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.
ORS 183.470(2) provides:
“A final order shall be accompanied by findings of fact and conclusions of law. The findings of fact shall consist of a concise statement of the underlying facts supporting the findings as to each contested issue of fact and as to each ultimate fact required to support the agency’s order.”
The dissent makes some additional arguments which can be briefly answered. It argues that the state has made no “demonstration” or “showing” of a need to safeguard public employees’ interest, but the dissent does not prescribe any particular manner in which it would require the legislature to make such a showing. Although statements of policy are not essential to statutes, this statement adequately describes such a need.
*285Also, the dissent notes that previous cases upholding state regulation of local procedures were based on safeguarding interests of a due process nature. See, e.g., City of Woodburn v. State Tax Com., 243 Or 633, 413 P2d 606 (1966); Boyle v. City of Bend, 234 Or 91, 380 P2d 625 (1963); State ex rel Heinig v. Milwaukie, et al, 231 Or 473, 373 P2d 680 (1962); City of Cascade Locks v. Carlson, 161 Or 557, 90 P2d 787 (1939); City of Klamath Falls v. Oregon Liquor Comm., 146 Or 83, 29 P2d 564 (1934); Woodburn v. Public Service Commission, 82 Or 114, 161 P 391 (1916); Branch v. Albee, 71 Or 188, 142 P 598 (1914). The fact that this court has historically upheld statutes safeguarding interests does not mean that other interests, such as collective bargaining, cannot also be safeguarded.
Finally, the argument of the dissent that firefighters have demonstrated adequate political effectiveness or “clout” to protect their interest without the need for legislative protection fails on several counts. The argument says nothing about the interests of other protected persons such as jail guards, for example, whose wages are apparently not intended by the legislature to depend upon their skills at political organization. The argument goes to the wisdom of PECBA, not to its constitutionality, and we express no opinion on the political or legislative wisdom of this legislation.
In applying this rationale, we express no view regarding other possible challenges. For example, much of the dissent is devoted to establishing the proposition that salary setting is legislation. For purposes of this opinion, we do not mean to imply that an arbitration plebiscite is or is not “legislation” subject to referendum. See: State ex rel Allen v. Martin, 255 Or 401, 465 P2d 228 (1970); Tillamook PUD v. Coates, 174 Or 476, 149 P2d 558 (1944); Whitbeck v. Funk, 140 Or 70, 12 P2d 1019 (1932); Monahan v. Funk, 137 Or 580, 585, 3 P2d 778 (1931); Campbell v. City of Eugene, 116 Or 264, 274, 240 P 418 (1925); Amalgamated Transit v. Yerkovich, 24 Or App 221, 545 P2d 1401 rev den (1976).
Rose v. Port of Portland, 82 Or 541, 562, 162 P 498 (1917), relied on by the dissent, is not controlling for several reasons. First, of course, we are dealing with the amendments themselves, not a voters’ pamphlet statement of their effect under the statutes in effect that year. Also, the dissent would apparently equate local “employees” with “officers.” There is no suggestion in Rose that the scope of local initiative and referendum authority over “salaries” of local “officers” reaches the wages of every local government employee.
Also, the dissent relies on the example of a small town in which the police chief has no police force and the fire chief supervises a force of volunteers. The analogy is not apt. A police or fire chief is management or supervisory. They are not “public employees” as that phrase is used by the statute. ORS 243.650. Hence, such officials would not be included in a bargaining unit. This argument of the dissent is therefore fallacious.
Finally, the so-called “post-mortem” is, like Mark Twain’s premature obituary, “greatly exaggerated.” Rather than reply point by point, it is sufficient to say here that the opinions in LaGrande/Astoria were carefully limited and that this decision probably demonstrates certain outer limits in the application of the home rule *288analysis in that case. Limitations upon the state were spelled out there and here. What we hold and what we do not hold should be discerned from this opinion and not from the “post-mortem.”