281ERB’s assigned arbitrators do their job unsatisfactorily, the governor is politically accountable to the electorate.

PECBA makes no organic change to the political form of local government. The Common Council, budget committee, and executive agencies of Roseburg and the allocation of responsibility among them are unaffected by the statute. On its face, PECBA does not subject any matters of local governmental organization to negotiation, cf., LaGrande/Astoria v. PERB, 281 Or at 156, n 31. It provides generally for collective bargaining only regarding “employment relations.” That term is defined at ORS 243.650 and the definition refers to no issues of local government organization.

“(7) ‘Employment relations’ includes, but is not limited to, matters concerning direct or indirect monetary benefits, hours, vacations, sick leave, grievance procedures and other conditions of employment.”

In summary, the authority of a city to make a particular decision of state concern is transferred to the state, but the political forms and procedures by which a city decides matters which remain within its legislative purview are unchanged. PECBA provides for certain decision-making arrangements at the state level as an implementive device for its primary substantive objectives of the state. Although the assumption by the state of decision-making authority affects the affairs of local government in certain ways, it does not mandate structural and organizational arrangements of local government contrary to Article XI, section 2. Therefore, we conclude, in the language of LaGrande/Astoria, that PECBA is a general law addressed primarily to substantive social, economic and other objectives of the state which does not affect the local community’s freedom to choose its own political form.

The dissent is a strong and able presentation of views which have much historical force. However, the dissent analyzes the application of the home rule amendments to PECBA in a way which reflects a misconception of our holding in LaGrande/Astoria. The dissent argues laboriously to prove a point which is not disputed: arbitration is a procedure. The first quoted statement from LaGrande/Astoria, upon which the dissent relies, does *282not refer to all statutes involving procedures, but only to those involving “procedures of local agencies.” For convenience, we repeat the quotation:

“When a statute is addressed to a concern of the state with the structure and procedures of local agencies, the statute impinges on the powers reserved by the amendments to the citizens of local communities. Such a state concern must be justified by a need to safeguard the interests of persons or entities affected by the procedures of local government.” (Emphasis added.) 281 Or at 156.

This case involves structure and procedures of state government, not local government. We elaborated on this point in our discussion above wherein we concluded that PECBA is not irreconcilable with the local community’s freedom to choose its political form. Hence, the quotation upon which the dissent relies is not dispositive.

Nevertheless, it is arguable that compelled collective bargaining, culminating in arbitration, can be regarded as a decisional procedure in which local government is required to participate and hence should be considered a procedure of local government. Were we to accept that contention and apply the quoted holding of LaGrande/Astoria, our conclusion would be the same. The next inquiry to make under LaGrande/Astoria would be whether the state concern is “justified by a need to safeguard the interests of persons or entities affected by the procedures of state government.” Clearly it is.

The interest of employees in organization for the purpose of collectively affecting their working conditions and compensation has long been a major force in American economic, social and political history. At least since the Great Depression and the New Deal, the general recognition of the interest of laboring people in effective collective activity has been reflected by law which protects that interest to various extent. Not to belabor the obvious, it is sufficient to say that the interest of working people in organizing to act collectively, including by collectively refusing to work as an economic tactic, has become a fact of American life. The approach of modern labor legislation has been to recognize that interest subject to exceptions in the public interest, and to regulate it with the objective of *283avoiding labor strife. In 1933, the Oregon legislature recognized the interest of laboring people in organizing for their common welfare. It did so in terms which left no question as to the importance of the interest and, although statutes regulating labor disputes have been modified over the years, the declaration of public policy in ORS 662.020 remains unchanged since its adoption in 1933:

“* * * [T]he public policy of Oregon is declared as follows: Whereas under prevailing economic conditions, developed with the aid of governmental authority for owners of property to organize in a corporate and other forms of ownership association, the individual unorganized worker is commonly helpless to exercise actual liberty of contract and to protect his freedom of labor and thereby to obtain acceptable terms and conditions of employment, wherefor, though he should be free to decline to associate with his fellows, it is necessary that he have full freedom of association, self-organization and designation of representatives of his own choosing to negotiate the terms and conditions of his employment and that he shall be free from the interference, restraint or coercion of employers of labor, or their agents, in the designation of such representatives or in self-organization or in other concerted activities for the purpose of collective bargaining or other mutual aid or protection * * *.” See Or Laws 1933, ch 355; cf. National Labor Relations Act, 29 USC § 151.

Whether or how to safeguard in public employment the interests of labor which are protected in private employment is a political question for the legislature and PECBA is the legislative determination of that issue in Oregon. The PECBA begins with a policy statement (parts of which were set out above) which recognizes that this interest of the persons and entities affected by PECBA in collective action exists in public as well as private employment, but that there are important public interests as well. ORS 243.656(2), (3) and (5) expresses this policy:

“(2) Recognition by public employers of the right of public employes to organize and full acceptance of the principle and procedure of collective negotiation between public employers and public employe organizations can alleviate various forms of strife and unrest. Experience in the private and public sectors of our economy has proved that unresolved disputes in the public service are injurious *284to the public, the governmental agencies, and public employes;

“(3) Experience in private and public employment has also proved that protection by law of the right of employes to organize and negotiate collectively safeguards employes and the public from injury, impairment and interruptions of necessary services, and removes certain recognized sources of strife and unrest, by encouraging practices fundamental to the peaceful adjustment of disputes arising out of differences as to wages, hours, terms and other working conditions, and by establishing greater equality of bargaining power between public employers and public employes;

* * * sj: *

“(5) It is the purpose of ORS 243.650 to 243.782 to obligate public employers, public employes and their representatives to enter into collective negotiations with willingness to resolve grievances and disputes relating to employment relations and to enter into written and signed contracts evidencing agreements resulting from such negotiations. It is also the purpose of ORS 243.650 to 243.782 to promote the improvement of employer-employe relations within the various public employers by providing a uniform basis for recognizing the right of public employes to join organizations of their own choice, and to be represented by such organizations in their employment relations with public employers.”

These legislative statements of the policy of the PECBA express a legislative determination, in the words of LaGrande/Astoria, of the existence of “a need to safeguard the interests” of public employees in procedures which affect them. PECBA restricts those interests by denying to specified public safety employees the right to strike, regardless of whether their local employers would choose to deny that right. It then safeguards the interests of those employees by providing a substitute protection in the form of binding arbitration. We are not able to say that there is no reasonable basis for legislative action.2

*285Therefore, even if the dissent were correct that PECBA should be deemed legislation addressed to the procedures of local government, it meets the LaGrande/Astoria test for validity of such legislation because it is justified by a need to safeguard the interests of affected persons in those processes.

III. MUNICIPAL REFERENDUM POWERS

Roseburg also contends that the authority of the city to allow its voters to arbitrate unresolved labor disputes immunizes the process from state regulation because it is an exercise of referendum powers reserved to the voters of local government under Article IV, section 1(5), supra. The framework for examination of this contention was also set in LaGrande/Astoria and, particularly, in our opinion on rehearing. There, we clarified that the amendments granted charter authority (i.e., organic legislation, whether by charter or ordinance) to municipalities and withdrew it from the legislature. They also empowered municipalities to legislate by vote of the people. We explained:

“* * * Together, the 1906 amendments provide two grants of power and one limitation of power. Article XI, section 2, grants power to the voters of every city or town to enact and amend their municipal charter. It withdraws power from the Legislative Assembly to enact, amend, or *286repeal such charters. Article IV, section 1(5), empowers local voters to initiate or to refer to popular vote ‘all local, special and municipal legislation.’

“It deserves to be reemphasized that the terms of the granted powers and of the accompanying limitation need not be and are not symmetrical. Much of the argument against these statutes has proceeded as though a constitutional grant of power to one level of government necessarily carries with it a corresponding withdrawal of power from the other. That this is not so has long been a truism with respect to the relationship between the powers of Congress and the states, and it is equally true of ‘home rule’ within a state. It is entirely possible to grant certain powers to local governments to act on their own initiative without at the same time limiting the powers of the state legislature. Indeed, as a practical matter this is essential if local government is to have any authority to legislate on its own in matters in which the state could also act, for otherwise local powers would have to be narrowly confined in order to save room for potential state legislation. The fact that the 1906 amendments gave municipal voters direct constitutional power to ‘enact and amend their municipal charter’ and to use the initiative and referendum for ‘all local, special and municipal legislation’ was a great achievement for home rule even though these two clauses did not of themselves take anything from the plenary legislative power of the state, for before 1906 these local powers had to be obtained from the Legislative Assembly.

“The withdrawal of power from the legislature is found in the other clause of the 1906 amendments quoted above, that ‘[t]he Legislative Assembly shall not enact, amend, or repeal any charter or act of incorporation for any municipality, city or town.’ Or Const art XI, § 2. * * *” (Footnotes omitted.) 284 Or at 176-77.

Article XI, section 2, supra, enlarged the scope of permissible subject matter for local legislation without need for state authorizing legislation. Article IV, section 1(5), differs in that it authorized popular vote as a means of exercising local legislative power. The latter amendment changed neither the scope of local legislative authority, nor the balance of state and local pre-eminence regarding substantive and organic legislation. Stated otherwise, under Article IV, section 1(5), local government may legislate either by popular vote or by representative vote, but either *287way, the legislation must be within the scope of municipal legislative authority. Initiative and referendum are a sharing of legislative power between the people and their representatives, not a grant of additional legislative power to either. See, State ex rel Pierce v. Slusher, 119 Or 141, 146-47, 248 P 358 (1926); Zilesch et al v. Polk County et al, 107 Or 659, 668, 215 P 578 (1923); Allison v. Washington County, 24 Or App 571, 581, 548 P2d 188 (1976). Hence the constitutional limitation of municipal initiative and referendum powers to “municipal legislation.” In this respect, Article IV, section 1(5), is analogous to the reservation of state initiative and referendum powers in the preceding subsections of Article IV, section 1. Article XI, section 2, however, granted pre-eminence in organic legislation to local government, regardless of the mode of legislating, but did not disturb the pre-eminence of the state in substantive legislation relating to subjects of state concern. Thus, a question of dominance of conflicting municipal and state legislation turns upon the substance of the legislation and not upon the manner in which either is enacted.3

*288The substance of the legislation was discussed in Part II. The primary social and economic objectives of the legislation have been determined by the legislature to be state concerns even though, like many statutes, PECBA affects activities of local government. In the absence of a superseding statute, the city would have been free to legislate an entirely different scheme of employment relations, with or without collective bargaining and impasse resolution provisions. By virtue of PECBA, however, the decision (impasse arbitration) is now beyond the city’s choice. It is immaterial to the validity of the statute that the city council has decided by ordinance to refer to a plebiscite a future decision which is no longer the city’s to make. Were it otherwise, local government could cripple the ability of the state to legislate regarding any matter of state policy which affected local governments by the simple expedient of local referendum. The home rule amendments were not intended to have that drastic, general effect.

We therefore conclude that ERB’s finding that Roseburg had unfairly refused to bargain collectively pursuant to PECBA was lawful.

Affirmed.

specially concurring.

I concur in all parts of the majority opinion by Justice Tanzer except that I limit my concurrence in part II *300of the majority opinion to the second ground stated in the opinion; that is, that if the state statutory procedure for compulsory arbitration is regarded as requiring a decisional procedure in which the city must participate, it is “justified by a need to safeguard the interests of affected persons in those processes.”

The dissent argues there was no showing of a “need.” I am of the opinion that the existence of the “need” required to permit the state to impose a decisional procedure in which the city must participate need not be shown by the introduction of evidence. In this case it is apparent to me that there is a “need” to substitute compulsory arbitration, or something akin, for the forbidden right to strike to secure reasonable wages and working conditions.

It may be there will be cases involving statutes in which the need for a state-imposed procedure is not apparent and, despite a legislative policy statement of need, we may conclude in such cases that a “need” is not shown.

dissenting.

I dissent from the opinion by the “majority”1 because it does not properly apply the new “criteria” adopted by this court in LaGrande/Astoria v. PERB 2 for application in “home rule” cases. Instead, the “majority” has gone far beyond those “criteria” and has based its decision on new grounds which make meaningless the LaGrande/Astoria distinction between matters of “procedure” and those of “substance.” Of more importance, these new grounds, as adopted by the “majority” as the basis for its decision, also have the effect of largely destroying any remaining area in which the power of “home rule” granted to cities by the Oregon Constitution to adopt charter provisions and ordinances can still prevail over the *301power of the legislature to adopt conflicting state statutes. This far-reaching result is made clear by an examination of the grounds on which the opinion by the “majority” is based and which are as follows:

(1) When, as here, a state statute and a city ordinance prescribe conflicting “procedures,” the statute is not subject to the LaGrande/Astoria “criteria” for statutes addressed to “procedures” of local government, so as to require the showing of a “need” for the procedures provided by such a statute, as required by LaGrande/Astoria, if it serves any “social, economic or other regulatory objectives” of the state. Because all state statutes presumably serve some “substantive social, economic or other regulatory objectives” of the state, the result is to destroy “home rule” on matters relating to “procedures” of local government, as set forth in LaGrande/Astoria.

(2) Even when a statute is one addressed to “procedures” of local governments which, by charter or ordinance, have adopted conflicting “procedures,” the “need” required to justify such a statute can be established whenever the legislature says that there is such a “need.” In addition, the “need to safeguard the interests” of the persons affected is not limited by the “majority” to a need to provide “procedural protections” for such persons, as held in LaGrande/Astoria, but is extended to a need to safeguard or promote any interests of such persons which the legislature may deem to be in need of protection. The result is to confer upon the legislature complete supremacy over cities in all matters involving “procedures” of local government, as well as in all “substantive” matters.

(3) Despite the fact that Article IV, Section 1(5) of the Oregon Constitution was amended to confer upon the voters of cities the right of initiative and referendum as to “all local, special and municipal legislation of every character” and despite the fact that in Rose v. Port of Portland 3 this court recognized that this amendment was submitted to the voters for adoption with an expressed representation that it would “give the people power to control the salaries of county and district officers,” the “majority” holds that *302the state has the power by statute to prescribe salaries and wages for all city officers and employees.

At the outset of a discussion of the problems presented in this case, it is important to keep in mind both the text of the “home rule” provisions of the Oregon Constitution and the purpose of those provisions.

In State ex rel Heinig v. Milwaukee et al, an opinion by Justice O’Connell, this court unanimously held that:

“* * * [T]he constitutional amendments * * * vested in the cities a part and an exclusive part of the power to legislate free from control of the state legislative assembly.”4

Indeed, as stated in the dissent in LaGrande/Astoria I:

“It has been recognized by authorities on this important subject that one of the basic purposes of ‘home rule’ is to ‘stake out a limited area where local government could legislate for itself and to ‘carve out an area in which the municipality enjoys a measure of local autonomy free from legislative interference or control * * *’ ” 281 Or at 159. (Footnotes omitted) (Emphasis added)

As recently stated by a recognized authority on this subject:

“The modern trend goes toward liberal construction of such grants [of home rule] with respect to local affairs and in several jurisdictions it has been held that home rule cities should be given the largest possible measure of self-government.” (Citing cases)5

In Heinig this court, by a unanimous opinion, held that the test to be applied in determining whether a particular matter is one of “local” or “state” concern is “not whether the state or the city has an interest in the matter, for usually they both have, but whether the state’s interest or that of the city is paramount” (231 Or at 481), i.e., “whether it is substantial enough to predominate over the interest of the city” (231 Or at 484), and that “[e]ach case requires a weighing of the state’s interest against the interest of the municipality.” (231 Or at 488).

*303In LaGrande/Astoria, this court did not disagree with the objectives of the “home rule” amendments as stated in Heinig, but in a four-to-three decision abandoned the test as stated in Heinig, together with the 20 years of precedent upon which Heinig was based, although holding that the reasoning of Heinig was still applicable in cases involving matters of procedure. This court in LaGrande/Astoria II was also critical (at p. 177) of the cities for their failure “to point to precise words” in Article XI, Section 2, to sustain their position. As a substitute for the Heinig test of “predominant interest,” this court in LaGrande/Astoria adopted a test based upon a distinction between statutes relating to “structure and procedures” of local government and those involving matters of “substance.” In formulating that new test, this court said in LaGrande/Astoria that:

“Article XI, section 2, for instance, is addressed to the legislative assembly and to the cities, telling the legislature what it may not do and the voters of the several cities what they may do. Judicial interpretations of such a provision must strive to articulate these directives and avoid formulations that give no guidance to government and leave every policy dispute to judicial decision.” LaGrande/Astoria I at 147. (Emphasis added)

and that for this purpose this court must adopt “identifiable criteria, derived from * * * constitutional command.” (LaGrande/Astoria II at 185).

This court in LaGrande/Astoria then purported to find “precise words” in Article XI, Section 2, to provide “identifiable criteria derived from * * * constitutional command” as the basis for the distinction made by it between statutes relating to matters of “structure and procedures” of city government and those involving matters of “substance,” and also to provide the basis for the “criteria” adopted by the court for application in each of those two types of cases.

That new distinction and these new “criteria” were adopted by this court in LaGrande/Astoria despite the fact that on rehearing all parties, including not only the cities, but also the unions and the Attorney General, representing the State of Oregon, conceded that such a distinction is *304inconsistent with the legislative history of the “home rule” amendments.

This is the first case since its decision in LaGrande/Astoria in which this court is required to apply the new “criteria” adopted in that case for application in determining whether statutes enacted by the Oregon legislature which conflict with the provisions of the charter or ordinances of a city prevail over such provisions in view of the “home rule” provisions of the Oregon Constitution, Article XI, Section 2, and Article IV, Section 1(5).

The more specific question presented for decision in this case is whether provisions of a state statute (ORS 243.742), which provides for a procedure for compulsory arbitration in the event that an “impasse” is reached in collective bargaining between a city and its firemen, prevail over the provisions of an ordinance adopted pursuant to a city charter and which also provides for collective bargaining between the city and such employees, but provides a different procedure to be followed in the event of such an “impasse,” under which the two “final offers” by the parties shall be referred to the voters of the city, whose vote is final and binding.

By its decision in LaGrande/Astoria I, and under the “criteria” adopted as the basis for that decision, this court held that the following distinction must be made:

(1) “When a statute is addressed to a concern of the state with the structure and PROCEDURES of local agencies, the statute impinges on the powers reserved by the amendments to the citizens of local communities. Such a state concern MUST be justified by a NEED to safeguard the interests of persons or entities affected by the procedures of local government.”

but that

(2) “* * * a general law addressed primarily to substantive social, economic, or other regulatory objectives of the state prevails over contrary policies preferred by some local governments if it is clearly intended to do so, unless the law is shown to be irreconcilable with the local community’s freedom to choose its own political form. In that case, such a state law must yield in those particulars necessary to preserve that freedom of local organization.” 281 Or at 156 (Emphasis added)

*305In order to properly apply these “criteria” as stated by this court in LaGrande/Astoria, the following are the issues which must be resolved in order to support a proper decision in this case:

(1) Is this state statute, which provides both a procedure for collective bargaining and also a procedure for determination of wages for city employees in the event of an “impasse,” a statute which is “addressed to a concern of the state with the structure and PROCEDURES” of local government?

(2) If so, has there been the required showing of a “need to safeguard the interests of persons or entities affected” by such a procedure?

(3) If not, and if this matter is one “addressed primarily to substantive * * * objectives” of the state, is this state statute “irreconcilable with the local community’s freedom to choose its own political form”?

1. This state statute, which provides a PROCEDURE for collective bargaining and a PROCEDURE for “post-impasse” settlement of disputes over wages for firemen and policemen by compulsory arbitration is a statute “addressed to a concern of the state with the * * * PROCEDURES” for collective bargaining and for “post-impasse” settlement of such disputes as provided by city ordinances, such as the ordinance in this case.

The opinion by the “majority” would hold that this statute, which provides a “post-impasse” procedure for compulsory arbitration, prevails over the Roseburg ordinance, which provides a different “post-impasse” settlement procedure, is not a “statute * * * addressed to a concern of the state with the structure and procedures” of local government, so as to require the showing of a “need to safeguard the interests of * * * persons affected” by the procedure provided by the Roseburg ordinance (the “criteria” prescribed in LaGrande/Astoria for “procedural” statutes), for the following reasons:

(a) “The dominant character of PECBA is that of ‘a general law addressed primarily to social, economic and other regulatory objectives of the state.’ ” 292 Or at 275-76. (Emphasis added).

*306(b) “Local governments may prefer other policies or may prefer other means [i.e., other procedures] of achieving these policies, but the legislature, by enacting PECBA, has made that substantive decision as a matter of state law. As substantive legislation, state law prevails unless it unlawfully interferes with the structure of local government.” 292 Or at 277. (Emphasis added)

As stated by the “majority,” however, this state statute includes the following policy statement:

“It is the public policy of the State of Oregon that where the right of employes to strike is by law prohibited, it is requisite to the high morale of such employes and the efficient operation of such departments to afford an alternate, expeditious, effective and binding procedure for the resolution of labor disputes * * ORS 243.742(1). (Emphasis added)

Indeed, ORS 243.650(5) also provides that:

“ ‘Compulsory arbitration’ means the procedure whereby parties involved in a labor dispute are required by law to submit their differences to a third party for a final and binding decision.” (Emphasis added)

Thus, the statute in this case provides a procedure for collective bargaining for disputes involving wages of firemen and policemen and also provides a procedure for compulsory arbitration in the event of an “impasse” in collective bargaining. By statutory definition, this is a “procedure.” The Roseburg ordinance also provides a procedure for collective bargaining for disputes involving wages of firemen and policemen, but provides a different procedure for the settlement of such disputes in the event of an “impasse” in collective bargaining.6

*307It would thus appear to be obvious not only: (1) that this statute is a state statute which provides a procedure for collective bargaining in disputes over wages for firemen and policemen and a procedure for the compulsory arbitration of such disputes in the event of an “impasse” in collective bargaining, but also (2) that in providing such statutory procedures for the settlement of such disputes, this statute is “addressed to a concern of the state with the * * * procedures” for collective bargaining in such disputes as provided by ordinances such as the Roseburg ordinance and, in particular, with the procedures provided by such ordinances for the settlement of such disputes in the event of an “impasse” in collective bargaining.

Indeed, this “concern” of the state with the procedures provided by such ordinances as the Roseburg ordinance is recognized by the “majority” (although not in terms of “concern”) by its statement that the purpose of this statute is to “protect their [firemen and policemen] economic welfare during their employment and to provide a means [i.e., a procedure] for them to affect certain negotiable working conditions” and (by the providing of such a means, i.e., procedure) to protect public safety by preventing interruption in the provision of essential government services.

The critical flaw in the opinion by the “majority,” however, is the “giant jump” in its reasoning, which is to the effect that (1) because the state has such a “concern” with the procedures for collective bargaining and “post-impasse” settlement procedures such as those provided by the Roseburg ordinance, and (2) because the state believes that its “policies” to protect the interests of firemen and policemen are better served by the procedures for collective bargaining and compulsory arbitration as provided by this statute, than by the procedures provided by such ordinances, it therefore follows that (3) this statute is not one “addressed to a concern of the state with * * * procedures of local agencies” (so as to be invalid in the absence of the showing of a “need to safeguard the interests of the persons * * * affected”), but is somehow transformed into a statute “addressed primarily to substantive social, economic, or other regulatory objectives of the state” (so as to be valid *308unless the statute is shown to be “irreconcilable with the local community’s freedom to choose its own political form.”).

Presumably, all state statutes are addressed to some “substantive social, economic or other regulatory objectives of the state,” including purely procedural statutes. It does not necessarily follow, however, that because a state procedural statute is “addressed” to such an “objective,” the proper test to be applied in determining whether a procedure provided by such a statute prevails over a procedure provided by a city ordinance is the test as stated in LaGrande/Astoria for “substantive” statutes, a rigid test which places the burden on the city and is more favorable to the state, rather than the test stated in LaGrande/Astoria for “procedural” statutes, a test which requires a showing of “need” and is more favorable to cities.

If this were so, the distinction made in LaGrande/ Astoria between matters of “procedure” and those of “substance,” so as to give cities the benefit of some measure of protection from legislative interference in matters involving “structure and procedures” of local government by imposing the burden in such cases to show a “need to safeguard the interests of persons or entities affected by the procedures of local government” to justify such interference, would be rendered largely meaningless and illusory. Yet this is the essence of the reasoning by the “majority” in support of its strange conclusion that this statute is not one that is “addressed to a concern of the state with the structure and procedures of local agencies,” so as to be subject to the “criteria” as stated in LaGrande/Astoria for application in determining the validity of procedures provided by such a statute when in conflict with procedures provided by a city ordinance, as in this case.

In support of its holding that the PECBA statute is “substantive” rather than “procedural,” the “majority” would make a distinction between: (1) a statute which mandates a city procedure by a city agency, and (2) a statute which does not mandate a city procedure by a city agency, but provides for a mandatory procedure by a state agency which supersedes and invalidates procedures provided by city agencies established under city charter or *309ordinance. According to the “majority,” because the PECBA statute is of the second type it is a “substantive,” not a “procedural,” statute.

I disagree. Statutes of both types are statutes which are “addressed to a concern of the state with the structure and procedures of local agencies” because both have the purpose, as well as the effect, of superseding all existing and conflicting procedures established under city charters or ordinances. In my opinion, the distinction urged by the “majority” is one which could “fool only a lawyer,” to paraphrase a previous holding by this court.7 As also recognized by this court in LaGrande/Astoria I, the legislature cannot do “by indirection” what it cannot do directly. (281 Or at 151-52).

For all of these reasons, I am of the firm opinion that this statute is clearly a statute “addressed to a concern of the state with the * * * procedures” of local government and, therefore, that the question whether that statute prevails over the conflicting procedures provided by the Rose-burg ordinance must be determined by application of the “criteria” or test as stated by this court in LaGrande/ Astoria for application in such cases, and as next discussed, 1. e., whether the “state concern” has been “* * * justified by a need to safeguard the interests of persons or entities affected by the procedures of local government.” (281 Or at 156).

2. There has been no showing of a “NEED to safeguard the interests of persons” affected by the procedure provided by the Roseburg ordinance.

The “majority” says (292 Or at 282) that “[w]ere we to accept” the contention that this statute is one “addressed to a concern of the state with the structure and procedures of local agencies,” as contended by the city, so as to require the showing of a “need” for such a statute, “our conclusion would be the same” for the following reasons:

(a) “Whether or how to safeguard in public employment the interests of labor * * * is a political question for the legislature and PECBA is the legislative determina*310tion of that issue in Oregon.” 292 Or at 283. (Emphasis added)

(b) “PECBA restricts [the] interests [of public employees] by denying to specified public safety employees the right to strike, regardless of whether their local employers would choose to deny that right. It then safeguards the interests of those employees by providing a substitute protection in the form of binding arbitration. We are not able to say that there is no reasonable basis for [such] legislative action.” 292 Or at 284. (Emphasis added)

These contentions beg the question to be decided in this case. It may be that in the absence of the provisions of city ordinances which guarantee to city employees the right of collective bargaining and which also provide a procedure for settlement of disputes after an “impasse” in collective bargaining, as in this case, the state would have the power to enact this statute.

The question to be decided in this case, however, is whether, in view of the powers reserved to cities by Article XI, Section 2, such a statute supersedes provisions of city ordinances which, as in this case, also guarantee to city employees the right of collective bargaining and provide a procedure for settlement of disputes after an “impasse” in collective bargaining, and which also restrict their right to strike. Under the “criteria” adopted by this court in LaGrande/Astoria, such a statute “MUST be justified by a NEED to safeguard the interests of persons * * * affected by the procedures of local government,” i.e., a showing that despite the availability of the procedures provided by the city ordinance, there is nevertheless a “need” for the procedures provided by the statute to supersede the provisions of such an ordinance in order to “safeguard the interests of persons or entities affected by the procedures” provided by the ordinance.

In LaGrande/Astoria this court recognized that under the “criteria” adopted in that case for application in cases involving a conflict between procedures under city charters or ordinances and procedures under state statutes, as in this case, the question of the “need” for the procedure under the statute is a question to be decided by the court, *311just as under the rule of Heinig the question of “predominant interest” was a question to be decided by the court.8 In other words, under the LaGrande/Astoria test, the crucial question is not whether the legislature has asserted a need for the statute by recitals in the statute, but whether the “need” is sufficient, as determined by the court, to justify the statute’s intrusion into “the structure and procedures” of local government.

The “majority,” however, has abdicated this function to the legislature by its holding that in such a case the legislature, by its enactment of the statute, may conclusively determine the question of “need.” The result would destroy any area of “home rule” by cities in cases in which procedures adopted by city charter or ordinance conflict with procedures adopted by state statutes and confer upon the legislature complete supremacy over cities in the area of “procedures” of local government by permitting the legislature to preempt that entire area.

This is contrary to the holding by this court in LaGrande/Astoria that “procedures” of local government as provided by city charters or ordinances are to prevail over procedures provided by statute unless there has been a showing of a “need” for the statutory procedures in a case involving such a conflict. It is also contrary to the purpose *312of home rule amendments to “carve out an area in which the municipality enjoys a measure of local autonomy free from legislative interference or control.” In other words, if the legislature can preempt even the entire area relating to “procedures of local government,” there is little, if any, area left to the cities in which they can have “local autonomy free from legislative interference or control.”

In addition, the “majority” goes far beyond the “criteria” adopted by this court in LaGrande/Astoria, in which this court described the nature of the “interests” which the court found were intended to be protected for the purpose of determining the existence of such a “need.” Thus, the “majority” says (in footnote 2) that such “interests” are not limited to those of a “due process nature,” but extend to the safeguarding of “other interests, such as collective bargaining.”

In LaGrande/Astoria II this court held that:

“A city’s choice of its frame of government in its charter, and even beyond the charter as such, is not subject to general or statewide laws, except for procedural protections of the kind cited at note 15 of the original opinion. With respect to general laws for governmental processes, our opinion reaffirmed the rule of State ex rel Heinig v. City of Milwaukie, 231 Or 473, 373 P2d 680 (1962). See 281 Or at 146, 576 P2d at 1210.” 284 Or at 182.9 (Emphasis added)

As noted by the city in this case, the cases cited in footnote 15 dealt with specific procedural protections for individuals in a particular governmental decision-making process: Brown v. City of Salem, 251 Or 150, 444 P2d 936 (1968) (notice of proposed street assessment under ORS 223.389 and 223.399); Bennet v. City of Oceanlake, 247 Or 539, 430 P2d 1004 (1967) (notice of public improvement to affected property owners under ORS 223.389); City of *313Woodburn v. State Tax Comm’n., 243 Or 633, 413 P2d 606 (1966) (notice to voters in state law requiring tax levy measure to be expressed in dollar as opposed to mill amounts); Boyle v. City of Bend, 234 Or 91, 380 P2d 625 (1963) (court appeal from assessment decision); City of Cascade Locks v. Carlson, 161 Or 557, 90 P2d 787 (1939) (bond issuance procedures).

As also noted by the city, in each of these cases, with the possible exception of Cascade Locks, a state procedure prevailed because an individual was harmed by lack of notice or an appeal from a governmental quasi-judicial decision that had particular impact on that person. In this case there are no asserted procedural infirmities in the local decision-making process which would result in a failure of that process to give proper notice or representation to an affected individual or entity.

Even assuming that the “majority” is correct in its holding that “need” for a state statute is not limited to a need for “procedural protections,” but may properly be extended to a need to safeguard any “interest” of municipal employees, the union still has not sustained its burden under the record in this case to make the required “showing” of “need.” In LaGrande/Astoria I, the “criteria” adopted by this court as a matter of “constitutional command” provide that “when a statute is addressed to a concern of the state with the * * * procedures of local agencies * * * such a state concern MUST be justified by a NEED to safeguard the interests of persons * * * affected by the procedures of local government.” (281 Or at 156).

In this case there has been no showing of such a “need.” There has been no showing that the procedure for arbitration under the PECBA statute would protect the interests of the Roseburg firemen better than those interests would be protected under the “post impasse” procedure provided by the Roseburg ordinance. There has been no showing that after an “impasse” in collective bargaining between a city and the union representing its firemen and police officers, a decision by an arbitrator selected by both the city and the union would be more favorable to the firemen or police officers than a decision made by the voters of the city, much less that the appointment of an *314arbitrator in such a case would “provide a measure of economic security” to firemen or police officers. It follows that there has been no showing of a “need” for the statutory procedure to protect these employees.10

*315Of even more fundamental and controlling significance is the position of the union on the question of “need.” It must be remembered that this is an adversary proceeding. This is an unfair labor practice proceeding in which the union has charged the city with an unfair labor practice in refusing to bargain under the procedure provided by the PECBA statute, contending that the procedure provided by the PECBA statute superseded and invalidated the procedure provided by the Roseburg ordinance. The union, however, has completely failed to make a showing of any “need,” as required before a “procedural” statute will be held to supersede and invalidate procedures provided by a city charter or ordinance.

In its brief as respondent in the Court of Appeals, its only response to the contention by the city that the PECBA statute was “procedural,” thus requiring such a showing of “need,” was that the statute was “substantive,” not “procedural.” In its subsequent memorandum in response to a written question by this court on the specific question of “need,” the response by the union was again that the PECBA statute is a “substantive” statute and is not a “procedural” statute, so as to require a showing of “need,” but that if the statute was “procedural,” the “contrary procedures of the state law” are “justified” (a) by the statements of “policy” of the statute, and (b) by an article published in a 1975 law journal11 to the effect that in the years 1969-1972 there were no work stoppages by firefighters in several cities in other states in which compulsory arbitration was required.

*316For reasons previously stated, the PECBA statute is not a “substantive” statute.12 Also, the fact that the legislature may have been “justified” in enacting the PECBA statute does not satisfy the requirement of LaGrande/Astoria that such a “procedural” statute “must be justified by a need to safeguard the interests” of the persons affected by the “procedures of local government” before the city may be prohibited from following the procedure provided by its ordinance and required to follow the procedure provided by the PECBA statute. As previously noted, the legislature, by enactment of such a statute, cannot “preempt” the question of “need,” which is a question to be decided by the courts.13

Moreover, “need” for such legislation in Oregon cannot properly be established by reference to a published article to the effect that over ten years ago there were no strikes by firefighters in several cities in other states in which compulsory arbitration was required, but with different, if any, constitutional requirements for “home rule.” For all we know from the record in this case, firemen in Oregon, who were previously forbidden from striking in 1963,14 have been able through collective bargaining to achieve wages at a level equal to or higher than the level of wages achieved by firemen in states in which compulsory arbitration is required.

For these reasons alone, if for no other reasons, I am of the opinion that the union has not sustained its burden to make a “showing” of “need” as required by the “criteria” adopted by this court in LaGrande/Astoria and that, as a result, the opinion by the Court of Appeals must be reversed.

*3173. If a procedure for collective bargaining and for the determination of wages for city firemen or policemen in the event of an “impasse” is a matter of “substance,” rather than “procedure,” the provisions of this statute are “irreconcilable” with the city’s “freedom to choose its own policital form. ”

Even assuming, but by no means conceding, that a procedure for collective bargaining and for the determination of wages for city firemen and policemen in the event of an “impasse” is not a statute “addressed to a concern of the state with the * * procedures of local agencies,” the imposition upon a city of the procedure provided by this state statute would be “irreconcilable” with its “freedom to choose its own political form.”

The “majority” holds that the procedure provided by this statute is not “irreconcilable” with the “freedom” of this city to “choose its own political form” for the following reasons:

(a) “PECBA does not deal with the structure of local government.” 292 Or at 277. “This case involves structure and procedures of state government, not local government.” 292 Or at 282. “* * * [T]he political forms and procedures by which a city decides matters which remain within its legislative purview is unchanged.” 292 Or at 281. (Emphasis added) See also 292 Or at 282.

(b) This is so because “an ultimate decision under PECBA is made by a process of arbitration provided and conducted by ERB, an agency of the state. * * * The statute * * * assumes that the arbitrator works [for] the state in the implementation of state policy” (292 Or at 279) and because “PECBA provides for certain decision-making arrangements at the state level as an implementive device for its primary substantive objectives of the state.” 292 Or at 281. (Emphasis added)

(c) There is a distinction between a state statute which provides that programs for city employes are to be administered by a state-mandated local “board” and a statute which, as in this case, provides for administration of such a program by a state board. 292 Or at 277-78.

It does not follow from the single fact that “PECBA does not deal with the structure of local government” and that “PECBA makes no organic change to the *318political form of local government” that this statute, which requires arbitration of disputes between a city and its employees, is not a statute which is “irreconcilable with the local community’s freedom to choose its own political form,” as the “majority” would conclude. For the same reasons, that conclusion does not follow from the finding by the “majority” that “the political forms and procedures by which a city decides matters which remain within its [the city’s] legislative purview are unchanged.”

With respect to the statement by the “majority” that an arbitrator who acts under the provisions of this statute is an “agent of the state,” it would appear, upon examination of ORS 243.746, that the arbitrator is not an “agent of the state,” but is a person who, in most cases, would be selected by the parties themselves (the city and union) to act as a neutral person on behalf of both parties.15 The further statement by the “majority” that the city has admitted that such an arbitrator is an agent of the state is at the least misleading, if not untrue, when that alleged

*319“admission” by the city is read in context.16 But even if the arbitrator is an “agent of the state,” it does not follow that the provisions of the statute, which strip from the city and its city manager the powers conferred by the city charter, are not “irreconcilable” with its “freedom to choose its own political form.”

The further statement by the “majority” that the “decision-making arrangements” provided by PECBA “at the state level” are an “implementive device for its primary substantive objectives of the state,” although perhaps material upon the question whether this statute is “substantive,” instead of “procedural,” is immaterial to the question whether, if “substantive,” its provisions are inconsistent with the freedom of the city “to choose its own political form.”

*320In LaGrande/Astoria I this court said that the “central object” of the Home Rule Amendments is to:

“* * * allow the people of the locality to decide upon the organization of their government and the scope of its powers under its charter without having to obtain statutory authorization from the legislature * * 281 Or at 142. (Emphasis added)

In LaGrande/Astoria II this court said that:

“* * * the limitation expressed in article XI, section 2, should not be read to hinge on whether a city chooses to place a particular policy into its charter or into some other form of enactment pursuant to its charter * * *.” 284 Or at 177-78. (Emphasis added)17

As for the distinction by the “majority” between statutes which provide for programs for city employees to be administered by a state-mandated local board and statutes which provide for such programs to be administered by a state board, any such distinction is no more controlling upon the question whether the statute is inconsistent with the freedom of the city “to choose its own political form” than upon the question whether the statute is “substantive” rather than “procedural,” as previously discussed.18

To illustrate, take a simple hypothetical case: The charter of a small eastern Oregon incorporated city with a population under 500 (of which there are many)19 provides for a mayor, recorder, treasurer, police chief and fire chief, and confers upon the city council the power to enact ordinances setting the salaries of each of these city officials. The police chief is the city’s only police officer. The fire chief is the only paid fireman and is the head of a non-paid *321volunteer fire department. It would seem to be obvious that for the fire chief and police chief to demand not only collective bargaining for an increase in salary, but compulsory arbitration in an effort to secure salaries higher than those provided by such an ordinance, would be “irreconcilable” with the local community’s freedom to choose its own “organization” of local government and “the scope of its powers.”20

Finally, and as contended by the city, in LaGrande/Astoria I this court said that the 1906 municipal home-rule amendments “are concerned with the structural and organizational arrangements for the exercise of local self-government, with the power of local voters to enact and amend their * * * municipal charters and to employ the initiative and referendum for ‘local, special [or] municipal legislation.’ ” (281 Or at 142-43); Roseburg’s provision for resolving by popular vote certain impasses between the city’s administration and the city’s firefighters concerns “the structural and organizational arrangements for the exercise of local self-government,” and to intrude an arbitrator into the resolution of such an impasse would modify, contrary to local preferences, “the structural and organizational arrangements for the exercise of local self-government.”21

*3224. The fixing of salaries and wages of municipal officers and employees is a proper subject of “municipal legislation” and, as such, is subject to the initiative and referendum powers expressly reserved to city voters by Article IV, Section 1(5) of the Oregon Constitution.

There is a further and independent reason why the decision by the Court of Appeals must be reversed. That is because the fixing of salaries and wages for municipal officers and employees is a proper subject of “municipal legislation” and, as such, is subject to initiative and referendum by city voters under Article IV, Section 1(5) of the Oregon Constitution.

Article IV, Section 1(5) of the Oregon Constitution provides that:

*323“The initiative and referendum powers reserved to the people by subsections (2) and (3) of this section are further reserved to the qualified voters of each municipality and district as to all local, special and municipal legislation of every character in or for their municipality or district. The manner of exercising those powers shall be provided by general laws, but cities may provide the manner of exercising those powers as to their municipal legislation. * * *” (Emphasis added)

As previously noted, this court in LaGrande/Astoria I stated, although in a somewhat different context, that:

“* * *cities sometimes place * * * rules for the conduct of government into ordinances, or perhaps resolutions, bylaws, or other forms of enactment allowed by the city’s charter. It is not the label that matters but the role of the provision in local self-government.” (Emphasis added) 281 Or at 150.

The Roseburg ordinance involved in this case provides not only procedures for collective bargaining between the city and the representatives of its employees and for the determination of wages for city employees in the event of an “impasse” in collective bargaining, as previously stated, but also provides that in the event of such an “impasse” and after findings by a factfinding board, the city council shall “call for a special election of the voters of the city” at which the best offers of the city and the union shall be submitted to the voters of the city, whose “decision” shall be “final and binding.”

In spite of the extremely broad reservation of power to voters of Oregon cities of the referendum power as to “all local, special and municipal legislation of every character”, and the statement by this court that “[i]t is not the label that matters, but the role of the provision,” the “majority” holds, in effect, that the Roseburg ordinance under which, in the event of an “impasse” in collective bargaining, the amount of the compensation to be paid to city firemen or policemen is to be submitted by referendum to a vote of the city voters is not “municipal legislation” subject to powers expressly reserved to city voters by Article IV, Section 1(5) for the following reasons:

*324“In the absence of a superseding statute, the city would have been free to legislate an entirely different scheme of employment relations with or without collective bargaining and impasse resolution provisions. By virtue of PECBA, however, the decision (impasse arbitration) is now beyond the city’s choice. It is immaterial to the validity of the statute that the city council has decided by ordinance to refer to a plebiscite a future decision which is no longer the city’s to make. Were it otherwise, local government could cripple the ability of the state to legislate regarding any matter of state policy which affected local governments by the simple expedient of local referendum. The home rule amendments were not intended to have that drastic, general effect.” 292 Or at 288.

With all due respect to the “majority,” this reasoning simply begs the question whether, conceding that a city cannot, by initiative or referendum, adopt legislation “regarding any matter of state policy” (such as by the adoption of a city ordinance providing the death penalty for murder), an ordinance which provides for a referendum to the voters of a city on the question of the amount of compensation to be paid to city officers or employees is “municipal legislation” subject to the powers expressly reserved to city voters by the provisions of Article IV, Section 1(5) of the Oregon Constitution. That question has never before been submitted to this court for decision. That question has, however, been considered by other courts, whose decisions would also be ignored by the “majority.”

In City of Las Vegas v. Ackerman, 85 Nev 493, 457 P2d 525 (1969), it was held (at 457 P2d 528) that:

“* * * The fixing of the salaries of municipal employees in the City of Las Vegas is a legislative function. The people have the power through the initiative process to enact legislation fixing such salaries.” (Emphasis added)

It is of particular significance to note that the controlling provision of the Nevada Constitution reserved to voters of cities the power of initiative and referendum in terms identical to those of Article IV, Section 1(5), i.e., “as to all local, special and municipal legislation of every kind. ” *325To the same effect, see Glass v. Smith, 150 Tex 632, 244 SW2d 645 (1952).22

In view of the broad language of Article IV, Section 1(5), which reserves to city voters the power of initiative and referendum “as to all local, special and municipal legislation of every character,” as well as the length and strength of the democratic tradition in Oregon, in which the right of initiative and referendum was first recognized in 1906 by constitutional amendment, I am of the firm opinion that all voters of Oregon, including all city voters, are entitled to a liberal construction by this court of that broad language of Article IV, Section 1(5), and that, as a result, this court should hold that the fixing of salaries and wages of municipal officers and employees is not only legislation, but also that an ordinance which fixes the amount of compensation to be paid to city officials and employees is “municipal” legislation, so as to be subject to the exclusive control of the voters of a city by the process of initiative and referendum under Article IV, Section 1(5).

Indeed, as noted by this court in Rose v. Port of Portland, 82 Or 541, 562, 162 P 498 (1917), when what was then Article IV, Section la was submitted to the voters for *326approval in 1906 as an amendment to the Oregon Constitution, following a statement by its sponsors which said, among other things:

“The adoption of this amendment will give the people power to control salaries of county and district officers.” (Emphasis added)

Again, to illustrate the point, consider the hypothetical case of a typical small incorporated city with a population under 500 and which employs one police officer, who may be designated by charter or ordinance as its “chief of police.” (As of 1980, there were 60 incorporated cities in Oregon with a population under 500.)23 Assume that in such a small city and under such a charter or ordinance the council should pass a new ordinance raising the annual salary of the “chief of police” from $12,000 to $50,000. Or assume, as a further hypothetical case, that the commissioners of the City of Portland should enact an ordinance raising the annual salaries of not only all commissioners, but also its police chief, to $100,000.

The holding by the “majority” would require one of two results in such a case: (1) that such an ordinance referring to the voters of either the small Oregon city or the voters of the City of Portland the question whether such an increase in compensation to city officals and employees is not “legislation” subject to the referendum powers expressly referred to city voters by Article IV, Section 1(5) of the Oregon Constitution, and that there is no right or power of referendum under such facts, or (2) that even if such an ordinance is “legislation” subject to such a referendum, an ordinance fixing the compensation of city officers or employees is not “municipal legislation” within the exclusive power of city voters to “control,” but that the state may by statute fix or substantially increase the amount of compensation to be paid to city officers and employees.

In view of the basis upon which what was then Article IV, Section la was submitted to the voters of Oregon for approval, either of such conclusions would surely come as a shock to all city voters in Oregon, not to speak of the sponsors of what was then Article IV, Section *3271(a) and those who voted for its adoption in 1906. Either of such conclusions would also make a mockery of the provisions of what is now Article IV, Section 1(5) of the Oregon Constitution, which in the broadest possible terms reserves to all city voters the power of initiative and referendum “as to all local, special and municipal legislation of every character.”

5. The concurring opinion by Denecke, C.J.

The concurring opinion by Chief Justice Denecke is far less drastic than the “majority” opinion in that it appears to agree with this dissent and to disagree with the “majority,” with respect to the following holdings by the “majority”:

(a) The holding by the “majority” that the PECBA statute, which provides a procedure for collective bargaining and compulsory arbitration and supersedes all conflicting city ordinances, is not a “procedural” statute, but is a “substantive” statute, because it is “addressed” to state objectives.

(b) The holding by the “majority” that even if the PECBA statute is a “procedural” statute, so as to require a “showing” of a “need” to safeguard the interests of city firefighters by providing a procedure for collective bargaining and compulsory arbitration, such a required “showing” of “need” is satisfied whenever the legislature “says” that a statute is “needed,” with the result that under the holding by the “majority” the question of “need” is no longer a question to be decided by the courts.

I disagree with the concurring opinion by Chief Justice Denecke, however, in his conclusion that:

“In this case it is apparent to me that there is a ‘need’ to substitute compulsory arbitration, or something akin, for the forbidden right to strike to secure reasonable wages and working conditions.”

As previously stated,24 this court held in LaGrande/Astoria that “procedural” statutes “must be justified by a need to safeguard the interests” of the persons involved. This court did not say how much a “showing of need” can or must be made. I do not contend, as stated by *328the Chief Justice, that such a showing must be made “by the introduction of evidence.” I do contend, however, that in a case such as this in which a union contends that a city is guilty of an unfair labor practice for refusing to proceed under the procedure provided by a statute and for preferring to proceed under the procedure provided by a city ordinance, and in which the union contends that the statute supersedes and invalidates such an ordinance, the union has the burden to make some sufficient showing of such a “need,” whether by evidence, by a “Brandeis-type” brief, or by argument based on reference to legal authorities or one based on logic. For reasons previously stated, the union has completely failed to do so in this case.25 This being so, and in the absence of any such “showing” of “need,” it would be improper, in my opinion, for this court, sua sponte, to do what would amount to a taking of judicial notice of such a “need,” as would be the effect of the concurring opinion by Chief Justice Denecke.

It may be “fair and equitable” to provide by state statute a procedure for collective bargaining and a “post-impasse” procedure for compulsory arbitration for the benefit of city firemen who cannot strike. Under the criteria adopted by this court in LaGrande/Astoria, however, the question to be decided by this court in this case is not whether such a statute is “fair and equitable,” but whether a “need” has been shown for the procedures provided by such a statute to supersede and invalidate the procedures for collective bargaining and post-impasse procedures provided by city ordinances. No “showing” of such a “need” has been made in this case.

6. The concurring opinion by Linde, J.

The scholarly concurring opinion by Justice Linde would sustain the validity of the PECBA statute on theories quite different from those urged by the parties in this case. For that reason, and because his concurring opinion was only recently filed, it has not been possible, as a practical matter, to analyze it in depth or to comment upon it beyond the following observations:

*329(a) The requirement of a showing of “need.”

I agree with statements by Justice Linde (292 Or at 290-292) of reasons why the PECBA statute may not be a “substantive” statute. The fatal flaw in his concurring opinion, however, is his attempt to “sweep under the rug” the “criteria” adopted by this court in LaGrande/Astoria I in an opinion by Justice Linde himself, in which it was held that as a matter of “constitutional command” (281 Or at 156) a statute “addressed to a concern of the state with the structure and procedures of local agencies * * * MUST be justified by a NEED to safeguard the interests of persons * * * affected by the procedures of local government.”

As previously noted, it is evident from the decision by this court in LaGrande[Astoria that the question whether there has been a sufficient showing of such a “need” is a question to be decided by the court.26 As also previously noted, the holding by the “majority” opinion is contrary to this requirement, as stated in LaGrande/Astoria, in that the “majority” now holds that the entire question of “need” is for the legislature to decide, thus abdicating that judicial function to the legislature and further emasculating what little was left by LaGrande/ Astoria as an area of “home rule” in which cities can legislate without interference by the legislature.27 The concurring opinion by Linde, J., completely ignores this fundamental question.28

That concurring opinion (at 292 Or at 296) discusses the nature of the “interests” which are subject to statutory “procedural protections” under those LaGrande/ Astoria “criteria” and concludes that “the interest of city employees in the process by which these matters [wages, *330working conditions and other benefits] are decided is * * * appropriate for legislative protectionThe question to be decided, however, is not whether such interests are “appropriate” for legislative protection, but whether legislative protection has been “justified by a NEED to safeguard” such interests. Such a “need” is not demonstrated simply by the state’s preference for the procedure provided by the statute over the procedure provided by the ordinance.

(b) The delegation of power to specify wages, etc.

The concurring opinion says that because the state could specify wages, hours and working conditions for city firemen, it can delegate such power to one arbitrator, if the arbitrator is an agent of the state. No such contention was made by the parties in this case.

I do not agree with the assumption that in Oregon, in which “home rule” is protected by the Oregon Constitution, the state legislature can “specify” wages, hours and working conditions for city firemen. Furthermore, under the terms of the PECBA statute, the arbitrator to be selected under the procedure provided by that statute is not a state arbitrator to whom any such power can properly be delegated, for reasons also previously stated.29

(c) The appealability of an arbitration award.

The concurring opinion also suggests (292 Or at 295) that a further reason why the PECBA statute cannot be held to be invalid in this case is that wages, hours and working conditions for firemen will not be determined in this case and that a second appeal may be taken from an award by an arbitrator selected under the procedure provided by the statute if the award “contain(s) provisions impinging on a city’s reserved home rule powers.” Again, no such contention was made in this case, and such a holding would inject a further element of uncertainty into the entire arbitration process.30

*331(d) The referendum question.

It is stated by Justice Linde (292 Or at 297) that:

«* *■ * a city’s own ‘municipal legislation’ is superseded by an otherwise valid state law, even a special or local law, this clause has no further application.”

This statement begs the question whether, under the provisions of Article IV, section 1(5), “municipal legislation” adopted by a city can be “superseded” by an “otherwise valid state law.” As stated above, in the discussion of this question in response to the “majority” opinion, it has been previously recognized by this court that Article IV, Section 1 was submitted to Oregon voters as a proposed constitutional amendment with the respresentation by the framers of that proposed amendment that its adoption would “give the people power to control salaries of county and district officers.”31 Such a promised purpose would be frustrated if a city ordinance adopted pursuant to Article IV, Section 1(5) prescribing salaries for city officers could be “superseded” by a state statute, as would be the result of the concurring opinion by Linde, J.

“Post Mortem”

Now that “home rule” for Oregon cities is “dead and buried” for most, if not all, practical purposes, a “post mortem” may be appropriate. I am concerned not only with the result of the decision in this case, but with the grounds on which the decision by the “majority” is based and its devastating effect in future cases by an opinion which would confer almost, if not complete, legislative “supremacy” over cities by its holdings that:

(1) Even a statute which provides procedures which conflict with procedures provided by city ordinances will not be considered as one “addressed to * * * procedures of local government” if such a statute “is addressed primarily to substantive social, economic or other regulatory objectives of the state.” Not only are all statutes presumably “addressed” to some such “objective,” but under the decision by the “majority” that determination is one which can now be foreclosed by legislative decision.

*332(2) Even statutes which are “addressed to structure and procedures” of local government will prevail over conflicting city ordinances whenever the legislature decides, either expressly or by implication, that there is a “need” for such a statute.

(3) City voters do not have the power by initiative or referendum to “control” the compensation paid by cities to their officers and employees, contrary to the representation to Oregon voters when Article IV, Section 1 was submitted to them for approval as an amendment to the Oregon Constitution.

The sole response by the “majority” to these three “post mortem” conclusions is a quotation from Mark Twain that reports of his death were “greatly exaggerated.” This is a clever, but evasive, response. The “majority” does not deny that its opinion in this case has these three drastic results.

In addition to the obvious effects of such needless and drastic holdings by the “majority” in this case, the additional consequences of these holdings, when considered with the two concurring opinions, include the following:

(1) The “criteria” adopted by this court in LaGrande/Astoria and purportedly based upon “precise” provisions of Article XI, Section 2, so as to constitute a “constitutional command,” are illusory, meaningless, and impractical in application, as demonstrated by the fact that a majority of this court cannot agree upon the meaning and application of such “criteria.”

(2) The distinction by the “majority” of this court in LaGrande/Astoria between statutes “addressed” to matters of “procedure” and those “addressed” to matters of “substance” (a distinction repudiated by the union, the Attorney General and the cities in that case) is also illusory and meaningless, as the same distinction has been found in other contexts.32

(3) The test of “predominant interest,” as previously reaffirmed by unanimous decision of this court in *333Heinig and approved by many authorities on the subject,33 but rejected by this court in LaGrande/Astoria in a four to three decision, is not only a more honest and straightforward test, but one far less difficult in application than the “criteria” adopted in LaGrande/Astoria. Tests involving a “balancing” of interests have been adopted by courts in many other contexts.34

(4) Under these holdings by the “majority,” the Oregon legislature could adopt statutes which, by “preemption,” not only supersede any “structure or procedures” provided by any city charter or ordinance, but also impose, either directly or indirectly, heavy financial burdens on cities without providing funds with which to meet such burdens.35

(5) The only remaining recourse available to the cities, as a practical matter, is to promote the adoption of an amendment to the Oregon Constitution which “spells out” more specifically the “home rule” rights of cities and counties.

For all of these reasons, I dissent from the opinion by the “majority” in this case.