concurring.

Since the facts of this case, as submitted to us by the parties, lack Clarity and precision, I am reluctant to concur in the decision written by Mr. Justice Goodwin which I recognize as carefully reasoned. However, the disposition of the case which the decision will make will do no more than remand it for further attention. I hope that if the case comes before us again we will be afforded a clearer impression of the facts. Whenever a state agency or a court withholds compensation from a workman under a belief that the group of which he is a member is so related to another group that the latter’s action in a strike automatically affects the first group the relationship and its purported automatic effect should be capable -of proof. I hope that clear and cogent evidence will be forthcoming upon remand. The relationship between the two groups, whatever it may be, surely is capable of proof.

specially concurring in part, and dissenting in part.

*432I cannot accept the definition of “class” devised by the majority. That definition would exclude the members of a separate and non-participating union consisting of an entirely distinct craft simply because the project on which they were employed was “integrated” and because they had a “community of interest” with a separate union involved in the labor dispute. Thus, the members of the carpenters union in the present case could be disqualified simply because they happened to be working on a project which required the co-ordination of their work with that of another craft and it was shown that their wages or other benefits traditionally rise sympathetically with those of the other craft. The majority judicially invents this broad definition of “class,” arguing that it is necessary to effectuate the legislative purpose which prompted the enactment of OBS 657.200 (3) (b) and its counterpart in other jurisdictions.

It is admitted that the purpose of this disqualifying provision is not evident in the legislative history of our statute or in the statutes of other jurisdictions. Finding no guide in the legislative history of these statutes, the majority indulges in the speculation that the legislative purpose was to disqualify any class of workmen employed on an “integrated” project if the workmen’s Wages or benefits happen to move sympathetically with those of another craft whose members participated in the strike. To support this speculation the majority points to the danger of the key man strike as one of the principal reasons for the enactment of disqualification statutes such as OBS 657.200. But, it is not pointed out why this extension of the definition of “class” to embrace two or more distinct crafts is necessary to obviate the supposed danger. I understand a key man strike to be a device, consciously em*433ployed, by which a union can hold up production by removing a handful of workmen from an integrated operation. The definition elaborated by the majority disqualifies members of a craft even though in all of its bargaining it remains separate and distinct from the striking union. In fact, the disqualification would apply even though the separate union was, in the particular instance, antagonistic to the striking union. It would seem that, if the majority is to sustain its position, it would be necessary to show that in the history of management-labor conflict separate crafts which bargain separately have joined together for the purpose of accomplishing a key man strike. We have no basis for assuming that this is so.

It may be noted that if one of the separate unions acts in consort with a sister union to effectuate a key man strike the members of both unions would be disqualified under ORS 657.200 (3) (a) because of their “participation” in the labor dispute.

In my opinion the elaborate and complicated definition of “class” developed by the majority not only exceeds the permissible bounds of judicial imagination in the interpretation of statutes, but it also sets up a test which is so indefinite and complex that it will be impracticable to apply in the administration of the act.

I believe that the term “class” should be given a simpler and more definite meaning. I agree that it should be related to the character of the dispute. Ordinarily the bargaining unit will constitute the class. However, under some circumstances the class may be smaller or larger than the bargaining unit. Thus, it is possible that where the labor dispute relates to conditions affecting only a part of the members of the bargaining unit the class may be smaller than the unit of which it is a part. Decisions under the Unemploy*434meat Insurance Acts, 1920-26, Case No. 8344 (July 10, 1924), reprinted in .Selected Decisions Given by the Umpire Prior to 19th. April 1928 respecting Claims to Benefit and Donation at 524 (1929); Shell Oil Co. v. Cummins, 7 Ill2d 329, 131 NE2d 64 (1955). Cf., Members of the Iron Workers of Provo v. Industrial Commission, 104 Utah 242, 139 P2d 208 (1943). On the other hand, where members of the same craft form separate local unions and the local unions do not join together for bargaining purposes, it is possible that the class would embrace all of the local unions thus recognizing that the craft is a class even though it nominally separates its bargaining units into local unions. It is not necessary for us to decide at this time the circumstances under which a class will be less or greater than the bargaining unit. We need only to decide in the present case whether, in the absence of any evidence of unity other than the unity of an “integrated” project and a so-called “community of interest,” different crafts, bargaining separately, are within the same class. I do not think that our statute intended them to be so considered.

There was some evidence of participation upon the part of individual workmen which would disqualify them and perhaps the other members of their union under OKiS 657.200 (3) (a). For that reason, I agree with the majority that the case must be remanded.

dissenting.

I dissent first because the court errs when it undertakes to adopt ground rules to guide the functions of an administrative agency. It is my view that the court has much exceeded the scope of review contemplated by the statutes here involved. Particularly when the *435majority undertakes this voluntary assignment without first deciding what its scope .of review may be. I read the statutes, particularly ORS 652.275 to 657.-290, as imposing a limited duty on the reviewing court. It is noted, for example, that 657.280 delegates to the Commissioner the authority to make necessary rules and regulations governing the entire process, even the procedure to be followed by the Appeals Board. 657.290 grants the Commissioner power to review any decisions at any stage of the process, apparently even after final review by this court. 657.275 permits doubt that the Appeals Board is required to making findings at all.

But perhaps more importantly the court reaches conclusions and imposes judgments on the administrator the repercussion of which we cannot know. We possess, at best, only a superficial knowledge of the problems this opinion may create. How this decision will affect the vexatious problems this ambiguous statute creates in the field of multi-union — multi-employer bargaining of varying geographical and jurisdictional limits, no one can know. I fear that we have added to the burden of the administrator, and the workers and employers, rather than relieved them.

It must, of course, be recognized that the compensation fund must not be used as an agency to foment or support strikes. Neither should the denial of compensation in a proper case be permitted to deter legitimate bargaining rights.

Secondly, in formulating its rules the majority has failed to give consideration to important facts and the knowledge that can, in part, be drawn from the facts. I refer particularly to the treatment of the test called integration. The majority only adumbrates *436the real scope of the facts that makes the word integration rather useless in this case.

For example, the majority gives little consideration to the fact that the Associated General Contractors is itself segregated into two divisions: the building division and the heavy and highway division. The significance of this separation, as it reflects on the subject of integration is plain. The two divisions negotiate separately with the crafts involved. Three of the crafts do not bargain with either division of the Association. The members of those crafts are usually employed by subcontractors and, therefore, bargain with the individual employers.

It cannot be denied that there would be substantial integration or inter-dependence of work between the different crafts employed upon the same construction project. It is also probable that there would be similar inter-dependence between the employes of one employer who would be engaged in two or more projects at the same time. But even in that situation geographical location, differences in crafts employed or other facts could minimize the amount of integration involved. ORS 657.200 (2) recognizes this probability.

I realize that the parties to this case did not press this question on this court for decision. But if the court is going to make a complete set of rules to govern the Commissioner, as it is doing, then these considerations cannot be ignored.

There is a similar lack of consideration for the definition of premises contained in ORS 657.200 (1). That section limits the disqualification to those employed in the “same factory, establishment or other premises . . .” And that 657.200 (2) is the only statutory exception to the limitation of the same premises. *437How a -premise is expanded to cover most of the state is unexplained.

But the inter-dependence between the employes of a contractor building a grocery building in one community and those -of a bridge builder some place else is difficult to -see. Even more difficult to find is any relationship between a sheet metal worker employed in the shop of a subcontractor and a teamster, for example, delivering concrete on a highway project. Particularly when the workers live in different communities and where the actual relationship between employer and employe may widely vary. It requires little imagination to visualize other completely incongruous relationships that the majority would say can be considered to be integrated and inter-dependent each upon the other.

The complexity of the problems presented by this case and the futility of this court’s attempt to make rules that will guide the administrator’s decisions could be demonstrated in much greater detail. It is not necessary. We should decide the case before us, nothing more. If the administrator decides he needs rules to apply to these cases he has the power to adopt them, OPS 657.280. That is an administrative function, not a judicial one.

In respect to most of these claimants there was no integration in any measurable degree. The findings of the Appeals Board that the evidence of integration was not impressive was supported by the evidence. It is admitted that if the Board was talking about integration on a particular job site that the evidence would not support the statement. It is obvious that the Board was talking about over all integration of the variety of construction projects involved.

N-or is there any evidence of indirect interest as *438suggested by the majority. I have grave doubt that the statute permits a test of indirect or limited interest but even so, there is none shown here for most of the claimants.

I agree that some of the claims require further evidence, but not all of them. The statute particularly enjoins the court to expedite decisions in these oases and to decide them. I think the judgment should be affirmed for all but a few of the claimants.

This is an appeal by certain employers① from administrative decisions, affirmed in the circuit court, which awarded unemployment compensation to certain workmen who became unemployed as the result of labor disputes during the summer of 1958.

The disputes involved the projects of some 78 employers engaged in the construction industry. There were two strikes, one in May by the laborers, and one in July by the teamsters and operating engineers. The other major unions bargaining for construction workers were the carpenters, pile-drivers, iron workers, and cement masons. None of these unions was a party to a strike; nor were the handful of sheetmetal *415workers and electricians who were employed on certain work which was suspended during the disputes. Virtually all the claimants were members of one of the nonstriking unions mentioned above.

Each of the named unions conducts its own negotiations with the employers on at least a state-wide basis. The region covered in the collective bargaining agreements may also be larger than the State of Oregon, but these consolidated cases concern only unemployment in Oregon. For collective bargaining purposes, the employers in the construction industry have banded together in the Associated General Contractors of America, Inc., commonly referred to as the AGC. As far as the record indicates, all the interested employers had assigned their bargaining rights to one or the other of the two divisions of the AGC, the Heavy and Highway Chapter and the Building Chapter. Each of these chapters has at least state-wide jurisdiction. The parties have made no distinction between the two chapters, nor do we see any reason to do so.

ORS 657.200 (1) provides:

“An individual is disqualified for benefits for any week with respect to which the commissioner finds that his unemployment is due to a labor dispute which is in active progress at the factory, establishment or other premises at which he is or was last employed or at which he claims employment rights by union agreement or otherwise.”

Every claimant became unemployed by reason of a labor dispute. There was either a lockout or a strike at the project where each was employed. These strikes and lockouts were part of an industry-wide, state-wide disagreement over the provisions to be included in a collective-bargaining agreement between the AGC and the striking union or unions. Either a strike or a lock*416out constitutes a labor dispute for purposes of this statute. Henzel v. Cameron, 73 Adv Sh 323, 228 Or 452, 365 P2d 498 (1961). Thus all the claimants are disqualified for unemployment benefits by OES 657.200 (1), unless they can show that they are requalified under Subsection (3), which provides:

“This section does not apply if it is shown to the satisfaction of the commissioner that the individual :

“(a) Is not participating in or financing or directly interested in the labor dispute which caused his unemployment; and

“(b) Does not belong to a grade or class of workers of which, immediately before the commencement of the labor dispute, there were members employed at the premises at which the labor dispute occurs, any of whom are participating in or financing or directly interested in the dispute.”

These cases involve the application of Subsection (3) to the individual claims. No member of a striking union has submitted a claim. Members of nonstriking unions made idle by the dispute seek to requalify for unemployment benefits under OES 657.200 (3). Since we are concerned almost exclusively with the two parts of Subsection (3), we shall hereafter refer to them as (3) (a) and (3) (b).

No particular difficulty is presented by the provisions of (3) (a). It is conceded that workmen who were not members of the respective striking unions neither contributed to strike funds nor had a direct interest in the outcome of the strikes. The record will show whether or not a given claimant participated in the labor dispute. Eefusal to work behind a picket line constitutes participation within the meaning of OES 657.200 (3) (a). See Baldassaris v. Egan, 135 *417Conn 695, 68 A2d 120 (1949); American Brake Shoe Co. v. Annunzio, 405 Ill 44, 90 NE2d 88 (1950); In re St. Paul & Tacoma Lumber Co., 7 Wash2d 580, 110 P2d 877 (1941); cf. Lexes, et al. v. Industrial Commission, et al., 121 Utah 551, 243 P2d 964 (1952); and see Annotation, 28 ALR2d 333. The evidence showed that several claimants honored picket lines before they were locked out by the employer. These claimants are therefore disqualified under (3) (a). Claimants who did not have an opportunity to observe picket lines because a lockout occurred before such lines were established are not disqualified under (3) (a), although, as will be seen, a number of them might be disqualified under (3) (b), to which we now turn our attention. This omnibus clause requires the individual claimant to prove that he:

“Does not belong to a grade or class of workers of which, immediately before the commencement of the labor dispute, there were members employed at the premises at which the labor dispute occurs, any of whom are participating in or financing or directly interested in the dispute.” OBS 657.200 (3) (b).

The legislature has thus seen fit to disqualify a claimant who did not participate in a given dispute if he falls within a disqualified “grade or class”. Accordingly, a claim may fail even though the workman has no other connection with a given dispute than his chance membership in whatever turns out to be a “class” for the purposes of administering the fund. Conflicting views have been expressed concerning the desirability of this vicarious disqualification.② How*418ever, it is not our function to pass upon the wisdom of the law. Bather, we must attempt to discern the legislative intent and give effect thereto. The cases and reviewers suggest several reasons which may have caused the lawmakers to think some such safeguard was necessary to protect compensation funds from abuse.

Collective bargaining strategists, it is suggested, ■might be encouraged to employ the “key-man” strike③ in the absence of such a clause. Bethlehem Steel Co. v. Board, 219 Md 146, 153-154, 148 A2d 403, 407-408 (1959), noted in 20 Md L Bev 59 (I960). The “keymán” technique permits a union to halt production by removing a handful of workmen from an integrated operation. Compensation claims of the other workmen who are not striking are then charged to the employer when lack of work throws the nonstrikers upon the compensation fund. By having to replenish the compensation fund, the employer is forced to finance a strike against himself. Feldman, supra note 2, at 616; Lesser, supra note 2, at 169.

Another possible reason for Subsection (3) (b) might have been the anticipation of administrative difficulties in trying to determine the degree, if any, of concerted action 'by ostensibly dissociated, but actu*419ally interested, parties. Queener v. Magnet Mills, Inc., 179 Term 416, 424, 167 SW2d 1, 4 (1942) (quoting English report); Note, 33 Minn L Rev 758, 763 (1949); 20 Md L Rev 59, 62 (1960).

Such a statute may have been deemed necessary to preserve the neutrality of the state in labor disputes. Matter of Ferrara (Catherwood), 10 NY2d 1, 8, 176 NE2d 43, 47, 217 NYS2d 11, 15 (1961). This rationale, however, has been subject to criticism. See Lesser, supra note 2, at 174-177; but see Williams, supra note 2, at 356-358.

Not only is the rationale of the grade and class disqualification uncertain but, in addition, efforts to establish criteria for its application have not been crowned with success. In some cases, the class has been held to be equal to (and limited by) the membership of the striking union. E.g., Queener v. Magnet Mills, Inc., supra. In ether cases, the work in which the claimants were engaged was integrated with that of the strikers so that only a fraction of the work force caused the plant or operation to shut down. In these cases the claimants were disqualified as members of a “class”, but the cases involved industrial unions. Thus, the claimants were members of the same union as those who were directly involved in the dispute. Accordingly, such cases may be treated as limiting the “class” to the membership of a single union, although such a rationale is not clear. E.g., Local No. 658 v. Brown Shoe Co., 403 Ill 484, 87 NE2d 625 (1949); Bethlehem Steel Co. v. Board, 219 Md 146, 148 A2d 403 (1959); Rusynko Unempl. Compensation Case [Bethlehem Steel Co. v. Unemployment Comp. Bd. of Rev.], 191 Pa Super 434, 156 A2d 576 (1959), affirmed 402 Pa 202, 166 A2d 871 (1961) (evenly divided court); Oliver Unempl. Compensation Case *420[U. S. Steel Corp. v. Unemployment Comp. Bd. of Review], 189 Pa Super 362, 150 A2d 361 (1959). In other oases, the class may he held to include two or more unions. 'This result may be based upon the relationship of the claimants to the work being done by the strikers as well as upon the nature of the dispute. E.g., Curcio Unemployment Compensation Case [Westinghouse Elec. Corp. v. Unemployment Comp. B. of R.], 165 Pa Super 385, 68 A2d 393 (1949); In re Deep River Timber Co., 8 Wash2d 179, 111 P2d 575 (1941). Where there are different crafts, organized into separate international unions, the unions have on occasion assigned their bargaining rights to a building-trades or other multi-union council. In one such ease, the court treated all the allied crafts as members of the same class under a statute containing a clause somewhat like (3) (b). Olof Nelson Const. Co., et al., 121 Utah 525, 243 P2d 951 (1952); contra, Shell Oil Co. v. Cummins, 7 Ill2d 329, 131 NE2d 64 (1955). Other courts in similar cases have not reached the question of classification under the equivalent of OES 657.200 (3) (b) because they held that the several crafts were so united in interest as to be barred by their direct .interest in the dispute under the equivalent of (3) (a). E.g., A. Borchman Sons v. Carpenter, 166 Neb 322, 89 NW2d 123 (1958); Stahlman Unempl. Compensation Case [Dravo Corp. v. Unemployment Compensation Bd. of Rev.], 187 Pa Super 246, 144 A2d 670 (1958).

A rule that all members of the same union, in the same plant, working at 'the same workbenches, are members of the same class under (3) (b) is no doubt a workable rule in many circumstances. It does not f ollow, however, that the converse is necessarily true. i.e., that nonmembers of .the union are not members *421of the class. Nevertheless, in some cases from other states, nonmembers of a striking union have escaped class disqualification by reason of their snonmembership. Shell Oil Co. v. Cummins, supra; Outboard, Marine & Mfg. Co. v. Gordon, 403 Ill 523, 87 NE2d 610 (1949); Queener v. Magnet Mills, Inc., 179 Tenn 416, 167 SW2d 1 (1942). In substance, this is the position urged by the claimants in the cases at bar. This mechanical formula, however, can produce an illogical result where union and nonunion workmen or members of two or more unions are doing the same work. Queener v. Magnet Mills, Inc., supra. Thus, statutory provisions such as (3) (b) must contemplate classes which may include nonorganized workers or members of more than one union.

The foregoing cases show that no single test can be applied universally to all labor disputes under statutes like ORS 657.200 (3) (b). The subsection can, however, be given a full and fair effect if two concepts, integration of the work and a community of interest between the claimant and the participants in the labor dispute, are carefully applied.

We turn now to the proceedings below. The administrative decision at the first, or deputy, level was to allow the claims. When notified of this result, the employers requested a hearing. The claims were heard by a referee, who took testimony, made findings, and denied all the claims. Then the claimants appealed from the referee’s decisions. The individual cases were consolidated for appeal under ORS 657.275. The Appeals Board reversed the referee and allowed most of the claims. Prom that decision, a consolidated appeal was taken to 'the circuit court under ORS 657.285. The circuit court affirmed the Appeals Board, and *422it is from that decision that the employers bring the present appeal.

The following findings by the referee in the MacInnes case apply generally to all the cases:

“The claimants worked in the nsnal trades utilized in the heavy construction work of the employer. Such trades include operating engineers and teamsters. 'The workmen are organized into .separate trade unions, by crafts, which bargain separately for labor contracts. Although the various trades or crafts do not confirm that they have grouped themselves under a building trades council on the state level, it is admitted they are so joined on a local level in some areas of the state.

“‘The employer is a member of the Associated General Contractors of America, Inc., Heavy and Highway Chapter, to which it has assigned its bargaining rights in labor contracts. In early 1958 the Association entered into negotiations with operating engineer and teamster bargaining agents for new contracts involving wage rates and working conditions. The negotiations broke down about July 1, 1958. On July 9, 1958, pickets of operating engineers appeared at some projects of Association members in southwest Washington .State, and on the following day at some projects in Oregon. A few days later pickets of the teamsters also appeared. At other projects threats of picketing, if the employer did not shut down, were made.

“The Heavy and Highway Chapter of the Association instructed its members on July 10,1958 that its policy was, ‘a .strike against one is a strike against all’, and requested that all members shut down their operations until the dispute had been settled. This employer shut down his operations in compliance with the Association policy and thereafter all projects on which either operating engineers or teamsters were employed remained shut down until the dispute was settled. Work resumed in the week of August 17,1958.”

*423The above findings are accepted by all parties and were adopted by the Appeals Board. Other findings were made by the Appeals Board, and are challenged in these appeal's. In weighing the evidence with reference to (3) (a), the Appeals Board asked itself these four questions:

“(1) Did claimant take any part in picketing?

“ (2) Did he contribute anything of value in support of carrying on the dispute?

“(3) Did he aid in forming the strategy of the dispute?

“(4) Does he stand to gain in the event the strike is successful in its purpose ?”

The board then answered each of the four questions in the negative and thereby concluded that (3) (a) had no application. That conclusion was in error, as we have seen, with respect to at least those few claimants who admitted that they had honored the picket lines. As noted above, observance of, as well as actually walking in, a picket line constitutes participation.

With reference to the more troublesome issues yet confronting us, the Appeals Board also made findings upon the issues under (3) (ib). First, the board asked these questions:

“(1) Do the employees affected have a common labor agreement.

“(2) Do the employees affected have a common bargaining agency.

“(3) To what extent are the duties of the employees affected integrated. This may occur as a totally integrated production line or may be most casual.

“(4) The specialized skill of the employees affected, if any, may have a bearing in some cases.”

*424The board answered these questions as follows:

i(* * # # #

“While it appears that a general increase in wages became effective in all the crafts in 1958 it may be assumed such increase was a result of the merits of the cause advanced at the separate negotiations of the various agencies bargaining for the separate crafts.

“It further appears from the record that the various crafts affected had separate labor agreements, had separate bargaining agencies, and a considerable variance of skill in the different crafts is apparent.

“On the matter of work integration various courts and appeal boards have come to many different conclusions and in this connection we believe each case must stand on its own bottom. The degree •of integration in this cause is not 'impressive.”

The board, then, relies on four factors: separate agreements, separate bargaining, a variation in skills, and nonintegration of work. Historically, a workman’s skill was germane to his “grade”, but not necessarily to his “class”. Cf. Barnett v. Port of London Authority, [1913] 2 KB 115, 124, 128 (workmen’s compensation case). For the purposes of this ease we assume that the workmen were of a variety of pay grades and skills. The board’s findings that there were separate agreements and separate bargaining units are relevant to classification and are supported by the evidence. However, with reference to the findings on the integration of the work, the board doubly erred.

First, the board placed the burden of proof on the employers. To requalify for compensation, a claimant whose claim is challenged under OHS 657.200 (3) has at least the burden of going forward with the evidence until he makes a prima facie case. A. Borchman Sons v. Carpenter, 166 Neb 322, 89 NW2d 123 (1958); *425Auker v. Review Board, etc., 117 Ind App 486, 494, 71 NE2d 629, 632 (1947) (dictum); Martineau v. Director of the Division of Employment Security, 329 Mass 44, 51, 106 NE2d 420, 425 (1952) (dictum); Annotation, 28 ALR2d 287, 331-332.

Furthermore, the board disregarded uncontradicted testimony that there was a high degree of integration of the work on many of the construction projects. The work is integrated when the effective utilization of the plant and labor force requires the continued service of substantially all the workmen. If a strike or slowdown on one part of a production line, or in one phase of a continuous operation, substantially detracts from the efficiency of the work, management is .under no duty to change its methods of operation in order to make work for its employes who happen not to be involved in the labor dispute. Thus, even though it might be possible to continue one step in an integrated operation for some time, perhaps until a stockpile of finished parts becomes too large, or until the inventory of materials becomes unbalanced, the mere possibility that a part of the labor force can be kept busy does not mean the work is not integrated. The test of integration is the regular, and usual, practice followed prior to the eruption of the labor dispute.

There are two reasons why integration of the work is important in cases under ORS 657.200: first, as noted above, the vulnerability of an employer’s integrated operation to the key-man strike, and second, the relationship of the workmen to labor disputes which may affect their common interest in wages, hours and working conditions. Because of the significance of the integration of work, we conclude that a claimant whose work is integrated with that of the strikers is prima *426facie disqualified, as a member of a class contemplated in (3) (ib). This does not, however, end the ease.

It must ¡be remembered that the labor dispute itself remains the key to the operation of all of OKS 657.200, including (3) (b). The scope of the class ultimately should 'be fixed in terms -of the relationship between the claimant and the dispute which caused his unemployment. In the cases at bar, a substantial number, if not all, of the claimants worked in integrated operations. Some may, however, escape the disqualified class if they can show that they 'had no interest, or only a negligible (interest in the outcome of the particular dispute.

A claimant, even though he works in an integrated operation, is not a member of a disqualified class if he has nothing to gain or to lose, directly or indirectly, as the result of the dispute. If, however, he works in an integrated operation and has an indirect interest in the dispute he is disqualified under (3) (b). A direct interest in the dispute disqualifies the claimant under (3) (a), and need not be considered further at this point.

Negligible interest, i.e., remote interest or mere sympathy, does not disqualify at all. Thus, we may disregard the kind of interest which might be characterized as generic in labor relations, i.e., the sympathy a member of one union might entertain upon general principles for the objectives of another union. The kind of indirect interest which will bring a claimant into a class contemplated by (3) (b) therefore lies between direct and negligible, or inconsequential, interest.

Whether or not such an indirect interest exists between a given claimant and the particular labor dis*427pute depends upon all the facts. If a group of workmen stop work because of a grievance the group may have agaiinst their foreman, workmen not supervised by that foreman would not ordinarily have a sufficient community of interest with the first group to throw nonstrikers into the disqualified class under (3) (b). In such a case, membership in a common union would be significant only if union policy or some other factor supported an inference .that concerted action actually involved a group larger than the ostensibly disputing group. On the other hand, if some of the unskilled workmen in a plant stop work in order to enforce a demand for higher wages, and management has traditionally maintained constant pay differentials between skilled and unskilled workers, then the skilled workers would have an indirect interest in the strike. The only difference between the economic interest of the strikers and that of the nonstrikers in such a case would be the fact that the resolution of the dispute would necessarily affect the strikers, whereas it would not necessarily affect the others. It is conceivable that management might choose to disregard the traditional pay differentials, at least for a time, with reference to the nonstrikers. Even here, however, the historical differential is a strong indication of the interest the skilled workers had in the dispute, and they may, in turn, strike to maintain the differential. In any event, such a state of affairs would indicate that the two groups shared a community of interest in the progress of each other’s negotiations. This community of interest would be equally strong, both from an economic point of view and from the employer’s point of view, whether the two groups of workmen were organized in the same union or in several unions. Thus we see that the most important factor in the application of *428(3) (b) is the presence or absence of a community of interest between the workmen involved in a labor dispute and their nonstriMng fellow workmen. This community of interest, as we have seen, may exist within a single union, or between several unions, depending upon :the facts of each case.

In dealing with the requalification of claimants who are prima facie disqualified under OBS 657.200 (1) because a labor dispute exists at their place of employment, the trier of fact must always consider union membership. This fact is hardly ever in controversy. As we have seen, membership in a common union, in at least a preliminary way, provides one basis for classification under Subsection (3) ‘(b).

Whether to include‘within a single class under (3) (b) workmen in a given case who may belong to different unions involves a choice between classification based exclusively on organizational lines and classification on functional as well as organizational Ines. If classes are to be assigned solely according to bargaining units, then workmen equally interested in the 'labor dispute which caused their unemployment will receive compensation or not according to the fortunes of union membership. In Illinois, two cases reached opposite results, apparently because one plant bargained with one industrial union while the other bargained with a variety of .separate craft unions. See Local No. 658 v. Brown Shoe Co., 403 Ill 484, 87 NE2d 625 (1949), and Shell Oil Co. v. Cummins, 7 Ill2d 329, 131 NE2d 64 (1955). If .the matter of classification is considered in a functional context, union membership remains relevant, but not necessarily controlling. .Since the addition of (3) (b) to the statute must have been intended to enlarge the class beyond that already ‘disqualified by *429direct interest in the dispute under (3) (a), and since all members of the same bargaining unit ordinarily have a direct interest in the outcome of their own union’s dispute, we believe the functional approach will achieve a more logical and workable result in a majority of the cases than the mechanical approach to classification based solely upon bargaining units.

The problem of determining whether, in the cases at bar, the nonstriking crafts should be included under (3) (b) with the striking crafts leads us to an examination of factors which may have occurred to the Appeals Board, but which, so far as its findings disclose, were largely ignored. We refer to the relationship, if any, between wage, working-condition and fringe-benefit improvements gained by the laborers, teamsters, or operating engineers as the result of their respective strikes and similar gains flowing to the various nonstriking unions. The findings assumed that such benefits were unrelated, but the board does not appear to have decided the question. There was evidence that some nonstriking workmen belonged to unions which had already entered into new agreements and accordingly had no interest in the bargaining of the strikers. In the case of other nonstrikers, bargaining was yet to be concluded. However, even if there were an established pattern of derivative benefits from the bargaining of one union flowing to members of other crafts, the members of the nonstriking unions could still requalify under (3) (b) by showing that they worked on projects where the work of the several crafts was not functionally integrated. If, however, the work was integrated, and the claimants could not show that their gains, if any, were unrelated to those of the strikers, then these two factors would bring members of different unions into the same class for *430the purposes of the particular labor dispute.④ The administrative agency should have taken account of the separate and divergent interests of the several claimants instead of treating them all alike.

Whatever findings the agency might reach with respect to the question of integration of the work and the .claimant’s relation, if .any, to the dispute must be set forth so that judicial review can be had in the manner contemplated by the statute. It should only be necessary for the court to study the record and see whether the findings are supported by evidence. In the oases at bar, many important operative facts are to be discovered in the transcribed testimony. But the findings reveal only that the hoard considered evidence of integration to be unimpressive and that it assumed certain bargaining benefits to be coincidental. Such findings were contrary to the evidence, at least with reference to part of the claims. The findings, moreover, did not give recognition to the honoring of picket lines which disqualified some of the claims under (3) (a). In these particulars the findings were in error.

Reluctant as we are to remand the numerous individual claims for further proceedings, we must conclude that the evidence did not support the findings *431entered by the Appeals Board in the particulars mentioned above, and accordingly it was error to approve them in the trial court.

The decree must be reversed with instructions to remand the cause to the commissioner as provided by ORS 657.285 (6). Claims disqualified under ORS 657.200 (3) (a) should be clearly identified. After disposing of the questioned claims under (3) (a), no other claim should be allowed unless it is supported by new findings of fact which specifically show that the claimant (or group- of similarly situated claimants) produced evidence that took the claim out of the class of those disqualified by ORS -657.200 (3) (b).

Reversed and remanded.