reason, contract and law in labor relations

8/8/2019 Reason, Contract and Law in Labor Relations

http://slidepdf.com/reader/full/reason-contract-and-law-in-labor-relations 1/30

A P P EN D IX A

REASON, CONTRACT, AND LAWIN LABOR RELATIONS*

H A R R Y S H U L M A N

Late D ean and Sterling Professor of Law

Yale Law School

Holmes did not have much occasion as judge to deal withthe organization of labor and collective bargaining. But w henhe did, he stated what he called "the less popular view of thelaw."

1To H olmes, our system of free enterprise and dem ocratic

government required the state, subject to the limitations ofpublic order, to permit workers to organize and to extend their

organization for the purpose of strengthening their bargainingposition in the struggle for a better share in return for theirservices. Th eir shares were to be determ ined by the parties to

the struggle, not by the state; and the state should not interfereso as to make the struggle uneq ual. Th us, when dealing withpicketing to enforce a wage demand, he said in 1896:

One of the eternal conflicts out of which life is made up isthat between the effort of every man to get the most he can for

his services, and that of society, disguised under the name ofcap ital, to get his services for the least possible ret ur n. Co m bi-natio n on the one side is pat ent and powerful. Com bination on

the other is the necessary and desirable co un terp art, if the battleis to be carried on in a fair and equal way.2

* This paper was delivered as the Oliver W endell Holmes Lec ture at H arv ard Law Schoolon February 9, 1955, a little over a month before Dean Shulman's death. It w as printed inthe Harvard Law Review, 68:999 (April 1955), and is here reprinted by permission of thepublisher.

1 Vegelahn v. Guntner, 167 Mass. 92, 104, 44 N.E. 1077, 1079 (1896 ) (dissenting opinion).2

Id. at 108, 44 N.E. at 1081.

169



170 MANAGEMENT RIGHTS AND THE ARBITRATION PROCESS

And when dealing with threatened strikes and boycotts to com-pel employment of members of defendants' union only, he said:[The defendants'] purpose was not directly concerned withwages. It was one degree more remote. The immediate objectand motive was to strengthen the defendants' society as a pre-liminary and means to enable it to make a better fight onques-

tions of wages or other matters of clashing interests. I differfrom my brethren in thinking that the threats were as lawfulfor this preliminary purpose as for the final one to whichstrengthening theunion was a means.

3

Perhaps the social and economic history of the United States

would have been significantly different if in these cases Holmes

were stating the prevalent view rather than speaking in dissent.

His dissenting view did prevail. But it was not until the Na-

tional Labor Relations Act4

became effective that the workers'

freedom of association was safeguarded by the imposition of a

correlative duty onemployers to refrain from interfering with

or restraining the workers' choice. And the workers' right to

collective bargaining was then enforced by an affirmative dutyon the employer to recognize and bargain with the representa-

tive designated by them, onwages, hours, and other conditions

of employment.5

This bare legal framework is hardly an encroachment on the

premise that wages and other conditions of employment be left

to autonomous determination by employers and labor. On the

contrary, it merely establishes the conditions necessary for the

exercise of that autonomy. But, as is perhaps inevitable, the

statement of a legal duty to bargain collectively on wages and

conditions of employment leads to demand for legal definition

of the scope of the du ty. "What m atters are included in the

phrase "con ditions of emp loym ent?" And is a party's legal duty

to bargain satisfied when it persistently demands unilateral dis-

8 P l a n t v. W o o d s , 176 Mass. 492, 505, 57 N . E . 1 0 1 1 , 1016 ( 1900 ) ( d i s s e n ting op in ion ) .

* 4 9 STAT. 449 ( 1 9 3 5 ) , as a m e nde d , 29 U. S . C . §§ 1 5 1 - 6 8 ( 1 9 5 2 ) .5 § § 8 ( a ) ( S ) , 9 ( a ) , 4 9 STAT. 453 ( 1 9 3 5 ) , as amended, 29 U. S . C . §§ 1 5 8 ( a ) ( S ) , 1 5 S ( » )

( 1 9 5 2 ) .



REASON , CONTRAC T, AND LAW IN LABOR RELATIONS 171

cretion on one or m ore items and refuses to concede to the oth erside any voice in their determination?

The questions of what the parties should bargain about andwhat they should leave to unilateral rather than joint determi-nati on could , of course, be left to the parties themselves. Th ey

could decide whether to bargain about pensions or the numberof shifts in the same way that they decide whether to have awage increase or how mu ch of an increase. Th at would involvethe possibility of a cessation of production because of stalemateon these issues; but such an interruption is an integral part ofcollective barg aining . Th e results m igh t the n differ from oneenterprise to another; one might bargain about pensions, theother might not; one might place a matter under unilateral con-trol, the other might make it a matter of joint determination.But such differences would be quite in accord w ith th e p ostulateof autonomous determination through collective bargaining.

In an enterprise in w hich collective bargaining is just m akingits appearance, if the law in its administration surveys the courseof the apparent bargaining and determines that it is apparentrather than real, because of the scope of the demands for uni-lateral discretion, the law may well be merely enforcing theduty to bargain rather than shaping the content of the bargain.But in an enterprise in which collective bargaining is an acceptedand going institution, if t

1ne law commands that some particular

item must be made the subject of bargaining and may not bethe object of a firm demand for unilateral control, then to thatextent the law interferes with the parties' autonomy and shapesthe conte nt of their bargain. Such decisions tend t o become no t

only definitions of the legal duty to bargain but also statementsof the maximum that the parties may in practice seek from oneanother.

Th e law intervenes in another way. Th e parties' bargainingnormally results in a collective labor agreement for a stated termor for an indefinite period. The agreement is made on the un -derstanding and with the expectation that both parties will re-spect it as a com mitm ent b inding upon them . In the business




world such com mitm ents are called contrac ts. An d the collec-tive labor agreement itself comes to be called the contract evenby the workers—as, for example, in the slogan "no contract, no

work." Does it not naturally follow, then, that the law whichprovides remedies for breaches of contract generally should also

provide remedies for breaches of collective labor agreements?If the parties are entrusted by law w ith the responsibility of de-termining conditions of employm ent, should they not be held to

their responsibility and their agreements be given the sanction oflegal enforcem ent? This is the line of reasoning which appar-ently persuaded theCongress to invest the federal courts withjurisdiction over actions for breach of contract between em-

ployers and labor organizations.6

Again, if the collective agreement provides for resort to vol-untary arbitration, it is argued that the law should enforce the

agreement; and provision is made for suits to enjoin or compelthe arbitration or to enjoin or enforce the resulting aw ards. This

limited intervention by the law, it is argued, is not an impair-ment of the freedom of contract but rather a means of makingit effective.

In my judgment, these are unwise limitations on the parties'autonomy. For me this conclusion follows from the analysiswhich I propose to make of the rule of law and reason whichthe parties' contract — the collective agreement — establishes.But the analysis has validity, whether or not you draw the sameconclusion.

While what I shall say may have wider application, my

archetype is a large industrial enterprise employing many thou-

sands of organized workers in one or more plants. It is neces-sary to bear in mind that such an enterprise involves not onlylarge groups of organized workers but also an employer who

acts through many hundreds or thousands of representatives at

various levels of authority from the job foreman through the

superintendents andmanagers to the vice-presidents and presi-

« Labor Managem ent R ela t ions Act, § 301 ( a ) , 61 STAT. 156 ( 1 9 4 7 ) , 29 U.S.C. § 1 8 5 ( a )

( 1 9 5 2 ) .



REASO N, CONTR ACT, AND LAW IN LABOR RELATIONS 173

dent and board of directors. W hile the organization on theemployer side is more monolithic than on the worker side, thefact that many people exercise its authority in various ways isof great practical significance.

Collective bargaining today is not concerned merely with the

return for the employees' services that Holmes talked about. 7

Th at is, of course, one im por tant concern. O n occasion allatte ntio n seems to be focused on it. Bu t wages are negotiatedonly periodically, once in six months, or twelve months, orperhaps even at longer intervals. Even when w ages appear tobe the chief or only matter in controversy, there is a great dealmore involved—more which is not only of at least equal impor-tance but which also affects the wage negotiations.

Collective bargaining is today, as Brandeis pointed out, themeans of establishing industrial democracy as the essential con-dition of political democracy, the means of providing for theworkers' lives in industry the sense of work, of freedom, and ofparticipation that democratic government promises them ascitizens.

8The modern industrial worker is not engaged to pro -

duce a specific result and left to himself for the performance.He is hired to work under continuous and detailed directionand supervision, in close association with hundreds or thousandsof fellow workers, each of whom performs a very minute por-tion of the work that ultimately results in a finished product.The enterprise requires the continuous co-ordination of thework of this multitude or employees; and this poses numerousdaily problem s wh ether or no t the employees are organ ized. Soelementary a matter as leaving the job for a few minutes "to

service the body," as they say in the shop, poses a serious prob-lem which must be carefully analyzed and provided for, other-wise one might find the work of a hundred men held up everytime one of them had to leave. Every day a number of em-ployees ma y be absent or repor t late. Da ily or almost daily

7 Seep.—* See Hearings Before the U.S. Commission on Industrial Relations, S. Doc. No. 415, 64th

Cong., 1st Sess. 991-1 011, 7657-81 (19 14- 15), reprinted in part in BRANDEIS, THE CURSEor BIGNESS 70-95 (Fraenkel ed. 1934).



174 MANA GEMEN T RIGHTS AND THE ARBITRATION PROCESS

some employees have to be laid off for a short period or indefi-nitely; some employees must be hired; changes must be made injob assignments, either by way of promotion or demotion orotherwise. An d daily there are thousand s of occasions for fric-tion between employee and supervisor which may erupt in

disciplinary action against the employee or a stoppage of work.These and a host of similar problems are inherent in the

necessity of co-ordinating the work of thousands of personsinto an efficient o peratio n. Even where there is no union , theemployer needs statements of policy to guide the hundreds ofpersons through whom he must act, though he may be readyto invest them w ith large powers of discretion. Add ition of theunion alters the situation in at least two ways: First, the em-ployees, through the union, must participate in the determina-tions. Second, the acceptanc e of unions and collective barg ain-ing has increased the employee's confidence and his sense ofdignity and importance; where previously there may have beensubmission, albeit resentful, there is now self-assertion.

One might conceive of the parties engaging in bargainingand joint determination, without an agreement, by consideringeach case as it arises and disposing of it by ad hoc decision. Butthis is, of course, a wholly im practical method , pa rticularly fora large enterprise. So the parties seek to nego tiate an ag reem entto provide the standards to govern their futu re action.

In this endeavor they face problems not unlike those encoun-tered wherever attempt is made to legislate for the future inhighly complex affairs. Th e parties seek to foresee the m ul ti-tude of variant situations that might arise, the possible types

of action th at m ight then be available, the practicalities of eachand their anticipa ted advantages or disadvantages. Choice be-tween the suggested possibilities is rendered more difficult bythe very process of bargaining and the expected subsequentadm inistration of the barga in. Th e negotiations are necessarilyconducted by representatives removed in variant degrees fromdirect confrontation with the anticipated situations. They acton the basis partly of their own experience and partly of the



REASON , CONTRA CT, AND LAW IN LABOR RELATIONS 175

more or less incomplete or clashing advice of constituents—the resolutions of councils, subcouncils, unit and departmentalmeetings in the case of the union, and the suggestions fromindividuals at the various levels of management in the case ofthe employer. W hile each area of problems— vacations, over-time, promotions, layoffs, and the like—must be separately and

carefully considered, each is nevertheless but a small part ofthe tota l negotiation. The pressure for trad e or compromiseis ever present.

No matter how much time is allowed for the negotiation,there is never time enough to think every issue through in allits possible applications, and never ingenuity enough to antici-pate all that does later show up . Since the parties earnestlystrive to comp lete an agreemen t, there is almost irresistible pres-sure to find a verbal formula which is acceptable, even thoughits mean ing to the two sides ma y in fact differ. Th e urge tomake sure of real consensus or to clarify a felt ambiguity in the

language tentatively accepted is at times repressed, lest theeffort result in disagreement or in subsequent enforced consentto a clearer provision which is, however, less favorable to thepa rty w ith the urge. "With agreement reached as to k now nrecu rring situations, questions as to application t o m ore difficultcases may be tiredly brushed aside on the theory that those caseswill never—or hardly ever—arise.

Then there is never, of course, enough time to do an im-peccable job of draftsmanship after substantive agreement isreached—apart from the hazard that such an effort might un-cover troublesome disagreement. Th ou gh the subject mat teris complex and the provisions intricate, the language must

nevertheless be directed to laymen whose occupation is notinterpretation—the workers in the plant, the foremen, theclerks in the pay roll office. For it is the y whose actions m ustbe guided by the agreement; and indeed, in the case of theunion, the membership is asked to ratify or reject what is priorto its action only a proposed agreement. W hile the inte rpre ta-tions or explanations made at the membership meetings can



176 MAN AGEM ENT RIGHTS AND THE ARBITRATION PROCESS

hardly bind the employer, it is nevertheless important that theagreement be not such as to become a promise to the ear but adisappointment to the hope of the membership.

To be sure, the parties are seeking to bind one another andto define "rig hts " and "obligation s" for the future . Bu t it is

also true that, with respect to nonwage matters particularly,the parties are dealing with hypothetical situations that mayor may no t arise. Both sides are interested in the welfare of th eenterprise. Ne ither would unashamedly seek con tractu al com-m itme nts tha t would destroy the other. Each has conflicts ofinterests in its own ran ks. Both m ight be con tent to leave thefuture to discretion, if they had full confidence in that discre-tion and in its full acceptance when exercised. An d even whenthe negotiating representatives have full confidence in eachother as individuals, they recognize that it will be many others,not they, who will play major roles in the administration of theagreement. So they seek to provide a rule of law w hich w ill

eliminate or reduce the areas of discretion. The agreem ent thenbecomes a compilation of diverse provisions: some provide ob-jective criteria almost automatically applicable; some providemore or less specific standards which require reason and judg-ment in their application; and some do little more than leaveproblems to futur e consideration w ith an expression of hope andgood faith.

Consider, for examp le, the role of seniority. Specifically, theparties seek to provide for the selection of employees for pro-motion, or for layoff should it become necessary to reduce force;and seniority is urg ed as the touc hston e for the selection. Sen-

iority here means not chronological age, or length of service inthe particu lar industry or indu stry generally. It means rathe rlength of service for the particular employer in the particularseniority unit of th at employer. W hen the union insists th atseniority shall govern promotions to better jobs, it is not becausethe union does not adm it the desirability of recogn izing superiorability and encouraging ambition and greater effort; nor is itbecause the union is unaware that the progress of some of its



REASON , CONTRAC T, AND LAW IN LABOR RELATIONS 177

worthy members will be retarded by a strict seniority rule.Again, when the union demands that layoffs shall be made bystrict order of seniority, it is not because of unawareness of theother factors that might appeal, for example, to the socialwork er. Th e union is well aware that layoff by seniority may

in some instances cause greater hardship tha n w ould be the caseif other factors were considered. It know s that in a particularcase, the senior employee in the given seniority unit may be arelatively young m an with no dependents and wi th considerablemobility that would enable him to find some work elsewhere;while the junior employee may be a relatively old man with alot of service in the industry, many dependents, and littlechance of finding w ork elsewhere. To be sure, there is thegeneral opinion that long service deserves of itself some rewardand preference. Bu t the seniority rules of which I speak applyat all periods of service—to the men with one or two years ofservice or less, as well as to those with ten, fifteen, or more.

And a difference of a day or a week is made determinative inthe selection. Moreover, seniority is comm only n ot a factor indeter mi ning the employee's pay for his wo rk. H e receives therate for the job whether he has been on it a year or ten years.

I suggest that the insistence on seniority, like the insistenceon single rates of pay for specific jobs, is based in large part onthe desire or need for an objective rule which eliminates judg-ment and discretion in particular cases. And that is not merelybecause the union is unwilling to lodge discretion in the em-ployer. I daresay that if the employer were willing to gran t tothe union complete control over promotions and layoffs, the

union would adopt the seniority standards in order to curb itsow n discretion. For the exercise of discretion in these cases is avery difficult task and its fairness or soundness is always subjectto attack— mo re or less violent. W hen it is recognized th at th epurpose is not merely that of rewarding seniority, but is alsoto provide a fair, objective measure which would curtail arbi-trary power or discretion, the range of possible adjustments issignificantly altered.




Now contrast the different matter of discipline. Here, too,

the parties recognize that occasion for disciplinary action willarise and that disciplinary action is something of a necessity.But whereas a reduction of force requires a selection amongemployees which necessarily means preference of some em-

ployees over others, disciplinary action poses no problem ofpreference. The union can generally seek to protect each em-

ployee threatened with disciplinary action without subordi-nating an interest of other employees. Here that protection re-

quires the exercise of fair and humane judgment and discretionwhich takes into account all mitigating factors that can be

mustered for the particular employee. The problem, at leastfo r theunion, is not tha t of eliminating the pains of discretion,but rather that of confining theemployer's power and provid-ing maximum opportunity for the union to challenge the

soundness of his exercise of discretion, while for the employerthe problem is tha t of reducing this vulnerability of the dis-

ciplinary action taken by him. So most collective agreementsdo not go much beyond recognizing the employer's power to

discipline or discharge and providing that the action shall be

for cause, or good cause, and shall be subject to challenge by

the employee and the union, subject to a few limitations or

exceptions.

Th e parties recognize, when they make their collective agree-ment, that they may not have anticipated everything and that,in any event, there will be many differences of opinion as to

the proper application of its standards. Accordingly the agree-ment establishes a grievance procedure or machinery for the

adjustment of complaints or disputes during its term. Theautonomous rule of law thus established contemplates that the

dispute will be adjusted by the application of reason guided by

the light of the contract, rather than by force or power.

While the details of the grievance procedure differ from one

enterprise to another, its essence is a hierarchy of joint confer-ences between designated representatives of the employer and

the union. But joint conferences even at the highest levels of




authority may not, and frequently do not, result in agreement.In the absence of provision for resolution of stalemate, the par-ties are left to their ow n devices. Since grievances are almostalways complaints against action taken or refused by the em-ployer, a stalemate means that the employer's view prevails. Ofcourse, in the absence of some restraint by contract or other-

wise, the union is free to strike in order to reverse the employer'schoice. But th e union can hardly afford an all out strike everytime it feels tha t a grievance has been unjustly denied. Th econsequence is either that unadjusted grievances are accumu-lated until there is an explosion, or that groups of workers, lessthan the entirety, resort to job action, small stoppages, slow-downs, or careless workmanship to force adjustment of theirgrievances.

The method employed by almost all industry today for theresolution of stalemates in the adjustment of grievances underthe private rule of law established by the collective agreement

is private arb itration b y a neu tral person. The largest enter-prises provide for a standing umpire or arbitrator to serve fora stated period of time or so long as he continues to be satis-factory to both sides. The great majority of agreements pro-vide for separate appointment of an arbitrator in each case.And the appointments in any case are made by the parties orby a method agreed upon by them . The wide acceptance ofarbitration as a terminal step in the grievance procedure—ascontrasted with its relatively limited use in the making ofthe contract in the first place—is explained generally on thegrounds, first, that grievances involve interests of lesser impor-tance than those in contract negotiation and, second, that the

discretion of the arbitrator is confined by the agreement underwhich the grievances arise. Both statemen ts require qualifica-tion. As um pire unde r one collective agreement, I have arbi-trated cases ranging all the way from the claim of a single em-ployee for fifteen minutes' pay to that of more than sixty thou-sand employees for a paid lunc h period th e direct cost of whichwas between seven and eight million dollars a year. An d the



180 MAN AGE MEN T RIGH TS AND THE ARBITRATION PROCESS

restraining bonds of the collective agreement are found onoccasion to be elastic indeed.

The parties do not generally restrict their own joint powersin the grievance procedu re. Bu t it is custom ary for the collec-tive agreement to limit the arbitrator's jurisdiction with ap-

pa ren t strictness. A pa rt from th e specific exclusion of certainsubjects, as, for example, rates for new jobs or productionstandards, he is commonly confined to the resolution of griev-ances or disputes as to "the interpretation or application of theagreem ent," or of claims of "violation of the agreem ent." An dquite frequently he is further enjoined not to "add to, subtractfrom, or modify any of the terms of the agreem ent." In the

agreement with which I am most familiar he is admonished alsothat he has "no power to substitute his discretion for the Com-pany's discretion in cases where the Company is given discre-tion" by the agreement, and no power "to provide agreementfor the parties in those cases where they have in their contract

agreed that further negotiations shall or may provide for certaincontingencies."

9

Doubtless these are wise, perhaps even necessary, safeguards—at least before the parties develop sufficient confidence intheir private rule of law to enable them to relax the restriction.And an arbitrator worthy of appointment in the first placemust conscientiously respect the limits imposed on his jurisdic-tion, for otherwise he would not only betray his trust, but alsoundermine his own future usefulness and endanger the verysystem of self-government in which he works. But these arehardly provisions which would be inserted in the agreement to

control the courts in an action on the contr act. The judge, too,mu st decide only "according to law ." Un like the case of thearbitrator, however, the judge's authority and the law whichhe must interpret and apply do not derive entirely from theagreement of the litigants before him.

9Agreement Between Ford M otor Co. and U nited Automobile Workers, CIO , art. I ll, § 21

(1949, 1950).



REASON, CONTRA CT, AND LAW IN LABOR RELATIONS 181

Let me consider some of the difficulties and limitations of thearbi trato r's function . Suppose the collective agreement is com -pletely silent on a m atter in dispute. Suppose, for exam ple, tha tthe agreement is silent on the question whether acceptance ofovertime work is mandatory or optional with the employee.This very issue was reported as the cause of the recent extensiveand vexing strikes on the English docks.10 It is an issue which anumber of arbitrators have had to decide under collective agree-me nts. N ow it is easy enough to say that the ma tter is notcovered by the agreement. But wh at follows? May the em-ployer, therefore, require the employees to accept the overtimeassignments on pain of disciplinary penalties, such as layoff ordischarge, or may the employees properly refuse the assign-ments? Answer w ould be aided, of course, if there were acommon presupposition as to the effect of the collective agree-me nt. In constitutional law terms, bu t witho ut pushing themetaphor far, is it a grant of limited powers or is it a set of

restrictions on otherwise unlimited powers? If it is the formerand the employer is not given the power to command overtimework, then his attempt to discipline employees for failure toaccept would be a violation of the contract; if it is the latter,then, since by hypothesis the agreement contains no relevantrestriction, the employer would have the "reserved power" toenforce the command.

Partly for the purpose of meeting this difficulty many agree-ments now include what is generally called a " managementprerogative " clause, sometimes more accurately and tactfullycalled a "managem ent responsibility" or "managem ent func-tions " clause. This norm ally lists certain m atters as "th e soleright" of management or for "sole determination" by manage-ment, subject, however, to such restriction as may be providedin the agreement. Th e inclusion of the manag emen t provisionin some agreemen ts m ay raise a question as to th e significance ofits exclusion in others; and it focuses attention on the precise

10 See N.Y. Tim es, Oct. 17, 1954, p. 22, col. I.



182 MANA GEMENT RIGHTS AND THE ARBITRATION PROCESS

language of the provision with possible reference to the maximinclusio unius est exclusio alterius. Apart from its specificationof items as to which there is normally no question, such as theprod ucts to be ma nufa ctured , the provision is norm ally couchedin broad phrases like " the right to manage the business" orto "direct the working forces." One may wonder about the

chances of the adoption of an agreement, in some enterprisesat least, if it states in unmistakable language that the employershall have the right to do anything at all with respect to thework of the employees except as he is expressly limited by theagreement.

Courts, if confronted with this problem, would doubtlessdeclare a general principle, whether or not it squared with theconception of the parties in the particular case. But th e powerof the arb itra tor to do so is at least questionable. Th e obviousalternative is for the arbitrator to refrain from affirmative de-cision and to remand the dispute to the parties on the ground

tha t it is outside of his jurisdiction. But w ould no t th at be ineffect a decision supporting the employer's freedom of action?If the validity of the employer's order requiring the overtimework is beyond the arbitrator's jurisdiction, he would seem tohave no power to restrain the disciplinary action taken by theemployer to enforce the order. On th e other hand, if he doesrestrain the disciplinary action, is he not in effect denyingvalidity to the employer's order? Again, the denial of jurisdic-tion presum ably leaves the dispute for resolution by the p arties.But whether the union may properly resort to economic pres-sure in the effort at resolution may depend upon the construc-tion of the "no strik e" provision of the agreement. The obliga-

tion not to strike may or may not be coextensive with the arbi-trator's jurisdiction.

The question of fundamental presuppositions arises in an-other way. The parties rarely start with an enterprise fromscratch; generally they negotiate an agreement for a goingenterprise w hich has been in operation for some time and whichhas developed practices or precedents of varying degrees of con-




sistency and force. W ha t is the significance of th e claimed"prior practice"?

For example, in the overtime case we have been consider-ing, suppose that evidence is tendered that the employer neversought to compel acceptance of overtime assignments, or that

the employees never refused such assignments without goodexcuse. Or suppose that , thoug h the agreement is silent on thematters, the employer had been giving the employees a restperiod of ten m inutes in each half of the shift, or a lunch periodon the employer's time, or a five minute wash-up periodbefore lunch or at the end of the shift, or a money bonus atChristmas. Or , to vary the nature of the example, suppose theclaim is that it had been customary for the employer to assign arigger to assist pipefitters when they were required to lift pipeof four inches or more in diameter, or to assign an employeeto hold the pieces which a welder had to weld. N ow suppose

that, during the term of the agreement, the employer changesthese claimed practices over the union's strenuous objections,which are then carried through the grievance procedure to theum pire . In these cases it is the u nion w hich relies on the prio rpractice. Bu t frequently the position is reversed. For example,an employer directs a punch press operator to paint his presswhen he has no punching to do; or he asks a crib attendant topain t the walls of his crib . In either case, the employee refuseson the ground that painting is not work in his classification,bu t rather in tha t of a painter. And th e employer points to aclaimed prior practice in accord with his direction.

Again the fundam ental question may be asked: Is the agree-

ment an exclusive statement of rights and privileges or does itsubsume continu ation of existing conditions? And again it maybe ventured that courts, if confronted with the question, wouldprobably give a general answer for all cases. For the arbitrator,particularly if his jurisdiction is limited to "interpretation"with a prohibition against "adding to, subtracting from ormodifying" the terms of the agreement, a general answer is notso clear.




Some have urged that established practices, at least if theywere in existence at the time of the negotiation of the agree-ment and were not considered in any way during the negotia-tions, are binding upon the parties and must be continued forthe dura tion of the agreem ent. This, it is said, is implied in th e

agreement itself—or in the "logic" of the agreement or in thecollective bargaining relationship. Lawyers are familiar w ith"im plied" terms. W e used to differentiate between imp lications"in fa ct" and implications "in law ." N ow scholars say the dif-ferentiation is not quite valid and the implication in any eventis based on morality, common understanding, social policy, andlegal du ty expressed in to rt or quasi-con tract. The com monunderstanding of the litigants in the particular action is onlyone factor in the implication—and not the most important.But the judges' authority for imposing the implication is notthe party's will; it is the superior authority of the law, whichtranscends the party's will.

The arbitrator of whom we are talking does not have suchsuperior auth ority to impose implied conditions. Th e implica-tions which he may find are only those which may reasonablybe inferred from some term of the agreement. Is there an im-plication "in fact" in the collective agreement that existingpractices must be continued until changed by mutual consent?It may be said parenthetically that the legal duty to bargain isnot quite relevant because, apart from the question whether thearbitrator may enforce that duty, the issue is whether the prac-tice may be changed without mutual consent when bargaininghas failed to achieve consent.

It is more than doubtful that there is any general understand-ing among employers and unions as to the viability of existingpractices during the term of a collective agreement. Ther e maybe some agreements which are negotiated upon a real or tacitassumption of con tinuance of existing practices except as mod i-fied by the agreement. The re are certainly some agreementswhich specifically provide for the continuance of existing prac-tices with va riant limitations. Bu t I venture to guess that in



REASON , CONTR ACT, AND LAW IN LABOR RELATIONS 185

many enterprises the execution of a collective agreement wouldbe blocked if it were insisted that it contain a broad provisionthat "all existing practices, except as modified by this agree-ment, shall be continued for the life thereof unless changed bym utu al consen t." An d I suppose tha t execution would also be

blocked if the converse provision were demanded, namely, that"the employer shall be free to change any existing practiceexcept as he is restricted by the term s of this agreem ent." Th ereasons for the block would be, of course, the great uncertaintyas to the nature and extent of the commitment, and the relent-less search for cost-saving changes. The larger the enterprise, themore varied its operations, the more dependent it is on techno-logical change, and the keener the competition the greater thisun cer tain ty and search. The agreement between BethlehemSteel and the Un ited Steel wo rkers steers a midd le course. Itprovides that if management changes any local practice or cus-tom, an affected employee may file a grievance and in "the

disposition of the grievance the burden shall be on Managementto justify its action."

uThe agreement does not state, how-

ever, wh at is to cons titute justification. Th at little question isleft to future judgment.

Assuming the prior practice to be at least relevant, we mayfind ourselves in further troub le. I have spoken of the p rac -tice as an ascertained or readily ascertainable matte r. Bu t co m-mo nly it is only a question. Com mo nly there is widely conflict-ing evidence as to wh at was in fact done in the past. Ascertain-ing the fac ts w ith re spect to an alleged pract ice is a difficult tasknot suggested by the assurance implicit in the word "practice."

Nor is it a task which can fortunately be cast on the broadshoulders of a jur y. Bu t even after the facts are ascertained,wh at is their significance? W hen do they add up to a prac -tice? And what practice?

Suppose tha t in the pipefitters' case, the em ployer says: "Surewe've used a rigger in the instances cited. Bu t we did that be-

1 1Agreement Between Bethlehem Steel Co. and U nited Steelw orkers, CIO , art. I I, § 3

(1952).



186 MANA GEMEN T RIGHTS AND TH E ARBITRATION PROCESS

cause we had a rigger available with free time and used him toexpedite the work. "We still do that . But w e never had anynotion that we would supply a rigger in other circumstances orthat the pipefitters can't be required to work without him."Or tak e the Christmas bonus. The employer says: "O f coursewe've paid the bonus. W e did it in our discretion when w ethought we could afford it and accomplish some good for ourbusiness. This year we are convinced that we cannot affordthe bonus and, in any event, th at it will do us no good." Orconsider the union's reply to the company's claim that cribattend ants always painted their cribs: "Sure they have. Bu tthat was their individual choice—not a collective determina-tion. Th e union is not o ut to stir up trouble . So long asnobody objected, we did not look into the question. But whena crib attend ant did object, we then took our position. An d wesay that the attendant has the choice of accepting or rejectingthe assignment." Such are the limitations comm only claimed

for alleged practices, and their reality cannot be gainsaid merelybecause they were not recorded at the time or communicated tothe other side. One cann ot acco mp any his every act in thecourse of a busy day with explanations which would avoidprejudice for the future.

I have been discussing situations where the agreement is silenton im po rtan t phases of the parties' dispute. Bu t frequently thesilence so assumed is a conclusion as to the very question in dis-pute. Generally one or the other of the parties urges stronglytha t, while the agreement may not speak to the issue directly, itspeaks to it indirectly but clearly.

A fairly common recurring dispute relates to the employ-ment of independent outside contractors to do work which hasbeen or can be done by the employer's own em ployees. Forexample, an employer may decide to engage an independentoutside painting contractor to paint the plant, though he haspainters in his own w ork force. Or he may decide to employan outside contractor to make an electrical installation in theplant though he has his own electricians available for the work.




The fact that some of his own employees may be on layoff whilethe outside contractor is working aggravates the situation, but

is not necessarily controlling on the issue of interpretationinvolved . Th e emp loyer's defense of his action in these casesnormally runs along these lines: He contends that the determi-

nation whether to have particular work done by his own em-ployees or by an outside contractor is part of his reserved

"prerogative" which is either unrestrained by the agreement orrecognized in the agreement by a provision of the kind men-tioned above, leaving to him the "management of the business,"the choice of "products to be manufactured," "the schedules ofof production," the "direction of the working forces," and the

like. An d he ma y add, wit h or wi tho ut full disclosure of thesupporting evidence, that he chose to engage an independentcon trac tor for reasons of econom y and business expediency.

The u nion's reliance is on the agreem ent. It points to the sec-

tion, normally called recognition, which usually states that theemployer recognizes the union as the collective ba rgaining agentfor his employees in stated categories of work, such as produc-tion, or maintenan ce, or shop clerical and the like. This means,it argues, that work of the stated categories must be done by

employees represented by the union. Its representation, it m ain-tains, is not of any specified individuals as of any one time, butof the categories of work in the plan t. Unless this meaning isaccepted, the argument runs, the employer could drasticallyreduce or destroy the bargaining unit for which the union was

designated.

Of course, if this meaning is accepted, the considerations of

economy and business expediency upon which the employerrelies become irrelev ant. Bu t ano ther po ssibility is suggested.

The recognition clause, it is said merely establishes the bargain-ing un it. But good faith, which m ust be an obligation in allagreements, requires that the employer refrain from deliber-ately imp airing that unit wit hou t sufficient justification. Inthis view the recognition clause is violated only if the letting




of the work to the outside contractor is without sufficient busi-ness justification.

Bu t if this is the view found to be required b y the agreement,then it launches an inquiry for which the agreement provides noguides at all: W ha t is sufficient business justification? T o w ha t

extent is the employer's own assertion of business judgmentsignificant? H ow m uch or wh at kind of evidence is necessaryto bolster his judgm ent? H ow m uch or how little economy isnecessary to justify the assumed impairment of the bargainingunit?

Or take the example of employee discipline discussed above.The agreement may be quite clear that the employer has thepow er to discharge or discipline for cause. It m ay be qui te clearin empowering the arbitrator to pass on grievances protestingthe employer's action and even to reduce or modify penalties.Bu t wh at and w here are the guides for his decision? "With theadvent of grievance procedures and arbitration, discharge has

ceased to be regarded as the only available disciplinary measure.Layoffs for various periods are now in general use; and sugges-tion is made of disciplinary demotions, transfers, reduction ofseniority, and the like. W ha t is pro per cause for d isciplinaryaction, and more particularly, for discharge rather than forsome other pe nalty? May such measures as demotion or redu c-tion in seniority be properly used for disciplinary purposes?How much weight is to be attached in each case to the em-ployer's judgment, particularly in view of the fact that it isprecisely that jud gm ent w hich is sought to be curbed b y th egrievance procedure? W ha t significance is to be attached tothe personality of the individual employee, his age, his seniority,his prior reco rd, his promise? W ha t consideration , if an y, is tobe given to probable effects on plant "morale," the morale ofsupervisors as well as of the workers, and the effects at the timethe decision is to be made as well as at the time the penalty isimposed? Th e frequent instances of stoppage of work in adepartment or a whole plant because of a disciplinary penaltyimposed on a single employee indicates that what is involved is




not merely the case of an individual but a group dispute. Fac-tors of this kind should be and doubtless are considered by theparties in the other stages of the grievance procedure. Do theybecome irrelevant when the case is appealed to the arbitra tor?

Here is, of course, the clearest illustration of the arbitrator's

role as creative more than in terpretiv e. It w ould be folly tosuggest that all his work is of tha t cha racter. Despite all plati-tudes as to the inherent ambiguity of language, there are casesin which the language of the agreement appears compelling and

leaves no room for consideration of other evidence of meaning;cases in which the dispute seems frivolous or captious, or pat-ently designed to shift the onus of decision from the party tothe arbitrator, or a desperate effort to recapture a concessionmade in negotiations and subsequently regretted. Assuming,how ever, a real difference of opinion, w hat c riteria m ay th earbitrator look to for the choice between conflicting interpreta-tions, each of which is more or less permissible?

Answer in the form of rules or canons of interpretation isneither practic al nor helpful. Long experience w ith sta tuto ryinterp retatio n has failed to produce such answer. In the lastanalysis, w ha t is sought is a wise judgm ent . It is jud gm ent , saidHolmes, that the world pays for.

12And we can only seek to be

aware of the kind of care and preparation that is necessary informing and pronouncing this judgment.

A proper conception of the arbi trator's fu nction is basic. H eis not a public tribunal imposed upon the parties by superiorauth ority which the parties are obliged to accept. H e has nogeneral charter to administer justice for a community which

transcends the parties. He is rather pa rt of a system of self-govern me nt created by and confined to the parties. He servestheir pleasure only, to administer the rule of law established bytheir collective agreement. The y are entitled to demand tha t,

at least on balance, his performance be satisfactory to them , andthey can readily dispense with him if it is not.

12 H O L M E S , John Marshall, in SPEECHES 9 0 ( 1 9 3 4 ) .



190 MAN AGEM ENT RIGHT S AND THE ARBITRATION PROCESS

To the extent that the parties are satisfied that the arbitratoris properly performing his part in their system of self-govern-ment, their voluntary cooperation in the achievement of thepurposes of the collective agreement is pro mo ted. W hen I speakof the satisfaction of the parties, I do not mean only the advo-cates who may present the case to the arbitrator, or the top

echelons of manag em ent or union represen tatives. I meanrather all the persons whose cooperation is required—all theemployees in the bargaining unit and all the representatives ofmanagement who deal with them, from the job foreman up.

Ideally, the arbitrator should be informed as fully as possibleabou t the dispute wh ich he is asked to resolve. H e should hearall the contentions with respect to it which either party desiresto ma ke. For a pa rty can ha rdly be satisfied th at his case hasbeen fully considered if he is not permitted to advance reasonswhich to him seem relevant and important.

The more serious danger is not that the arbitrator will hear

too much irrelevancy, but rather that he will not hear enoughof the relevant. Indeed, one advantage frequently reaped fromwide latitude to the parties to talk about their case is that theapparent rambling frequently discloses very helpful informa-tion which would otherwise not be brou ght out. Rules of pro-cedure which assure adequate opportunity to each party toprepare for and meet the other's contentions, or rules designedto encourage full consideration and effort at adjustment in theprior stages of the grievance procedure may be quite desirable.But they should not be such as to prevent full presentation ofthe controversy to the arbitrator before he is required to makefinal decision. For th at w ould n ot only limit his resources for

sound judgment, but would tend also to create dissatisfactionwith the system.

The arbitrator may have to take a more active part in theinvestigation tha n does a trial co urt . This is not merely because,being charged with the responsibility for decision, he should besatisfied th at he knows enough to be able to decide. A jud gestarts with some legal premises as to burden of proof or burden



REASON, CONTR ACT, AND LAW IN LABOR RELATIONS 191

of going forward, which are presumably known to the lawyerswho conduct the litigation and are binding on their clients.Even there these burdens are considerably eased by the modernpractice of pretrial examination and discovery. But a collectiveagreement—the arbitrator's law—rarely states any burden of

proof; and the presentation to the arbitrator is not always inthe hands of skilled advocates having the same training for thework and operating on common premises. A cou rt's erroneousfindings of fact in a particular litigation may work an injusticeto the litigants but rarely disturb the future; similar error byan arbitrator may cause more harm by disturbing the parties'continuing relationship than by the injustice in the particularcase.

Moreover, notions of burden of proof are hardly applicableto issues of interpre tation . Even courts do not confine them -selves to the parties' presentations in their search for the mean-ing of the law. Inte rpre tatio n of the agreement requires, how -

ever, appreciation by the interpreter of relevant facts; and thearb itrato r mu st assure himself as well as he can th at he has them .

Finally, in this connection the arbitrator must be quite cir-cumspect in explaining his decision on the ground of indequatepresentation, for his usefulness may depend in large part onthe very people so designated for respon sibility. A nd so, forseveral reasons, the arbitrator cannot simply sit back and judgea deb ate. H e mu st seek to inform himself as fully as possibleand encourage the parties to provide him with the information.

His choice from the more or less permissible interpretationsof the language of the agreement, keeping the basic conceptions

in mind, requires an appraisal of the consequences of each ofthe possibilities. Tho ugh all the parts of the agreem ent do no tnecessarily make a consistent pattern, the interpretation whichis most compatible with the agreement as a whole is to be pre-ferred over one which creates anom aly. Th e effects on effici-ency, productivity, and cost are important factors to be con-sidered. So are also the effects on the a ttitud es and interests ofthe employees. The tw o sets of factors are not always in oppo-




sition. An apparent increased cost may in some circumstancesbe more than repaid by the increased productivity resultingfrom the greater stimulus to volu ntary cooperation. Pra c-ticality of the interpretation in its day-to-day applications is

a related value. The interpretation, no matter how right in

the abstract, is self-defeating and harmful to both sides if itsday-by-day application provides further occasion for contro-versy and irritation.

Appraisal of probable consequences and practicality is no easytask and is not made on the basis of indisputable proof. The

parties, too, make the appraisal. They differ with one anotherand they maydiffer with the arbitrator. But disagreement withthe arbitrator by one or the other of the parties is normal and

expectable and, of itself, not at all unhealthy. Indeed, the sur-

prising thing is the extent of agreement that his award may

meet within the ranks of both parties. For while a party may

speak with one voice at the hearing, the fact is that there may

be considerable difference of opinion among the many peoplewho make up the artificial entity called the party.

The im portant question is not whether the parties agree withthe award but rather whether they accept it, not resentfully,but cordially and willingly. Again, it is not to be expected that

each decision w ill be accepted w ith the same degree of cordiality.But general acceptance and satisfaction is an attainable ideal.Its attainment depends upon the parties' seriousness of purposeto make their system of self-government work, and their con-

fidence in the arbitra tor. Th at confidence will ensue if the arbi-trator's work inspires the feeling that he has integrity, inde-

pendence, and courage so that he is not susceptible to pressure,blandishment, or threat of economic loss; that he is intelligentenough to comprehend the parties' contentions and empatheticenough to understand their significance to them; that he is not

easily hoodw inked by bluff or histrionics; that he makes earnesteffort to inform himself fully and does not go off half-cocked;

and that his final judgment is the product of deliberation and




reason so applied on th e basis of the standards and th e auth oritywhich they entrusted to him.

An important factor tending toward such general acceptanceis the opinion accom panying the arbitra tor's aw ard. It has beenurged by some that an arbitrator's award should be made with-

out opinion or explanation in order to avoid the dangers ofaccumulating precedents and subjecting arbitration to the rigi-dities of stare decisis in the law. Perhaps this view has me ritwhen the particular arbitration is regarded as solely a meansof resolving the particular stalemate and n othin g else. It is anerroneous view for the arbitration which is an integral part ofthe system of self-government and rule of law that the partiesestablish for their continuing relationship.

In this system opinions are necessary, first, to assure the par-ties that the awards are based on reason applied to the agree-ment in the manner I have described.

13To be sure, the opinions

may convince the parties that their arbitrator is inadequate and

should be replaced. This may wo rk a hardship, and at timeseven an injustice, on the arbitrato r. Bu t tha t is a risk w hichthe parties are entitled to impose on his occupation and whichis a necessary feature of the system.

Secondly, in this system a form of precedent and stare decisisis inevitable and desirable. I am not r eferrin g to the use in oneenterprise, say United States Steel, of awards made by anotherarb itrato r in another enterprise, say General Motors. Becausethe publishing business has made arbitration awards generallyavailable, they are being used in this way both by the partiesand by arbitra tors. But they are not so used in the belief tha t

they are entitled to any particular precedential value, for theyare not so entitled. Their value, if any, lies rathe r in the ir sug-gestion of approach or line of argument, or perhaps in theircharacte r of evidence as to practice in other enterprises. Assuch evidence, it must be used, of course, with great circum-spection because of its limited character, and with ample oppor-tunity for the parties to consider it.

13 I pass over the desirability of an opinion to assure the arbitrator himself that he hasreached his conclusion in that way.



1 94 M A N A G E M E N T R I G H T S A N D T H E A R B I T R A T I O N P R O C ES S

But the precedent of which I am now speaking refers to thesuccessive decisions wi thin the same enterprise. Even in theabsence of arbitration, the parties themselves seek to establisha form of stare decisis or precedent for their own guidance—bystatements of policy, instructions, manuals of procedure, andthe like. This is bu t a means of avoiding the pain of rethink ingevery recurring case from scratch, of securing uniformity ofaction among the many people of co-ordinate authority uponwhom each of the parties must rely, of assuring adherence intheir action to the policies established by their superiors, andof reducing or containing the possibilities of arbitrary or per-sonal discretion.

When the parties submit to arbitration in the system of whichI speak, they seek not merely resolution of the particular stale-mate, but guidance for the future, at least for similar cases.They could hardly have a high opinion of the arbitrator's mindif it were a constan tly changing mind . Adherence to prior de-

cisions, except when departure is adequately explained, is onesign that the determinations are based on reason and not merelyrandom judgments.

The arbitrator's opinion can help in rationalizing the agree-ment and the parties' contentions with respect to it and in fos-tering greater appreciation by them of each other's views andneeds with respect to the problem at hand. Its greatest utilitylies in its effect, not merely on the advocates who presented thecase or the higher authorities in the enterprise, but on whatmight be called the rank and file—the workers in the shop andtheir supervisors. It is the rank and file th at m ust be convinced.For the temptation to resort to job action is ever present and iseasily eru pted . The less their pr ivate ru le of law is understoodby the workers and the more remo te from their participation arethe decisions made on their grievances, the greater is the likeli-hood of wildcat stoppages or other restraints on productivity.Th e likelihood can be decreased by b ringing the arbitratio n closeto the shop, not only in the hearings and investigations, but alsoin the opinion which explains the award.




The awards must necessarily set precedents for recurring casesand the opinions must necessarily provide guidance for the fu-ture in relating decision to reason and to more or less mutuallyaccepted principle. Consistency is not a lawyers' creation. It isa normal urge and a norm al expectation. It is par t of the ideal

of equality of treatm ent. Th e lawyer's con tribu tion, indeed, ishis differentiation of rational, civilized consistency from appar-ent consistency. Let me give you an example. In m any appealsfrom disciplinary penalties imposed by the employer, I heardthe union argue earnestly that the penalty should be reducedbecause of the employee's long service record. I was persuadedand held that the employee's seniority should be considered infixing the size of his pena lty. The n came a case in which twoemployees commited the same offense at the same time, and onewas given a larger penalty th an the other. The union protestedthe larger penalty as being an obvious impairment of the prin-ciples of equality . This was no t necessarily conscious op po rtu -

nism, althou gh th ere is always a good deal of th at. A period ofeducation was required to effect the realization, not only bythe advocates, but by the rank and file that the equality forwhich they themselves contended in the area of discipline nec-essitated different penalties for the same offense whenever fac-tors other than the offense itself were considered.

The arbitrator's opinions may thus be a valuable means ofseating reason in labor relations. Bu t the opinions m ust be care -fully restrained. I ven ture to think th at the greater danger tobe guarded against is that too much will be said rather than toolittle. If the opinion wanders too far from the specific p rob -

lem, in order to rationalize and g uide, it runs great risk of errorand subsequent embarrassment to the arbitrator himself. Evenmore unfortunately, it may lead the parties to distrust himbecause he has gone beyond the necessities of the case and hasassumed to regulate their affairs in excess of their consent.

The danger of deciding too much or too early appears in an-other way. The parties themselves, each confidently expectinga decision its way , may press the arbit rato r to decide issues which



1 96 M A N A G E M E N T R I G H T S A N D T H E A R B I T R A T I O N P R O CE S S

might better be left undecided or at least delayed until time andexperience provide greater assurance of wise judg m ent. To thedogmatic and the partisan, there is no need for delay; theirminds are made up and, to the m , delay is confusing and exasper-ating. The U nited States Supreme Co urt has seen the dangers

of premature decision and has developed standards for avoidingit, such as the insistence upon a "case or constroversy" and therefusal to pass upon a constitutional question when a narrowergro und w ill suffice for th e case in hand. Th e conscientious arbi-trator sometimes yearns for similar means of avoiding or delay-ing decision on issues wh ich he feels unrea dy to decide. For itmust be remembered that the arbitrator's decision has a strengthand a carry-over which does not exist in the case of an adjust-ment made by the parties in the lower stages of the grievanceprocedure.

Consider this example: The agreement sets forth certain clas-sifications with attached rates of pay—ironworker, millwright,

crib builder, sashman, belt repairm an, and the like. The w ork ofall these classifications is related by features common to all ofthem. In some plants all the work mig ht be covered b y perhapsone or two classifications r ather than by a half-dozen. Theagreement contains no job descriptions outlining the work ofeach classification, or if there are job descriptions they are eitherunilaterally adopted, or sketchy and expressly not exhaustive, orboth . Disputes arise as to whethe r pa rticular assignments madeon certain days by supervision fall properly within one or an-othe r of the classifications. Th e part icu lar cases m ay come tothe abitrator on appeal of disciplinary penalties imposed on em-ployees who refused the assignments on the ground that they

were not within the classifications of these employees; or theymay come on the grievances of employees claiming that theywere deprived of work belonging to their classification when thewo rk w as assigned to others. Such cases are vexing indeed, forthe parties as well as the arb itrator . Even after long experience,he may find it practically impossible to draw clear and fine linesof dema rcation be tween th e several classifications. If he at-



REASON, CONTRA CT, AND LAW IN LABOR RELATIONS 197

tem pts to prick po ints in a future line by deciding the individualcases as they arise, his task is not much easier because he lacksconfidence as to the direction in which he is going and knowstha t each case may be a prelude to m any others. To decide tha tthe issues are beyond his jurisd iction , because the a greem ent does

not demarcate the classifications, is unsatisfactory because thatmay in effect be a decision for one of the parties and becausethe fact is that the dispute relates to a provision of the agree-ment .

In cases of this character, and others in which the arbitratorconscientiously feels baffled, it may be much wiser to permit himto mediate between the parties for an acceptable solution. I donot suggest it for all cases; nor do I urge that settlement is al-ways bette r th an decision. I suggest it only for those cases wheredecision with confidence seems impossible and where the arbi-trator is quite at sea with respect to the consequences of his deci-sion in the operation of the enterp rise. In such cases, an adjust-

ment worked out by him with the parties is the most promisingcourse. A nd th e possibility of adjustm ent is enhanced if he isable to exert the gentle pressure of a thr ea t of decision. In thisactivity, as in the case of the arbitrator's socializing or meetingwith the parties separately, the dangers envisaged with respect tojudges or other governmental personnel are not equally appli-cable. For the parties' control of the process and their individualpower to continue or terminate the services of the arbitrator areadequate safeguards against these dangers.

Th e example I cited comes from m y own experience. W iththe parties' indulgence, though not with their prior consent, I

withheld decision and let numerous cases accumulate, mean-while gaining more illustrations of the scope of the problem andencouraging the parties to search for solution. W e finally cameup with a mu tual und erstanding which amalgamated the sev-eral classifications into one with an appropriate adjustment ofrate, reclassified the affected employees, disposed of the accu-mu lated cases, and eliminated the problem for th e future . Toavoid certain internal difficulties the understanding was re-



198 MAN AGEM ENT RIGH TS AND TH E ARBITRATION PROCESS

corded not as a signed agreement, but rather as a decision of theum pire, the parties having waived for this case the norm al limi-tations on his jurisdiction.

I have attempted in this paper to sketch the autonomous ruleof law and reason which the collective labor agreement estab-

lishes. It has, of course, its limitation s and its faults. It reliesupo n wholehearted acceptance b y the parties and requires a con-genial and adequate arbitrator, as I have explained, who isneither timid nor rash and who feels a responsibility for the suc-cess of the system. The arbitra tion may be resented by eitherpa rty as an impairm ent of its auth ority or power. It is suscep-tible of use for buck-p assing and face-saving. An d it may some-times encourag e litigiousness. Bu t when the system wo rksfairly well, its value is grea t. T o consider its feature of ab itra-tion as a substitute for cou rt litigation or as the consideration fora no- strike pledge is to take a foreshortened view of it. In asense it is a substitute for both—but in the sense in which a

trans por t airplane is a substitute for a stagecoach. Th e a rbitra -tion is an integral part of the system of self-governm ent. An dthe system is designed to aid management in its quest for effi-ciency, to assist union leadership in its participation in the en-terprise, and to secure justice for th e employees. It is a meansof making collective bargaining work and thus preserving pri-vate enterprise in a free gov ernm ent. W hen it works fairly well,it does not need the sanction of the law of contracts or the lawof arbitration. It is only when the system breaks down com -pletely tha t the co ur ts' aid in these respects is invoked. Bu t thecourts cannot, by occasional sporadic decision, restore the par-ties' continuin g relationship; and their in tervention in such cases

may seriously affect the going systems of self-government.When their autonomous system breaks down, might not theparties better be left to the usual methods for adjustment oflabor disputes rather than to court actions on the contract or onthe arbitra tion award? I suggest tha t the law stay out— but,mind y ou, not the lawyers.

reason, contract and law in labor relations

Documents