LEVIN, J.

This is an action to review the award of an Act 312 arbitration panel resolving a collective bargaining dispute between the Detroit Police Officers Association and the City of Detroit.

We granted bypass of the Court of Appeals to review a judgment of the circuit court enforcing the award.

I

The city challenges Act 312, and the award in the instant case, on constitutional grounds. It additionally contends that the panel's award on the economic issues of wages and COLA and of a hardship exemption to the city's residency requirement is not supported by competent, material and substantial evidence on the whole record, and that the chairman of the panel erred in refusing to admit relevant evidence concerning economic demands of the DPOA.

We dissent from the Court's disposition:

1) We would hold the act unconstitutional on the ground that it delegates the power to prescribe policy (law) to the chairman of the panel without adequate safeguards circumscribing the exercise of that power.

2) We are of the opinion that the last-best-offer feature of the amended act exacerbates rather than cures the deficiencies which lead us to conclude that the act is unconstitutional, so that, entirely apart from whether the act would be unconstitutional without this feature, it is unconstitutional with this feature.

3) Although we would hold the act unconstitutional, we would, for the reasons set forth in Dearborn Fire Fighters, give that decision prospective effect only.

For over a year the police officers have been paid the increased wages and COLA awarded by the panel. It would cause undue hardship to call upon them to repay what has already been paid.

4) That consideration, however, does not require that we enforce the full unpaid retroactive portion of the award.

The chairman of the arbitration panel in the instant case indicated (see fn 60) that he might have made an economic award less costly to the city had he not been required to select the last best offer of one of the parties.

It would not cause undue hardship to readjust the amount of the award as of July 1, 1977 to eliminate the excess implicit in adoption of a last best offer and to offset the excess already paid against the unpaid retroactive compensation.

We would remand the awards in the instant case and the Lieutenants and Sergeants Association case (see ¶ 5, infra) to the respective panels for redetermination of the economic awards unrestrained by the last-best-offer limitation, on the basis of the records already made, supplemented in this case by the economic evidence previously excluded.

Until such a redetermination, the police officers should continue to be paid on the basis of the award and the stay on payment of retroactive amounts should continue in force. After redetermination of the economic award, i) current payments should be readjusted, ii) the retroactive amount payable should be based on the new award, and iii) that amount should be reduced by the amount of the overpayment, but in no event should the officers be required to repay any amount.

5) The hearing on remand should be deferred until the Court decides the city's appeal from an arbitration award to the Detroit Police Lieutenants and Sergeants Association.

After oral argument in the instant case this Court entered an order holding that case in abeyance pending this decision.

The chairman of the panel in this case indicated that he was persuaded to follow the decision in that case so that police officers would not be treated differently than lieutenants and sergeants.

The merits in this case thus depend in part on the merits of the city's appeal in that case and therefore there can be no final decision in this case before decision on the city's application for leave to appeal in that case.

A

The question on which the four Justices participating in Dearborn Fire Fighters (decided in June, 1975) were equally divided is again presented in the instant case.

Two (Justice KAVANAGH and myself) would have held the act unconstitutional as an unlawful delegation of legislative power because it entrusts chairmen of ad hoc arbitration panels "with the authority to decide major questions of public policy concerning the conditions of public employment, the levels and standards of public services, and the allocation of public revenues" but, because their appointments are only for an individual dispute and they have no continuing responsibility, insulates their "decision[s] from review in the political process" which "is not consonant with the constitutional exercise of political power in a representative democracy". We added that the Legislature could vest "authority to resolve disputes concerning public employees in a governmental officer or agency with continuing responsibility for the day-to-day exercise of that delegated power". "Although the present law's provision for hit-and-run arbitrators is constitutionally defective, a law providing for a continuing politically responsible arbitrator could meet the Constitution's demands."

Justice WILLIAMS said that while selection of the chairman by the municipal and employee representatives might not satisfy the requirements of public accountability, the act was constitutional insofar as it provided for his appointment by the chairman of the Michigan Employment Relations Commission when the parties fail to agree on a chairman (as was done in Dearborn Fire Fighters). He said that the "proximity of the appointment to the electoral process guarantees a high degree of political accountability".

Justice COLEMAN said that the statute was constitutional because the Legislature intended to vest in the panel the "sovereignty necessary to accomplish its mandates" and the "panel is a public body performing a public function".

In the instant case, a majority of the Court, in separate opinions, concludes that the act is constitutional. Justice WILLIAMS, speaking for himself, and Justices RYAN and MOODY, states that, if accountability is constitutionally required, "as a practical matter" "the scheme exhibits adequate accountability".

B

We would hold the act unconstitutional as violative of the delegation doctrine.

The delegation doctrine seeks to safeguard against excessive delegation and misuse or abuse of delegated law-making power.

The chairman of the panel is empowered to formulate public policy on such subjects as wages, hours of employment, pensions, retirement, residency, affirmative action, health insurance and crew size. He does not merely find the facts and apply thereto established policy. Rather, he fashions the public policy for the case at hand — at once drawing the policy line and deciding on which side of the line the case falls. If the chairman's only role was to find the facts and apply to them policy predetermined by the Legislature or, say, the Michigan Employment Relations Commission — if all he did was decide the "on which side of the line" question — there would be less, perhaps no, need for the safeguards the absence of which we believe renders this act unconstitutional. He is not confined by the statutory factors; they are sufficiently amorphous to allow the chairman to reach and justify a result in favor of either party on virtually any issue.

The power to prescribe public policy having the effect of law is generally delegated only to full-time public officers or agencies. This provides a structure for the development of policy in a uniform and politically responsive manner.

Act 312 is novel in that the policy-making power is dispersed among ad hoc arbitrators. The dispersal of policy-making power prevents the emergence of visible and intelligible principles. It is often difficult to separate the policy-making from the fact-finding in any individual award, and no Act 312 chairman has any duty to develop a coherent body of law or to follow the decision of another Act 312 chairman.

The failure to develop a body of policy masks the Legislature's responsibility for the policies made by the chairmen in its stead and is not consonant with the concept that public policy should be developed in a politically responsive manner. Constructive political response presupposes that the electorate and the Legislature can determine the policy being applied.

This requires that the delegation of policy-making power be structured in a manner which does not inhibit the emergence of intelligible principles and coherent policies to guide the resolution of individual cases.

The potential for non-uniformity in policy inherent in the diffusion of state power among ad hoc chairmen is itself an evil. Just as the Legislature cannot enact laws which treat the same situation differently, neither can it empower others to do so.

The vague and malleable factors enumerated in the statute neither confine the chairmen's power nor supply a meaningful legislative statement of public policy. Because the factors are so general and the chairman decides how they are weighed and applied, any reasonable decision can be justified. The limited judicial review provided by the act does not safeguard against misuse or abuse of delegated policy-making power.

Last-offer arbitration exacerbates rather than cures the constitutional deficiencies of the act by divesting the chairman of the power to fashion the most appropriate award. Unless one of the last offers coincides with what in the chairman's judgment is the most appropriate award, the last-best-offer feature prevents the process from yielding results reflecting the best public policy, producing instead a result which may arbitrarily assess a penalty on the losing party measured by the difference between the last offer selected and the award the chairman, if not so constrained, would have made. Policy-making power is thus taken from government and conferred on the parties, the possibility of principled decision-making is reduced, and the risk of non-uniformity is increased. Because he is limited to the last best offer, the chairman can disassociate himself from the award. It is unrealistic to argue that the structure is politically responsive when even the chairman need not take responsibility for the award.

Finally, the revision of the process for selecting panel chairmen has not altered the basic structural defects of ad hoc interest arbitration of public policy. As before, the structure does not provide for consistency of the decision-maker, decisional principles or fundamental policy, or for the development of a body of precedent from which policy may be deduced.

A delegation of the power to make basic policy decisions declaring the priorities of government and the allocation of revenues is permissible only where the structure of such a delegation does not cloak the legislative nature of the power being exercised and of the policy decisions being made, promotes uniformity in the development of policies and in the application of the policies so developed, and does not mask the accountability of elected officials for the actions of those who wield the delegated power and the responsibility of the Legislature to rectify erroneous policy judgments made on its authority.

In Dearborn Fire Fighters, we contrasted the prototypical delegation with arbitral delegation: Where "decision making is concentrated in one person or agency exercising the delegated power on a continuing basis, the public can focus on the manner in which the power is exercised and can hold the appointing authority accountable. Here decision-making power has been dispersed through so many individual, independent arbitrators that it is not possible to hold any public official or authority responsible for the manner in which the delegated power has been exercised."

We added that "[t]he primary obligation of government is to govern. We seek to safeguard the final authority of government. Decision making by an independent outsider in operative effect excludes government as the final authority", and that:

"It is the unique method of appointment, requiring independent decision makers without accountability to a governmental appointing authority, and the unique dispersal of decision-making power among numerous ad hoc decision makers, only temporarily in office, precluding assessment of responsibility for the consequences of their decisions on the level of public services, the allocation of public resources and the cost of government, which renders invalid this particular delegation of legislative power."

II

Implicit in the Legislature's constitutionally conferred power to "enact laws providing for the resolution of disputes concerning public employees" is the power to delegate the resolving authority. The city argues that Act 312 is unconstitutional because it improperly delegates policy-making power to politically unaccountable arbitrators.

Justice WILLIAMS' analysis of the delegation issue concludes (1) that the Act 312 standards "are as reasonably precise as the subject matter requires or permits" and adequately channel the chairman's decision-making under the principle of Osius v St Clair Shores, and (2) that the revised method of selecting the chairman adequately provides for public responsibility and accountability.

We disagree with his second conclusion and believe that the first answers a question which is too narrowly framed. The delegation doctrine and other constitutional principles seek to protect against excessive delegation or misuse and abuse of delegated law-making power. "As a practical matter", Act 312 does not provide adequate safeguards respecting the law-making power delegated to the chairman of the arbitration panel.

A

The delegation doctrine is rooted, at least in part, in the constitutional provision vesting the "legislative power" in the Senate and House of Representatives and reflects the fundamental democratic premise that the law should be made by representatives chosen by the people. The history of the doctrine reflects attempts to reconcile that precept with the complexity of the governing process. Legislative declarations frequently require interstitial interpolation and not every aspect of every subject lends itself to prospective regulation.

By the 1950's much state delegation jurisprudence had crystallized around the concept that a legislature must articulate "standards" to guide the exercise of delegated power. Osius exemplifies that mode of analysis.

Writing in dissent in Arizona v California, Mr. Justice Harlan explained the purpose of this test:

"The principle that authority granted by the legislature must be limited by adequate standards serves two primary functions vital to preserving the separation of powers required by the Constitution. First, it insures that the fundamental policy decisions in our society will be made not by an appointed official but by the body immediately responsible to the people. Second, it prevents judicial review from becoming merely an exercise at large by providing the courts with some measure against which to judge the official action that has been challenged."

B

We believe those objectives to be sound. The judicial search for adequate statutory standards has, however, proved ineffectual in assuring legislative responsibility for "fundamental policy decisions" and in structuring the exercise of delegated power.

Professor Davis, a leading advocate of acknowledging the necessity and propriety of delegating policy-making power to administrators, has forcefully criticized the standards test, has argued persistently and influentially that the more appropriate inquiry is whether the totality of standards and safeguards surrounding the delegation of legislative power sufficiently protects "against unnecessary and uncontrolled discretionary power".

Other commentators, calling for the doctrine's re-emergence in the federal sphere, have emphasized its underlying function of guaranteeing ultimate legislative responsibility for the formulation of public policy (law).

Thus, the delegation doctrine is not today monolithic. Professor Cooper has "demonstrated that it is not possible to devise a 'true test' for determining whether a particular delegation will be sustained, one articulating and enumerating all the factors that motivate the judicial response. In the final analysis, courts, 'weighing the advantage of [the] delegation against the hazards involved,' make a pragmatic analysis of whether the consequences of the delegation are so undesirable as to require judicial intervention".

The hazards involved may range from the legislature's parting with or abdicating its policy-making role to the risk of arbitrariness or injustice resulting from the delegate's failure to forge coherent policy to guide the day-to-day exercise of the delegated power.

A court reviewing a challenged delegation of legislative power should, we agree, examine whether adequate checks have been provided against arbitrary or uncontrolled official action. Such an inquiry cannot, however, supplant the basic inquiry whether the legislatively devised framework for official action — considered in its application — secures the fundamental goal of the delegation doctrine: preserving legislative responsibility for the determination of public policy.

Mr. Justice Brennan has said that "if state power is to be exerted," the choices among competing social and economic theories in the ordering of life

"must be made by a responsible organ of state government. For if they are not, the very best that may be hoped for is that state power will be exercised, not upon the basis of any social choice made by the people of the State, but instead merely on the basis of social choices made at the whim of the particular state official wielding the power. If there is no effective supervision of this process to insure consistency of decision, it can amount to nothing more than government by whim. But ours has been 'termed a government of laws, and not of men.' Marbury v Madison, [5 US (1 Cranch) 137, 163; 2 L Ed 60, 69] (1803)." (Emphasis supplied.)

III

In Dearborn Fire Fighters we said that Act 312 was violative of the delegation doctrine "because it is designed to insulate and, in fact, does insulate the decision-making process and the results from accountability within the political process". We adhere to that view.

Accountability requires a structure which does not inhibit the emergence of intelligible principles and coherent policies to guide the resolution of individual cases. The possibility of constructive political response presupposes that the electorate and the Legislature can determine the policy being applied.

Act 312 arbitration is novel in that the policy-making power is dispersed among ad hoc arbitrators, which prevents the emergence of visible and intelligible principles. For this reason safeguards generally regarded as adequate in the context of delegations to administrative agencies do not assure accountability in this context.

A

Act 312 arbitrators formulate public policy.

The factual components of a dispute between a city and a police or fire fighters union and the term "arbitration" obscure the policy-making components of the arbitrator's decision and the safeguards requisite to constitutional exercise of delegated policy-making power. The arbitrator decides not only whether the particular controversy is on one side or another of a line, but where the line is to be drawn. In drawing the line, although guided by statutorily prescribed standards, he decides public policy in the exercise of law-making power delegated to him by the Legislature.

We observed in Dearborn Fire Fighters:

"When the same term, here 'arbitration,' is used in different contexts, the analysis may become blurred. While both 'interest' arbitration and 'grievance' arbitration concern disputes, the nature of the dispute in one case is considerably different than the other.

"Grievance arbitration concerns disputes arising under written agreements negotiated and agreed upon by the parties. In grievance arbitration, the labor arbitrator acts in a judicial or quasi-judicial capacity. He determines the facts and seeks an interpretation of the agreement in accord with the understanding of the parties as gleaned from the writing and the relationship.

"In interest arbitration, the functions and prerogatives of the arbitrator are significantly different. He is not bound by the agreement or understanding of the parties. He does not interpret a contract, he makes one. He then imposes his concept of what the 'agreement' ought to be on the parties." (Emphasis in original.)

The notion that an Act 312 arbitrator is merely deciding a dispute is inaccurate. "[H]is decision does more than resolve the differences between the parties. It affects the allocation of public resources, the level of public services provided the community as a whole and the cost of government. It also establishes precedents affecting the terms and conditions of public employment generally in both the directly-affected and other units of government".

"It may not be prudent or possible to cut expenditures or reduce personnel. There is a level below which police and fire services cannot safely be reduced. * * * Viewed from the governmental perspective, from the perspective of the citizens/taxpayers, the decision is legislative-political."

B

The power to prescribe policy having the effect of law is generally delegated only to full-time public officers or agencies. This provides a structure for the development of policy in a uniform and politically responsive manner.

While some writers assert that the Legislature may entrust the making of policy to others so long as the one to whom the power is delegated develops meaningful principles to guide the exercise of the power in individual cases and others would require that more specific policy direction be provided by the Legislature, virtually all discussions of the problem assume that the power in question is reposed in some body with continuing responsibility, typically an administrative agency.

Implicit in a delegation to such a full-time public officer or agency are the safeguards resulting from such continuing responsibility and political accountability. Generally agencies develop policies through quasi-judicial decision of individual cases or promulgate rules pursuant to an administrative procedures act. Their published decisions, separately stating the facts and the policies enunciated, and their published rules, tend to create a sense of agency responsibility for developing a coherent body of law.

The delegation doctrine developed in this context and thus it has not heretofore been necessary to articulate, and therefore few precedents recognize, the importance of the safeguards implicit in the centralization and continuing exercise of policy-making power and the mischief resulting from the unsupervised dispersal of policy-making power. It does not follow that such safeguards, generally implicit, are not required.

C

Most courts have sustained interest arbitration statutes but they have done so without considering whether the traditional safeguards suffice when the policy-making power is dispersed among ad hoc arbitrators. A few decisions have invalidated such statutes for the absence of the political accountability which Justice KAVANAGH and I would have required in Dearborn Fire Fighters and would now require. About the same number have concluded that political accountability is not required.

1

The decisions of Act 312 arbitrators involve fundamental questions of policy, principally the allocation of public funds. Yet it is left to the chairman to draw the lines and thus determine the policy which will govern his decision.

The act's standards, rather than "significantly channel[ing]," or "trenchantly circumscrib[ing]" the panel's decision-making authority, are sufficiently amorphous to allow a panel confronted with a given set of evidence to reach (and justify) a result in favor of either party on virtually any issue. See Section IV.A.1., infra.

Let us hypothesize that two panels chaired by different arbitrators are simultaneously convened to resolve disputes in two different communities. Each dispute involves the same issues: residency requirement and crew size. In each case, the union insists upon no residence requirement and no one-officer patrol cars; the city offers residency and one-officer crews.

Assuming no relevant city ordinance, only two § 9 statutory criteria (see fn 48) — clause (c), interests and welfare of the public and clause (d), "comparables" — seem to be applicable. Suppose that testimony shows that officer morale would be higher with two-officer patrol teams and no residency requirement, but all "comparable communities" utilize one-officer patrol cars and require officers to reside in the community.

One chairman might award two-officer patrol cars and no residency requirement because, in his judgment, maintaining officer morale protects the public and is more important than maintaining uniformity with comparable communities. Another chairman might find that the data on "comparables" outweighs any evidence adduced by the union because uniformity should be maintained.

Two different results, expressing two conflicting policy judgments, would have been reached in effectuating the aim of the act to resolve police or fire fighter labor disputes. Yet each of the awards would represent a judgment made on behalf of the people of the state concerning the policy which should govern residency and staffing requirements.

If we assume no significant differences between the communities involved in our hypothetical arbitrations, it is impossible to reconcile the two awards as expressions of the same legislative policy. If we suppose that factual differences accounted for the differing results, we still may not know what rule of decision, what intelligible principle explains the divergent outcomes.

No Act 312 arbitrator has any duty to further the development of a coherent body of law or to follow the decision of another Act 312 arbitrator. The absence of principled decision-making is inherent in the ad hoc structure of the act.

2

This framework, inhibiting the emergence of a coherent body of law, precludes meaningful political accountability.

When sensitive policy judgments of the sort made by Act 312 arbitrators are not based upon any known principles, they cannot be said to be based upon legislatively derived principles.

Because questions of policy are not being decided in the first instance by the Legislature, the decisions once made must be accessible to legislative scrutiny. As one commentator has noted, the delegation doctrine "has an underlying core of validity in that it requires that those who have been selected by a given process and from a given constituency retain the power to make ultimate policy decisions and override decisions made by others". (Emphasis added.)

The Legislature is uninformed whether to override policy decisions made by Act 312 arbitrators because those decisions are made only for the individual case. Any consistency of judgment that might be discernible in the multitude of decisions may be sheerly fortuitous.

A shift from ad hoc decision-making to a structure providing for principled development of arbitral guidelines would heighten the visibility of the process; the absence of discernible principles not only prevents the Legislature from effectively supervising and participating in the process, but also prevents the public from being able to scrutinize what is being decided in its interest.

3

The dispersal of policy-making power among ad hoc arbitrators invites non-uniformity of policy in the dispute resolution process, in itself an evil.

A fundamental precept of both the state and federal constitutions is the concept of equal justice under law. Whatever its doctrinal basis, it is clear that it is not consonant with governance in a constitutional democracy for those who exercise the powers of government to apply different policies in the same situation.

Just as the Legislature cannot enact laws that treat the same situation differently, neither may those who exercise delegated law-making power. The Legislature cannot constitutionally empower state functionaries, whether called administrators or arbitrators, to apply different policies of state government in the same situation on such questions as residency, affirmative action, crew size, pensions or wages, or the allocation of tax revenues between the police, fire protection and other functions of local government. If the rule imposed by the state for a community having the characteristics of, say, Brighton is to require officers to be residents, then the same rule must be applied to all other "Brightons". Otherwise disparate rules for the same situation, which the Legislature could not enact, are being applied.

Inherent in the diffusion of state power among ad hoc arbitrators is potential non-uniformity in the policy-making function. While a chairman may choose the line drawn by another chairman, each is free to draw his own. There may or may not be substantial uniformity. The difficulty or impossibility of determining what policy decisions have been made by the many chairmen and the extent of uniformity or diversity in state policy in similar situations highlights that the Legislature has enacted a policy-making procedure which does not assure that like situations will be treated in a like manner. Two chairmen faced with identical disputes might employ different rules to reach the same or different resolutions. Even if each chairman fully explained his reasoning, there would be no assurance that the rule enunciated would be applied in future cases by the same arbitrator or others.

4

A related vice flowing from the absence of principles is that the Legislature is effectively accomplishing indirectly that which it cannot do directly: the creation of what are, in practical effect, local laws. If the arbitral process were designed to culminate in statements of principles with general applicability, then the act would be creating not local, but general, laws.

D

We do not wish to be understood as saying that the Legislature may not adopt a mandatory procedure for resolving collective bargaining disputes. Although the inability to produce observable principles is inherent in the present ad hoc structure of the act, case-by-case resolution of disputes can be structured so that consistency and continuity (and hence visibility and amenability to legislative supervision) are built into the system.

It cannot be said that the kind of issues presented in Act 312 arbitrations are not capable of resolution by principles. Just as agency action can be undertaken in such a way that rules evolve, so might principles develop in the impasse resolution framework. We recognize that such principles may of necessity be complex in many cases. But some of the reasoning employed by arbitrators belies any claim that the decision is based on the unique facts of the case at hand.

In the instant case, for example, it appears that the reason COLA was awarded is that COLA was thought to be fair in a contract of this duration. It is difficult to conceive of a situation to which that reasoning would not be applicable; whatever characteristics are unique to Detroit played no part in the chairman's reasoning.

Other issues, we would agree, depend on more complex analysis of numerous factors and characteristics. In such cases, however, the interplay of the applicable factors can be described in general terms which may evolve into increasingly specific models through the accretion of precedent from a range of factual situations. "While it is sometimes inevitable that the rule of decision be developed on a case-by-case basis, this development should not be confused with a system under which decisions are made without rules."

What is important is that there be a continuing body responsible for the development of coherent principles and duty-bound to explain the basis for distinction when a result different from that of past cases is reached. We repeat our observation in Dearborn Fire Fighters that we would "not preclude the Legislature from vesting the authority to resolve disputes concerning public employees in a governmental officer or agency with continuing responsibility for the day-to-day exercise of that delegated power".

IV

We turn to more specific consideration of Act 312 as it has been amended and of the arguments advanced in support of its constitutionality.

A

The traditional safeguards — standards, advance rule-making and judicial review — provide no meaningful protection against excessive delegation and misuse and abuse of delegated law-making power by Act 312 arbitrators.

1

Although the factors provided by the Legislature as guides for arbitral decision-making may satisfy traditional standards-oriented inquiry, they do not adequately protect against excessive delegation or unnecessary and uncontrolled discretionary power. Indeed, they reveal the policy-making power of the arbitrator.

The statement of "public policy" in § 1 is no more than a general directive that arbitrators shall go forth and settle disputes. Nor do the § 9 standards, ranging from the specific to the catch-all "[s]uch other factors not confined to the foregoing which are normally or traditionally taken into consideration", supply the coalescing element of policy. They are simply suggested ingredients in a recipe whose proportions are unknown until the end product is served up by the arbitrator. In most cases, particularly in the context of last-offer arbitration of economic issues, any reasonable decision can be justified in terms of one or more of the § 9 criteria.

Those factors may serve to focus the arbitrator's inquiry but they provide only an illusory safeguard against haphazard decision-making. As a practical matter, the Legislature has offered the arbitrator little direction concerning what he should consider or how he should reach a decision and the electorate still less assurance that Act 312 arbitration awards will be fashioned through a principled, consistent, or fathomable process.

The two § 9 factors identified by our colleague as the "[m]ost salient * * * legislative directives" are also the least specific in defining their own substantive content and the most amenable to ad hoc definition and result-oriented characterization by the arbitrator.

Factor (c) requires the arbitrator to consider the "interests and welfare of the public", but he retains the essentially legislative freedom to decide what the public welfare requires before evaluating what course of action will better serve those interests.

In this case, the chairman thought that the public welfare required awarding the DPOA-requested COLA provision because to do otherwise after the Lieutenants and Sergeants Association had won COLA through Act 312 arbitration would "imperil morale and effective performance for the protection of the public". The chairman noted that the "interest and welfare of the public * * * would include, of course, the quality of services delivered" (emphasis supplied) but made no mention of the public interest in the quantity of services delivered, i.e., the reduction in police manpower or other city services that the cost of the police officers' COLA might necessitate. Under the act he was free in this manner to selectively determine which considerations affected the public welfare.

Factor (c) also speaks of the "financial ability of the unit of government to meet those costs". The chairman of the DPOA panel said:

"[T]he city has not [pled] nor has the city proved in the traditional sense an inability to pay, but has, very specifically said, it is not pleading inability to pay but rather the financial condition of the city as a 'limiting factor.' The difference may be subtle, but the difference is fundamental since the panel was not in a position, the budget not having been introduced and debated, to ascertain what other needs of the city were met by those funds already spent in relation to the importance and the amounts necessary for funding of the future award of this panel."

The chairman thus suggests that a city's claim of inability to pay the union's wage demands will not receive serious consideration unless the asserted inability is absolute.

To attach significance to a city's financial position only if payment would bankrupt the city is effectively to eliminate this factor from the equation. Seldom will a single arbitration award threaten to exhaust the entire city budget. The chairman's statement demonstrates an assumption that other budgeted items are presumptively of a lower priority than police or fire fighter salaries and can be sacrificed to provide additional funds to meet police and fire fighter demands — the prototype of a decision which should be entrusted only to a politically accountable official or body. No specific guidelines for making this determination are provided by the act. Another chairman might opt for a policy which considers the fiscal integrity of the city of paramount importance.

Another factor — "comparables" — is also unspecific and does not guide the chairman's evaluation of the evidence because it implicitly accords him complete discretion in determining what are "similar services" and what are "comparable communities". While the chairman may accept the parties' stipulations in this regard, two chairmen confronted with the identical evidence could either give it controlling weight or exclude it from consideration, depending upon the readiness of each to concede its validity as a basis for comparison.

A striking illustration of the pliability of this criterion appears in the successive opinions of the chairman in the Act 312 arbitration between the City of Detroit and the Detroit Police Lieutenants and Sergeants Association (see Part V, infra), the COLA award in which was central to decision in this case. The chairman's original opinion and award on the wage issue stated in its entirety:

"The Association last offer on salaries for members of the bargaining unit is reasonable and not out of line with other settlements in other police departments in the comparison cities. It is also not out of line with private sector settlements during the past year. No persuasive evidence was offered to the panel to warrant that the settlement to which the parties agreed in the 1974-77 contract not be continued during the 1977-80 contract.

"The panel adopts the Association wages proposal." Detroit v Detroit Police Lieutenants Sergeants Ass'n, Opinion and Award, p 25 (Docket No. 64065).

Following remand by the circuit court for supplemental findings relating the record evidence to the applicable § 9 factors, the chairman in effect disavowed his earlier reliance upon wage settlements in other cities. Although the parties had stipulated that certain cities would be regarded as comparable, the supplemental opinion and award gave overriding importance to maintaining the real level of compensation provided by the previous, negotiated agreement and virtually dismissed all evidence on "comparables" introduced by both sides:

"[T]he panel finds additionally that there has been no evidence submitted as to what Sergeants and Lieutenants do in those 'comparable' communities.

"Since the panel finds that the surveys, without relevant testimony regarding the function of those persons represented in the surveys, are less than persuasive with regard to whether increases in wages should be granted on the basis of those surveys, the panel must place great emphasis on the fact that the prior collective bargaining agreement was arrived at by mutual agreement and a meeting of the minds. The panel finds that with regard to those surveys the fact that the prior agreement was mutually agreed to by the parties is most relevant when considering the comparable data submitted."

The chairman thus initially indicated considerable reliance on "comparables" in reaching his decision only to dismiss the same evidence in a later opinion because no testimony established the functions of lieutenants and sergeants in cities the parties had agreed to regard as comparable. This change of position exposes the act's failure to give meaningful guidance or assure a measure of consistency in the reasoning of the same arbitrator, let alone different arbitrators.

The remaining § 9 criteria are perhaps less protean but still incapable of imparting direction to the process of arbitral decision-making. Factor (a), the lawful authority of the employer, is often developed in terms of the city's ability to raise revenues and thus becomes ancillary to the financial ability component of factor (c). Factors (e) and (f) — cost of living and present overall compensation — are clear enough in reference but not in significance. Must wages keep pace with inflation? Is COLA legislatively encouraged? May an arbitrator award a city offer calling for a pay cut? Factors (b) and (g) — stipulations of the parties and changes in any of the other factors during the pendency of the proceedings — depend on the particular course of proceedings for their content, and factor (h) is open-ended, leaving room for consideration of any other factor "normally or traditionally taken into consideration in the determination of wages, hours and conditions of employment" through collective bargaining, mediation or arbitration (and for the exercise of broad arbitral discretion in determining whether a factor fits that description).

The point is not that the criteria of § 9 fall short of being "as reasonably precise as the subject matter requires or permits". The Legislature need not attempt to identify every circumstance that might bear upon resolution of a police or fire fighter bargaining impasse or to provide a universal formula for weighting the various factors and combining them into a single grand calculation that will compel the chairman to the correct result. The point is that Act 312 arbitrators make legislative-political decisions on an ad hoc basis within a framework which, although it directs their attention to several specific considerations, does not provide for consistency of decision maker, decisional principles or fundamental policy, or for the development of a body of precedent from which policy may be deduced.

2

The judicial review provided by Act 312 is not a safeguard against the improper exercise of delegated policy-making power. The scope of judicial review provided in § 12 seeks only to confirm that the panel's decision is supported by "competent, material and substantial evidence on the whole record".

Justice WILLIAMS would require that the chairman's finding and opinion reflect consideration of all applicable § 9 factors. But such a requirement provides no assurance that decisions made with reference to the § 9 "standards" will evidence consistent use of intelligible principles of broader application than the particular case.

As our colleague's opinion in this case illustrates, the act does not provide for meaningful judicial review of substantive questions. The statutorily prescribed standards are so general that any reasonable decision can be justified. In the ten years that Act 312 has been operative, not one decision of an Act 312 arbitrator has been reversed by the Court of Appeals or this Court on a substantive question.

Given the likelihood of record support for any reasonable decision the chairman might reach, affirmance of the award is assured unless the chairman has considered evidence outside the record or employed decisional criteria not contemplated by the statute. As we said in Dearborn Fire Fighters:

"Most disputes in public employment will present a wide range of reasonable alternatives each of which is supported by 'competent, material and substantial evidence on the whole record.' Providing for judicial modification of those few decisions not so supported is not a substitute for political accountability and review of the choice between reasonable alternatives."

One may ask how even arbitrariness can be detected if the rules of decision-making are unknown. Courts ordinarily determine arbitrariness by comparing official conduct with some objective notion of how a set of principles should be applied to a given case. If assigning weight to the § 9 factors and determining how they interact in a given case rests in the judgment of the chairman, any choice he makes between two positions having evidentiary support is, as a practical matter, final. For example, a chairman may be able to justify awarding a cost-of-living allowance in any case where testimony is offered that such an adjustment is fair and necessary to prevent employees from losing ground to inflation, however compelling the evidence the employer produces on the other factors.

Moreover, if a chairman's opinion fails adequately to explain his award in terms of the § 9 criteria, a circuit judge to whom such an award is appealed is likely to remand the matter for amplified findings of fact and explanation of the chairman's reasoning rather than reverse outright. Assuming that the prevailing party did not utterly fail to make a record, the arbitrator can, by selective reference to the evidence and the § 9 criteria, inflate his initial conclusion like a balloon: the space it consumes expands but the message it bears remains the same.

Ordinarily, the opportunity for judicial review provides a measure of protection against arbitrary decisions which would apply different rules in the same situation. However, the decisions of Act 312 arbitrators are not required to be published, which prevents the accretion of a visible body of precedent. Moreover, while Act 312 arbitrators are required to make findings of fact, they have not been required — although this Court might impose such a requirement in the construction of the act — to separate the policy element from fact finding.

In all events, judicial review for non-uniformity in policy is not practicable. Even if such decisions were to be published, since no arbitrator's decision is more important than another's, a court confronted with non-uniformity could not resolve the conflict. It would not be for the court to make the policy judgment which of the two conflicting policies should prevail.

In another vein, we find a practical, if not a logical, inconsistency in a mode of judicial review which refuses to uphold the award of a hardship exemption to the residency requirement because the panel did not consider "comparables" although neither party had offered comparisons, yet refuses to remand the award on economic issues although the panel excluded clearly relevant evidence which indicated that choosing the DPOA's, rather than the city's, last offer on economic issues would cost an additional $24,000,000 because of parity agreements with the Detroit Fire Fighters Association and the Detroit Police Lieutenants and Sergeants Association. In considering the hardship exemption issue, it is said that the panel must solicit evidence from the parties if they initially fail to introduce evidence on an applicable factor, even though their failure to produce evidence may reflect a consensus that the particular factor is inapplicable. In considering the chairman's rejection of proffered relevant evidence, however, the dispositive inquiry on appeal is framed as whether the exclusion of the proffered exhibit "was so egregious as to cause the award * * * to be unsupported by competent, material and substantial evidence on the whole record".

The latter approach implies that as long as the record before the panel contains support for its conclusions, the exclusion of additional evidence, no matter how material, is of no moment. This is a consequence of the latitude accorded the chairman in weighing and considering evidence on the § 9 factors. If the chairman has discretion to ignore properly admitted evidence, a court's power to assign error for the erroneous exclusion of relevant evidence is meaningless.

B

The act provides that "[a]s to each economic issue, the arbitration panel shall adopt the last offer of settlement which * * * more nearly complies with the applicable factors prescribed in section 9".

The chairman thus must choose a party's proposal on each economic issue; he is forbidden to fashion his own award. Both last offers may be far from what the chairman may consider the best result. Nevertheless, his power stops at the point of choosing the better of the two offers. His options may be hopelessly inadequate, but choose he must.

1

Implicit in every decision, even on economic issues, are policy judgments. The judgment must be made whether wages should be increased without regard to increases in productivity or only to the extent of improved efficiency, whether they should be increased to compensate for all increases in the cost of living or only a portion or not at all. The relationship between wage increases and fringe benefits, pensions, vacation and sick leave, holidays, hours of employment and tasks assigned must also be determined.

In making a last best offer the union and city both enunciate implicitly the policies which they advocate with regard to these judgments. In requiring the chairman to pick the offer which "more nearly complies with the applicable factors", the Legislature in effect forces the chairman to adopt a policy which may be non-uniform as compared to the policy applied to similarly situated communities. The likelihood of non-uniformity is exacerbated by the last-best-offer feature because the chairman cannot formulate his own policy but must choose between the policies advocated by the parties.

There can be no administrative policy when the administrators cannot fashion the rules. The vitality of any policy that a chairman might develop is in the parties' hands. If they by chance learn of the policy and fashion their offers in accordance with it, the policy stands a chance of having some life. If the parties can discern no policy or feel safe in rejecting it because they are certain the other side will also, the policy is for naught. If there is no discernible policy, the parties have inadequate information to guide them in fashioning their bargaining strategy or devising a last offer that they can expect the chairman to find reasonable and adopt except what they can discern of the individual chairman's personal preferences.

If there is no policy each individual chairman will decide not only which offer is best but also what is the measure of "best". If there is no known policy by which to evaluate an award, there can be no effective judicial review. In sum, where the chairman has no power to fashion the award but must choose between the parties' last offers, there can be no principled development of policies to govern in all cases as each case is shaped and determined by the parties' last offers.

2

Last-offer arbitration further reduces political accountability. It tends to reduce the responsibility of the chairman who no longer is required to announce the result which he thinks most in accord with the statutory factors declared by the Legislature or the particularized policy choices implicit in whichever decision he makes. He is required, rather, as to economic issues, only to announce the result which "more nearly complies with the applicable factors." Having been relieved by the Legislature of responsibility for fashioning the best award, he need take no real responsibility for the result. The chairman can, as did the chairman in the instant case, disassociate himself from the result which he himself announced, by intimating that if he were not restrained by last best offer he would reach a different result, presumably lying between the last best offers.

It is the shift of responsibility from the chairman to the parties which advocates find attractive about final offer arbitration:

"Under a conventional procedure the award rarely resembles either party's position but is a compromise between the two. As a result, both parties may respond to constituent complaints that any dissatisfaction with the award should be directed at the arbitrator. In contrast, the final offer arbitrator selects one party's offer in toto after comparing it with the other party's offer, offering each side considerable opportunity to influence the outcome. This influence opportunity tends to increase each party's responsibility for the final outcome compared to conventional arbitration."

It is unrealistic to argue that there is any political accountability when not even the chairman of the panel need take responsibility. Responsibility for the public policy implicit in the award has moved even further from the Legislature. Under the present system it falls somewhere between the parties who fashion the last offers and the chairman who chooses between them.

3

The most striking characteristic of last-best-offer arbitration is the elimination of the most appropriate result as an option. It is excluded unless it coincides with one party's offer.

Putting delegation issues aside, it is a question of some moment whether the Legislature can constitutionally deprive itself, or its delegate, of the authority to make the best public policy.

Could the Legislature appropriate $100 million for revenue sharing to be allocated between urban, suburban and rural communities on a last-best-offer system, the purpose being to encourage the competing governmental units to reconcile their differences?

Could the Legislature announce that it was hearing proposals for a new criminal code and that it planned to enact the last best proposal on each controversial issue, the purpose being again to encourage the contending forces to be "reasonable" and thereby achieve a consensus that would have broad public support?

We are inclined to the view that the Legislature could not constitutionally take such action because the legislative power is the power to make the best laws which can result from the legislative process. The Legislature cannot legislate itself out of business before the fact. It cannot provide for law by lottery, the result to be determined by how close the interested parties approach the resolution the Legislature would pick if it were making the decision. Nor can it authorize an officer or agency to decide upon public policy in this fashion.

It might be argued that in the examples the Legislature denies itself the ability to make the best possible law without effecting a corresponding benefit in so doing. Act 312, however, promotes effective collective bargaining by forcing both parties to present serious settlement offers lest the last offer of the other side be chosen, and thereby increases their incentive to reach their own agreement and the likelihood that they will succeed.

The Legislature is indeed authorized to encourage collective bargaining, but peaceful resolution of collective bargaining disputes is not mandated by the Constitution. The constitutionally conferred authority to provide therefor is permissive rather than obligatory and does not override other constitutional precepts concerning the manner in which governmental power may be exercised and governmental decisions may be imposed.

Further, the efficacy of binding arbitration as a method of resolving collective bargaining disputes does not depend on the last-offer procedure. Conventional binding arbitration operated in this state for several years, and other jurisdictions provide for binding arbitration with the arbitrator free to fashion the award. Even last-offer enthusiasts argue that it is an improvement over conventional binding arbitration, not that it is an essential feature.

4

The last-offer system yields results which are unconstitutionally arbitrary.

While it may be sound public policy to encourage the resolution of labor disputes before the arbitration stage, once that stage is reached, the Legislature cannot constitutionally penalize the party who, although proceeding in good faith, presented a final offer which in the chairman's judgment did not as nearly satisfy the § 9 criteria as did the opponent's. The chairman's inability to fashion the most appropriate award and the need to pick one party's proposal which, although fairer, is still likely to err in the prevailing party's favor, is apt to effect a costly penalty.

The need to accept a city's wage offer, because more reasonable than the union's, rather than award the most appropriate wage could mean a significant difference in an employee's standard of living. The need to accept a union's COLA proposal because more nearly in compliance with the applicable factors than the city's could cost the city millions of dollars. When the city loses, the penalty is on the taxpayer — more taxes or reduced services.

In the instant case the dollar difference in the cost of the union's and the city's last best offers on the issues of wages and COLA is substantial. The Lieutenants and Sergeants Association case and the Fire Fighters case (which on the principle of parity will be governed by the disposition of this case) contribute to that difference. Had the chairman been free to announce the result which he regarded most in accord with the factors delineated by the Legislature, rather than forced to choose the last best offer which more nearly accorded with such factors, the total cost of the award may well have been substantially less.

Unless one party's offer duplicates what would be the most appropriate award, the losing party will always incur a penalty. The size of the penalty is arbitrary, having no relation to the good- or bad-faith quality of the bargaining or the losing party's last offer. It is determined by the reasonableness of the winning party's last offer; the closer that offer is to the most appropriate award, the less the penalty, and the less reasonable, the greater the penalty.

Having entered the field of public employee dispute resolution to provide for binding resolution having the force of law, the Legislature cannot constitutionally provide for so arbitrary a solution and penalty. While peaceful resolution of public employee disputes is an important value, something less draconian is required.

Peaceful resolution of judicial controversies has, with clogged court calendars, become a social goal. We have moved from the pretrial conference to mediation to arbitration and still other alternatives are being considered, all designed to avoid a trial on the merits. If the parties cannot agree on a settlement the case goes to trial. A jury, like an arbitration panel, hears evidence and receives documents. Suppose the Legislature provided that plaintiffs and defendants in personal injury, product liability and medical malpractice actions would each make a last best offer of settlement and that after the jury's verdict was announced the offer nearest the jury's verdict would become the final verdict. If the defendant's last offer is $100,000, the plaintiff's $500,000, and the jury's verdict is anything between $1 and $300,000, the plaintiff would receive $100,000. If the jury's verdict is over $300,000, even $750,000, the plaintiff would receive $500,000.

Such a system could not survive constitutional challenge. Government action, whether legislative or adjudicative, cannot be arbitrary even when there is an arguable rationale for the arbitrariness — the promotion of private settlements by penalizing those who did not try hard enough to settle.

C

The 1976 amendments to the process for selecting a chairman, made following Dearborn Fire Fighters, have not provided any significant measure of political accountability in the arbitral process.

The MERC was directed to establish and appoint a panel of arbitrators composed of "impartial, competent, and reputable citizens" of the United States and residents of Michigan, who "qualify by taking and subscribing the constitutional oath or affirmation of office", serve for indefinite terms, and are subject to removal by the commission without cause. Within seven days of a request from one or both parties for initiation of arbitration proceedings, the commission submits to both parties the names of three persons selected from its panel of arbitrators to be nominees for chairman of the arbitration panel. Each party is authorized to peremptorily strike one of the names within five days, and the commission then designates one of the remaining nominees as chairman of the arbitration panel.

Justice WILLIAMS finds "as a practical matter that the act as now amended sufficiently provides for public responsibility and accountability" because these modifications of the selection procedure "have greatly altered the atmosphere of accountability surrounding the service of arbitration panel chairpersons". We conclude that, while the set of persons eligible to become the chairman of an arbitration panel may in theory have been narrowed and the likelihood that some arbitrators will chair a number of panels increased, the alterations in the chairman selection process do not supply the necessary accountability because it remains impossible to assign meaningful responsibility to any public official or authority for the manner in which a chairman exercises delegated power.

1

Constituting a continuing panel of arbitrators under the auspices of MERC does not localize continuing responsibility for the decisions of arbitration panels or establish a readily apparent link between the arbitrator and politically accountable elected officials. Arbitrators continue to be selected to determine the terms of contracts on a case-by-case basis, and their tenures as decision makers last only so long as the individual disputes for which they were selected remain unresolved by agreement or award.

The panel is large enough to allow assignments to be dispersed among a large number of persons. Membership on the panel is no guarantee of nomination or selection, and as before, all members have "non-public" occupations, often as grievance arbitrators in the private sector. It is entirely conjectural whether the new method of selection motivates arbitrators to conduct themselves so as to enhance their chances of continued employment in the resolution of Act 312 disputes. It is equally likely that most members of the panel are fully occupied with other endeavors and are indifferent to whether they receive additional Act 312 appointments. Moreover, since the act places a premium on expeditious resolution, availability may be a key factor in determining which arbitrators are nominated and selected at any particular time.

It is also argued that the state residency and oath-taking requirements, as well as the indefinite term of membership on the panel of arbitrators — subject to removal without cause — promote accountability by encouraging arbitrators to act responsibly. The chairmen of Act 312 panels before and after the 1976 amendments were Michigan residents. Oath taking is more form than substance under the circumstance that the arbitrator must be "impartial, competent, and reputable". It is most unlikely that any arbitrator would be removed from the panel because of a particular decision. The decisional scheme of the act insures that few awards can objectively be labeled irresponsible, for evaluation of the evidence in light of the § 9 decisional criteria is largely a matter of individual judgment. Moreover, the last-offer system narrows the arbitrator's role in deciding economic issues to choosing between two alternatives, either of which will be a legitimate choice on most records.

2

In all events, Justice WILLIAMS' arguments misconceive the issue. Even if the chairman's sensitivity to the impact of his decisions is heightened by his awareness of personal consequences, that circumstance provides no safeguard against non-uniformity of policy and lack of political responsiveness.

A chairman may not even be aware of the policy content of his decision. The public is even less likely to appreciate that policy decisions concerning its governance have been made when an Act 312 decision is announced. The public sees, rather, that a highly charged controversy with conflicting claims and factual allegations has been resolved one way or another. The arbitration decision seems more like a jury verdict than a declaration of policy subject to legislative oversight and rectification.

It is said that the revised selection system produces a "high order of political accountability" because all members of the panel of arbitrators have been approved and appointed by MERC, whose three members are, in turn, appointed by the Governor. It is implausible that public reaction to a decision made by one of some 100 or more panelists bearing the imprimatur of a three-member gubernatorially appointed commission will be directed to the Governor. Revision of the selection process has not altered the reality that the members of MERC and the Governor do not regard themselves, and are not generally regarded by the citizenry, as responsible or accountable for the decisions of the panel chairmen. Indeed, as we observed in Dearborn Fire Fighters:

"Political intervention would be regarded as an unwarranted intrusion on the independence of the 'impartial' decision maker. It would be regarded as improper for the chairman of the MERC or higher governmental authority to attempt to influence the arbitrator/chairman."

The commission performs its statutory duty by empanelling "impartial" arbitrators. When an impartial arbitrator resolves the dispute, the people of the state, if they look to the Governor or Legislature at all, can hardly complain that a selection process with a large element of unpredictability did not yield a chairman who would have reached a different result.

The connection between the chairmen and the Governor or the Legislature is too attenuated to provide any measure of political accountability for the arbitrators' decisions.

D

We expect independence and impartiality in a judge or grievance arbitrator. But an Act 312 arbitrator does not decide grievances arising under an already established agreement. Rather, acting for the state, he decides and imposes (legislates) the terms of the employment relationship. "What is sound in the exercise of judicial power and the quasi-judicial power of the grievance arbitrator, when applied to interest arbitration in the public sector, is not consonant with a core concept of a representative democracy: the political power which the people possess and confer on their elected representatives is to be exercised by persons responsible (not independent) and accountable to the people through the normal processes of the representative democracy."

The MERC panel of arbitrators is composed largely of persons with grievance arbitration experience. The Legislature is undoubtedly within its prerogative in creating a system which allocates to persons with that experience, belonging to that fraternity, and who are professionally committed to the collective bargaining concept, the power to resolve public policy concerning the terms and conditions of public employment. But the persons chosen to make those decisions for the state may not be isolated from the political process. They may exercise such state power only within a structure which elevates the public interest to predominance, requiring that those who act for the state identify and align themselves primarily with that interest to the exclusion of competing concerns, and which provides means by which the political process may effectively influence the policy-making process and change the policies decided upon and imposed by them as functionaries of state government.

It is expected that government officials, while in office, "will develop a quickening sense of their public responsibility transcending insular concerns and that they will eschew outside endeavors which may be or appear to be incompatible with their public responsibility".

Chairmen, only temporarily in office, all have outside occupations. Most accept employment from labor unions and employers as grievance arbitrators. "That is not wholly compatible with development of the kind of solitary concern for the public interest which the citizenry properly expects of its public officials."

An arbitrator whose primary professional experience and skill has revolved around the identification and accommodation of the interests of two private parties cannot be expected to focus predominantly upon the public interest, subordinating the special interests of the parties, where his Act 312 service is occasional and his primary professional activity remains as before.

E

We address additional arguments advanced in favor of the constitutionality of Act 312.

1

It is argued that the provision of the constitution authorizing the Legislature to "enact laws providing for the resolution of disputes concerning public employees" constitutes a "vast" grant of authority to the Legislature to establish whatever mechanism it might deem appropriate for the resolution of public employee labor disputes. It is noted that at the 1961 Constitutional Convention a proposed amendment to delete the operative language of the forerunner of this provision, offered by a delegate who specifically stated his opposition to "laws to settle the disputes beyond the 2 parties interested, so that * * * the 2 parties have to listen to the third party for a solution", was defeated. It is contended that, because it was the intent of the Convention to leave the choice of procedures to future Legislatures, the Legislature and not this Court is the appropriate body to determine whether the delegation of power to the chairmen of Act 312 panels is an appropriate allocation of authority in a representative democracy.

In our form of government, the power to determine whether acts of the Legislature comport with the principles enshrined by the people in a constitution is confided to the judicial branch. The record of the 1961 Constitutional Convention does not indicate that the delegates, even if they specifically intended to authorize compulsory interest arbitration, intended to approve any particular method of implementing such arbitration, including the appointment of ad hoc arbitrators on a dispute-by-dispute basis. Nor does the inclusion of art 4, § 48 in the Constitution indicate that the delegates or the people intended the grant of authority to be so expansive that measures for the resolution of public employee labor disputes could be adopted without regard to other constitutional precepts. It does not disparage the power granted the Legislature by Const 1963, art 4, § 48 to require that that power be exercised in accordance with other constitutional principles, explicit or implicit.

2

It is claimed that to declare Act 312 unconstitutional because the structure does not provide for political accountability would be inconsistent with the constitutional authorization of similar dispute settlement mechanisms in Const 1963, art 11, § 5, establishing and declaring the powers of the Civil Service Commission. It is said that the commission is more conspicuously lacking in political accountability than the Act 312 arbitrator and that this Court's decisions upholding the broad constitutional grant of authority to the commission require us to uphold the delegation of legislative power effected by Act 312.

Const 1963, art 11, § 5 establishes the Civil Service Commission, prescribes the number, method of appointment and tenure of its members, and enumerates its duties, powers and obligations. The entire framework is constitutional because embodied in the Constitution. In contrast, Const 1963, art 4, § 48, relied upon as the source of constitutional authorization for Act 312, simply permits the Legislature to "enact laws providing for the resolution of disputes concerning public employees, except those in the state classified civil service". No particular framework for the resolution of such disputes is constitutionally approved, nor is it declared that the Legislature may exercise this power without regard to other constitutional principles.

Moreover, certain features of the constitutional charter for the Civil Service Commission endow that body with a measure of political accountability and subject its policy-making powers to limited legislative review.

The commission consists of four non-salaried persons, "not more than two of whom shall be members of the same political party, appointed by the governor for terms of eight years, no two of which shall expire in the same year". It is asserted that, although the commissioners are appointed by the Governor, their extended tenure renders it unlikely that they will be responsible or responsive to him; indeed, the concept of a bi- or multi-partisan body whose members serve staggered and substantial terms is designed to insulate the commissioners from, not expose them to, political accountability.

It may be that few commission members are politically accountable in the sense that they are motivated by the desire to enhance their prospects of reappointment for a second 8-year term (although such a situation is hardly inconceivable).

However, the commission's continuing responsibility for state personnel policy promotes consistency in the determination of that policy and encourages a commissioner's development of "a quickening sense of [his] public responsibility transcending insular concerns" and of "the kind of solitary concern for the public interest which the citizenry properly expects of its public officials". The commission has also exercised its constitutional power to "make rules and regulations covering all personnel transactions" to promote uniform treatment of comparable individual cases.

Finally, Const 1963, art 11, § 5 provides that "the legislature may, by a two-thirds vote of the members elected to and serving in each house, reject or reduce increases in rates of compensation authorized by the commission". Thus, where the fiscal integrity of the state is implicated, the Legislature is empowered to override the policy judgment implicit in a commission recommendation on rates of compensation if two-thirds of the Legislature disagree with that judgment. Even if the Legislature has never exercised this veto power, its inclusion in Const 1963, art 11, § 5 indicates that the provision contemplates legislative oversight and rectification of the commission's judgment on the budgetary component of personnel policy.

3

Finally, it is contended that a 1978 initiatory amendment to Const 1963, art 11, § 5 authorizing collective bargaining for Michigan State Police troopers and sergeants and the submission of issues unresolved through collective bargaining "to binding arbitration for the resolution thereof" "the same as now provided by law for public police and fire departments" amounts to a ratification by Constitution of the Act 312 arbitration scheme.

Compulsory interest arbitration as presently provided by Act 312 poses fewer dangers in the case of the State Police than in the varied cases of municipal police and fire fighters. Because there is only one State Police force, there is no danger that different policies will simultaneously be applied in comparable situations. The arbitrator will be setting a state policy which has statewide implications and which will be funded by the state and not by local units of government.

More pertinently, the 1978 amendment approves the compulsory interest arbitration concept only for the State Police. The Legislature has enacted implementing legislation providing for compulsory interest arbitration of labor disputes concerning only the State Police. Neither the constitutional amendment nor the enactment of this enabling legislation has any bearing on the constitutionality or administration of Act 312. Just as the Legislature is free to modify or repeal Act 312, the constitutionality of that act is to be decided without regard to this amendment which concerns only the State Police.

F

Justice WILLIAMS states that "the majority of jurisdictions which have considered [the accountability question] have found accountability", and that the courts which have ruled otherwise "have done so primarily on the basis of distinguishable constitutional provisions". We read the cases differently.

In support of the first proposition he cites Medford Fire Fighters Ass'n v Medford; Richfield v Local No 1215, International Ass'n of Fire Fighters; Arlington v Board of Conciliation Arbitration and Division 540, Amalgamated Transit Union v Mercer County Improvement Authority. Division 540 is inapposite. The statement in that opinion to the effect that compulsory arbitration schemes are an innovative way to avoid deadlocked labor disputes is not responsive to the accountability argument; clearly it is not a finding of accountability.

As becomes clearer in our colleague's footnote, the other courts did not find accountability; one court misconceived the argument, and the others found that accountability was not required.

The Oregon Court of Appeals did not find accountability in Medford Fire Fighters. The court misconceived the accountability argument, seeing in it only a concern that the arbitrator's decision might be based on his private interests.

The Massachusetts court did not find accountability in Arlington. Instead, it found that the town "fail[ed] to give this argument constitutional content".

Only the Minnesota Supreme Court, in Richfield, said that there was accountability. It concluded that "accountability to the public is like delegation of power; both are a matter of degrees. Although the arbitrators are not directly accountable to the public for their decisions, various provisions ensure the competence and accountability of the arbitrators". The court came to this conclusion in much the same way as does our colleague, referring to the standards in the act (specifically the requirement that the arbitrator consider the financial impact of his decision), the requirement that the arbitrators be qualified by experience, and the removal power of the Minnesota Public Employment Relations Board.

Thus, the courts of only three states — Massachusetts, Minnesota and Maine — have addressed and rejected the political accountability argument made in Dearborn Fire Fighters, and only one of those courts found accountability.

Courts in an equal number of states — Utah, Colorado and Connecticut — have found an unconstitutional absence of political accountability; these courts did not, as our colleague claims, do so on the basis of distinguishable constitutional provisions.

The decision of the Utah Supreme Court in Salt Lake City v International Ass'n of Fire Fighters is not distinguishable. Although, as our colleague observes, the Utah scheme lacked specified standards and safeguards of the sort which would be required by many courts, the court specifically stated that that was not dispositive. The basis of the court's decision was that the "complexities of budgeting and the selection of programs are duties elected officials owe to the electorate; these policy decisions cannot be delegated to a private ad hoc panel of arbitrators in violation of [Utah Const] art VI, § 1 [which provides that the legislative power is vested in a Senate and a House of Representatives]", and that "[t]he power conferred on the panel of arbitrators is not consonant with the concept of representative democracy". Even if the act had provided standards, this constitutional objection would not have been overcome.

Although the decision of the Colorado Supreme Court in Greeley Police Union v City Council of Greeley involved a city charter provision, not a state statute, the court struck down the provision because "[a] contrary holding, in our view, would seriously conflict with basic tenets of representative government". Placing specific reliance on the lead opinion in Dearborn Fire Fighters, the court reasoned:

"Fundamental among these tenets is the precept that officials engaged in governmental decision-making (e.g., setting budgets, salaries and other terms and conditions of public employment) must be accountable to the citizens they represent. Binding arbitration removes these decisions from the aegis of elected representatives, placing them in the hands of an outside person who has no accountability to the public."

Nor is the opinion of the Connecticut Superior Court in Berlin v Santaguida distinguishable. To be sure, the court's recognition that the standards in the act fell far short of those included in the acts of other states, including Michigan, provided one basis of decision. But the court, also relying on the lead opinion in Dearborn Fire Fighters, additionally held:

"The arbitration scheme in question, insulating the arbitration panel from the electorate and representatives of the electorate, contravenes [Conn Const] Article Second, Article Third, § 1 [providing that the legislative power is vested in the legislative branch] and the delegation doctrine set forth in [State v Stoddard, 126 Conn. 623, 628; 13 A.2d 586 (1940), articulating the traditional 'intelligible principle' formulation of the delegation doctrine]. In addition, the concept of politically insulated, compulsory arbitration panels is so inconsonant with the proper exercise of political power in our representative democracy as to violate Article First, § 2 of the State Constitution, which provides in part that '[a]ll political power is inherent in the people.'"

There is, thus, an even division of authority on the accountability question among courts which have addressed the issue.

We recognize that a majority of those jurisdictions which have considered the constitutionality of legislation mandating binding public sector arbitration have upheld such enactments, but most did so without adverting to the issue of political accountability.

Two of these decisions rested upon distinguishable constitutional or statutory provisions.

The decisions of the Wyoming and Rhode Island courts were discussed in Dearborn Fire Fighters.

We are unpersuaded by the reasoning employed by the remaining courts comprising the majority which upheld their statutes ultimately by perfunctory application of a conventional standards and safeguards test.

V

Justice WILLIAMS would affirm the chairman's award on the economic issues contested by the city because he finds the award supported by "competent, material and substantial evidence on the whole record". He states that this statutorily prescribed standard of review requires the order of the panel to be based upon evidence relating to all § 9 factors applicable to the case at hand, and that if an award finds the necessary support in the record, "we are mandated to uphold it — whatever we believe to be its wisdom or its folly".

We have previously indicated that the Legislature has indeed articulated a narrow scope of judicial review of Act 312 awards and that most awards will be adequately supported when viewed in the dim and deferential light in which judicial scrutiny must be conducted.

Nevertheless, we cannot properly approve the instant award because the critical factor in the chairman's decision to award COLA was the award rendered by another panel in the arbitration between the city and the Detroit Police Lieutenants and Sergeants Association (LSA). The validity of that award has not been finally determined because the LSA case is being held in abeyance pending resolution of this case.

The opinion of the chairman of the DPOA arbitration panel declared that COLA was "the central issue in this dispute". The chairman emphasized that, because prior agreements achieved through collective bargaining evidenced a relationship between the provisions obtained by the LSA and those won by the DPOA, and because members of both bargaining units worked together to perform police functions, acceptance of the DPOA's COLA offer was virtually mandated by the award of an identical COLA provision in the LSA arbitration.

In his item-by-item analysis of the § 9 criteria, the chairman of the DPOA panel specifically relied upon the LSA COLA award in discussing four factors.

In light of the "central" importance attached to the LSA award, the validity of the DPOA panel's reasoning in awarding the union's last best offer on COLA hinges upon the validity of the LSA panel's award on that issue. The DPOA award can be regarded as supported "by competent, material and substantial evidence on the whole record" only if the LSA award was properly arrived at.

The city appealed the LSA award to circuit court and, following affirmance at that level, to the Court of Appeals. The city's application for leave to appeal to this Court from the Court of Appeals decision affirming the award was not filed until November 8, 1979, the date on which we granted leave to appeal prior to decision of the Court of Appeals in the instant case. The city subsequently filed a motion to consolidate the cases on the ground that they related to the same subject and had in common controlling questions of law, but we took no action in the LSA case until four weeks after oral argument in the instant case, at which time we ordered the application for leave to appeal held in abeyance and denied the motion to consolidate as moot.

Speaking with the benefit of hindsight, it appears that we failed to appreciate the extent of the relationship between the two awards, at least on the central COLA issue.

The gist of the city's argument in the LSA case is that a panel is not empowered selectively to determine which of the § 9 criteria are "applicable" or to place greater weight upon one or more criteria upon which evidence has been received at the expense of other criteria also addressed in the record. My colleague's opinion implicitly rejects that argument by holding that the statute mandates consideration of whichever § 9 criteria are applicable but does not evince a legislative intention that each factor be accorded equal weight or deprive the panel of discretion to determine which applicable factors are more important in resolving a particular issue.

Other questions raised by the city in the LSA case, however, remain to be answered, including arguments that: the award is not supported by competent, material and substantial evidence on the whole record; the panel, although paying lip service to the § 9 criteria of lawful authority of the employer, financial ability of the city and wages of comparable employees, in reality disregarded applicable criteria upon which it had received evidence and thus reached its decision in a manner not authorized by the statute; the opinions and awards of the panel do not contain sufficient written findings of fact to satisfy the requirements of § 8 and of due process; and the Court of Appeals opinion upholding the panel's award improperly considered a fact outside the record: because the city's last offer on wages was related by a percentage formula to the compensation fixed for police officers in the DPOA case, the city's last offer on wages ultimately exceeded the union's last offer, which was accepted.

In his original opinion and award, rendered October 18, 1978, the LSA chairman's total discussion and presentation of his decision on COLA read:

"In the panel's opinion, employees under a three-year contract are entitled to a cost-of-living allowance. The panel therefore adopts the Association proposal on cost of living."

The Wayne Circuit Court found this portion of the award insufficient to permit judicial review and remanded the award to the panel for supplemental findings, based upon the applicable § 9 factors and the record, on the issues of cost of living, wages and economic aspects of sick leave. Two days later, the DPOA arbitration panel entered its award, recognizing the LSA remand order parenthetically in a footnote.

On January 24, 1979, the chairman of the LSA panel filed his supplemental opinion and award pursuant to the remand order. The city's brief in support of its application for leave to appeal in the LSA case argues that the "January 24, 1979 Supplemental Opinion and Award * * * is nothing more than a transparent after-the-fact attempt to justify [the] award." [p 21]. The city argues that the chairman's later effort disclaims reliance on comparative wage surveys acknowledged as relevant and influential in the original opinion and conceals rather than reveals the actual basis of the panel's award on economic issues.

The supplemental opinion discusses the wages and cost of living issues together in relation to the § 9 criteria. Close examination of the supplemental opinion, however, reveals little more than an inflated version of the bald conclusion originally stated: the certainty of inflation entitles employees to a cost-of-living allowance. In considering criterion (e) of § 9, the cost of living, the panel concluded:

"[A] cost of living allowance would be necessary to supplement the salary proposals of either of the parties in order to preserve the purchasing power of Lieutenants and Sergeants during the life of this contract. The panel finds that maintaining the purchasing power of affected employees is a legitimate aim of the economic provisions of a labor contract. In any case, the city has failed to provide any testimony that would support the proposition that the purchasing power of Lieutenants and Sergeants should not be maintained.

"Since the city has offered no cost of living allowance, and has proposed salary increases that would significantly reduce the purchasing power of Lieutenants and Sergeants, the panel finds that the cost of living factor supports the Association's economic proposals."

The chairman treated the other applicable factors as follows:

(a) Although recognizing that constitutional, statutory and charter provisions precluded the city from raising taxes or realizing revenue beyond the figures stated in its exhibits and budget, the chairman declared:

"The question presented by this arbitration is what proportion of the total budget will be allotted for wages and benefits for Detroit Police Lieutenants and Sergeants. * * * The panel also recognizes that if city officials can limit the scope of an arbitration panel's authority simply by declaring budget limitations, there often would be no purpose for conducting arbitration proceedings."

(c) The chairman found that "the city did not claim an inability to pay but rather indicated that the present appropriation and budgeted amount for Lieutenants and Sergeants was insufficient to pay salaries, COLA and fringe benefits that were the subject of the union demands". While claiming to be cognizant of the interests and welfare of the public and the city's ability to pay, the panel declared that "the real issue" was whether a cost-of-living allowance approximately identical to that contained in the prior contract should be continued, and that while "the city's financial position and legal authority is indeed a valid consideration * * * other criteria are more important, including cost of living".

(d) In considering the position of public employees in comparable communities, the panel exhaustively reviewed the data submitted by the parties on wages and other provisions in cities stipulated to be comparable, only to reject those figures as a basis for comparison because "no evidence [was] submitted as to what Sergeants and Lieutenants do in those 'comparable' communities". Instead, the panel declared that "the prior collective bargaining agreement * * * arrived at by mutual agreement and a meeting of the minds" was the most relevant comparison in determining adequate compensation for the services rendered by Detroit Police Lieutenants and Sergeants, and concluded a cost-of-living allowance was necessary to maintain that level of compensation.

(f) The chairman adverted to the overall compensation presently received by Lieutenants and Sergeants to emphasize that the previous agreement on wages and COLA had been negotiated and that the Association sought only to include identical provisions in the new contract.

The chairman found factors (b), (g), (h) and (d)(ii) — wages and conditions of those performing similar services in private employment and comparable communities — inapplicable.

The circuit court subsequently upheld the award and the Court of Appeals affirmed.

While we do not wish to be understood as intimating an opinion on the merits of a case not formally submitted for decision, we do not view the validity of the LSA arbitration award as a foregoing conclusion. Surely the Court has the power to give some content to the statutory factors. Nor do we understand how the DPOA award can be upheld before judicial review of the LSA award is completed. To do so would imply that it is immaterial to the validity of the instant award whether another decision repeatedly identified as a "central" consideration in the chairman's reasoning is invalidated. Such an approach confirms the general lack of principled decision-making inherent in Act 312, the emptiness of the judicial review provided by the statute, and the paucity of safeguards attending the exercise of legislative-political power by Act 312 arbitrators.

COLEMAN, C.J., and KAVANAGH, J., concurred with LEVIN, J.