In re Chicago, Milwaukee, St. Paul & Pacific Railroad

713 F.2d 274

Opinion

POSNER, Circuit Judge.

These consolidated appeals by the Railway Labor Executives’ Association (RLEA) from orders of the district court, sitting as the Milwaukee Railroad Reorganization Court, require us to consider the adequacy of the labor-protection arrangements ordered by the court for two classes of employees. The first consists of 11 employees of the Chicago, Milwaukee, St. Paul, and Pacific Railroad who did not elect to receive optional benefits negotiated between the railroad and the RLEA and embodied in an agreement signed March 4, 1980, and who are demanding that their statutory benefits be paid immediately rather than deferred till the end of the reorganization proceeding. The second group, also seeking statutory benefits, consists of employees of the Burlington Northern and Union Pacific railroads. These railroads purchased some of the Milwaukee Railroad’s lines and claim that their labor-protection obligations to their own employees are limited to the obligations imposed by the March 4 agreement. As the Union Pacific is not a party to these appeals we shall ignore it and call the two groups of employees whose claims are in issue the Milwaukee group and the Burlington group.

Closely related aspects of the Milwaukee Railroad reorganization were before us two years ago in In re Chicago, Milwaukee, St. Paul & Pac. R.R., 658 F.2d 1149 (7th Cir.1981), and familiarity with our previous opinion is assumed. Briefly, the Milwaukee, which by 1977 was the seventh largest railroad in the country, went broke that year and, not for the first time, sought shelter under section 77 of the Bankruptcy Act of 1898, 11 U.S.C. § 205 (1952 ed.), which though since repealed remains applicable to cases filed under it. Bankruptcy Reform Act of 1978, Pub.L. 95-598, § 403(a), 92 Stat. 2683. It soon became clear that to avoid complete collapse the railroad would have to get rid of about two-thirds of its lines. There were some potential purchasers but the sticking point was that under the Interstate Commerce Commission’s rules any purchaser would have to assume potentially astronomical obligations to the workers made redundant by the purchase. In fact, in the name of “labor protection,” each such worker would be entitled to six years of full pay. See 658 F.2d at 1156. With thousands of the Milwaukee’s workers likely to be discharged, the total cost of protection would have been hundreds of millions of dollars, see id. — far more than the lines were worth to prospective purchasers. It seemed that the only way out was for the reorganization court to “embargo” (authorize cessation of operations on) the lines the Milwaukee wanted to get rid of; the hope was that an embargo would not require the ICC’s authorization and hence no labor-protection conditions would be imposed. When we held that such an embargo would be proper in the circumstances, In re Chicago, Milwaukee, St. Paul & Pac. R.R., 611 F.2d 662, 668-70 (7th Cir. 1979) (per curiam), not only was a major shutdown of rail transportation imminent but thousands of railroad employees could look forward to being laid off permanently with no severance pay. See H.Rep. No. 225, 96th Cong., 1st Sess. 2-3 (1979).

At this point Congress stepped in and passed the Milwaukee Railroad Restructuring Act, 45 U.S.C. §§ 901 et seq., in 1979. Among other things the Act transferred primary authority over sales of the Milwaukee’s lines from the ICC to the reorganization court, but provided in section 5(b)(1), 45 U.S.C. § 904(b)(1), that in authorizing any such sale “the court shall provide a fair arrangement at least as protective of the interests of the employees as that required under” 49 U.S.C. § 11347. If Congress had stopped there, however, it would not have achieved its purpose of averting the Milwaukee’s collapse, for it is under the aegis of 49 U.S.C. § 11347 and its predecessor provisions that the ICC has devised the incredibly expensive labor-protection arrangements that the Milwaukee Railroad could not afford to pay. See New York Dock Ry. v. United States, 609 F.2d 83 (2d Cir.1979). Congress therefore went on to require, in section 9 of the Milwaukee Act, 45 U.S.C. § 908, that the unions and the Milwaukee negotiate a labor-protection arrangement the benefits under which would be treated as administrative expenses of the bankrupt estate (section 9(d)). Under principles of equity receivership, which govern railroad reorganizations, see 11 U.S.C. § 205(b) (1952 ed.); In re Chicago & N.W.R. Co., 110 F.2d 425, 430 (7th Cir.1940), such benefits would be entitled to highest priority, ef. 3 Collier on Bankruptcy § 507.04[1] (15th ed. 1983), and hence as a practical matter would be payable immediately. Section 13 of the Milwaukee Act, 45 U.S.C. § 912, required each employee to choose between receiving benefits under the section 9 agreement and receiving statutory (that is, section 5(b)(1)) benefits.

A section 9 agreement was negotiated— the agreement of March 4, 1980 — and it provided for severance pay equal to 80 percent of a worker’s pay for three years. Faced with a choice between this bird in the hand and more than two birds (100 percent for six years) in a very remote bush, almost all of the Milwaukee’s employees chose the section 9 benefits, thus waiving, by virtue of section 13, any right to section 5(b)(1) benefits. A handful of employees did not elect to receive such benefits; they are the Milwaukee group, whose claim to immediate payment of section 5(b)(1) benefits we turn to first.

Although the Milwaukee Act nowhere states that statutory benefits shall be deferred, it is not easy to make sense out of the Act without assuming that Congress wanted deferral. Since the statutory benefits are not only more generous than the contractual benefits but impossibly more generous, the only way of persuading the employees to waive them — hence the only way of preventing the shutdown of the railroad that it was the statute’s main object to avert — was to give them their substitute contractual benefits immediately. Section 9(d) did this, in effect, by making those benefits administrative expenses of the bankrupt estate.' But all this might have been futile if statutory benefits had also been payable immediately; for then each employee would have had an incentive to hold out for the statutory benefits in the hope that enough other employees would choose the contractual substitute to make his statutory benefits affordable, and such attempts to beggar-thy-neighbor could have put the Milwaukee right back where it had been before the statute was passed. Thus the structure of the statute, though no specific language, seems to require the deferral of statutory benefits. Compare Railway Labor Executives’ Ass’n v. Southeastern Pennsylvania Transport. Auth., 534 F.Supp. 832, 847-48 (Spec.Ct. R.R.R.A.1982) (Friendly, J.). We so held the last time this case was here. See 658 F.2d at 1158-60.

But since then, the Supreme Court has decided Railway Labor Executives’ Ass’n v. Gibbons, 455 U.S. 457, 102 S.Ct. 1169, 71 L.Ed.2d 335 (1982), and we are asked to reexamine our earlier decision in light of it. Gibbons held the labor-protection provisions of another special railroad reorganization statute, the Rock Island Railroad Transition and Employee Assistance Act, 45 U.S.C. §§ 1005,1008, invalid under the bankruptcy clause of Article I, which empowers Congress to enact “uniform Laws on the subject of Bankruptcies throughout the United States.” The Rock Island Act was passed after the Rock Island reorganization court authorized a total shutdown of the railroad and ruled that no labor-protection benefits had to be paid. The Act provided for such benefits; and the provisions that were invalidated in Gibbons required treating the benefits as an administrative expense. Since the provisions “cover[ed] neither a defined class of debtors nor a particular type of problem, but a particular problem of one bankrupt railroad,” the Court held “that such a bankruptcy law is not within the power of Congress to enact. A law can hardly be said to be uniform throughout the country if it applies only to one debtor and can be enforced only by the one bankruptcy court having jurisdiction over that debtor.” 102 S.Ct. at 1177.

Although the invalidated provisions of the Rock Island Act resemble section 9 of the Milwaukee Act, which authorizes one class of the creditors of one bankrupt railroad to have their contractually negotiated severance pay treated as an administrative expense, we need not decide how close the resemblance is. Section 9 is not under attack here. The RLEA is not arguing, and given its representative function hardly could argue, that the thousands of Milwaukee Railroad employees who chose to receive immediate benefits under section 9 obtained an unconstitutional preference. Its quarrel is with section 5(b)(1) of the Act, as interpreted in our previous decision to require the Milwaukee’s trustee to defer payment of statutory benefits to the employees who did not choose section 9 benefits.

Perhaps concerned with the potential constitutional problem, Congress in section 17(b)(1) of the Milwaukee Act extended the labor-protection requirements of section 5(b)(1) to all other railroad bankruptcies pending when the Act was passed. The relevant language of the two sections is identical. If section 5(b)(1) requires deferral of statutory benefits for the Milwaukee’s employees because the Act authorizes them to receive immediate benefits under section 9, then section 17(b)(1) presumably requires the same kind of deferral in any railroad bankruptcy where the employees of the bankrupt line have a right to choose immediate receipt of substitute benefits. The rub is that other bankruptcy laws do not have counterparts to section 9. Whether directly challenged in these appeals or not, section 9 is hard to ignore in determining the effect of section 5(b)(1) on the rights of creditors. Applied to a reorganization under another bankruptcy law that had no counterpart to section 9, section 17(b)(1) might not require any postponement of statutory benefits, in which event the Milwaukee Act would seem to have a nonuniform effect after all.

Fortunately, however, we need not decide whether this indirect effect would be enough to invalidate the Act as interpreted in our previous opinion to require postponement of statutory benefits. First, having taken advantage of section 9 of the Milwaukee Act to the extent of obtaining many millions of dollars in labor-protection benefits for the workers that it represents — money it does not propose be repaid if it wins this appeal — the RLEA may not turn around and challenge the constitutionality of the statutory scheme of which section 9 is an integral part. Section 5(b)(1), the nominal target of the challenge, is, in combination with section 17(b)(1), a uniform bankruptcy law. It becomes vulnerable only when read together with section 9. Having exploited section 9 for its own purposes in this reorganization proceeding, the RLEA will not be heard to challenge its constitutionality, whether directly or indirectly, in the same proceeding.

Second, the objection in Gibbons was to accelerating payment to a class of creditors; here it is to delaying payment to a similar class of creditors. The difference is important. A reorganization court has unquestioned power to defer the payment even of high-priority claims against the bankrupt estate, as we pointed out the last time this case was here, see 658 F.2d at 1164-66, citing Continental Illinois Nat’l Bank & Trust Co. v. Chicago, Rock Island & Pac. Ry., 294 U.S. 648, 55 S.Ct. 595, 79 L.Ed. 1110 (1935). We may assume without having to decide that the reorganization court could not in the absence of the Milwaukee Act force the Interstate Commerce Commission to authorize, without requiring immediate payment of the customary labor-protection benefits, the purchase of a bankrupt railroad’s lines; and we know the Commission has been reluctant to allow postponement of such benefits. See Central R. Co. of New Jersey — Abandonment, 342 I.C.C. 227, 300 (1972); Penn Central Transport. Co. Reorganization, 347 I.C.C. 45, 91-92 (1973). But the Commission has no blanket rule that such benefits may never be postponed and we think it well-nigh inevitable that, irrespective of any compulsion the Milwaukee Act might have laid on it, the reorganization court, with the Commission’s consent, would have used its equitable powers to defer the payment of benefits to the Milwaukee group. The equities of prompt payment were weak. The members of the group were no longer employees; and they could not have had an urgent need of immediate payment, or else, like most of their brethren, they would have elected to receive immediately the generous benefits provided by the arrangement that their representative had negotiated with the railroad under section 9.

It is no answer that the railroad could have “afforded” to pay statutory benefits to 11 employees at the same time as, or shortly after, it paid section 9 benefits to thousands of other employees. There would not have been 11 — there would have been hundreds or thousands — of hold-outs if statutory benefits had been expected to be paid as soon or nearly as soon as section 9 benefits; and with enough hold-outs, the railroad would have shut down and the reorganization would have failed. Postponing payment of statutory benefits was merely honoring the reorganization court’s implicit commitment to create a sequence of payments that would scale down labor-protection benefits to a realistic level. See also 658 F.2d at 1166.

If we were certain that because of these equitable considerations the Milwaukee group would not be paid any earlier even if the Milwaukee Act did not require deferral of statutory benefits, we would conclude that the group lacked standing to question the constitutionality of deferral. We are not so certain as to be able to place decision on that ground, but our analysis suggests an alternative interpretation of the Act that eliminates any problem created by Gibbons. Earlier we said the Act had to be read to require deferral, as otherwise its objective of preventing the Milwaukee from being destroyed by labor-protection obligations would have been thwarted. But this assumes that workers electing statutory benefits could be confident of having them paid promptly unless the statute required that payment be deferred, and they could not be confident; in fact, they could only be confident of the opposite — that the reorganization court would exercise its equitable powers to defer payment. Since a statutery provision requiring deferral was thus not necessary to induce most employees to elect section 9 benefits, the absence of such a provision may not, as we suggested earlier, have been a congressional oversight. And if the statute need not be read to require deferral of statutory benefits but can be read merely to have assumed without requiring that the reorganization court would use its preexisting equitable powers to defer such benefits, the statute itself creates no lack of uniformity in the bankruptcy laws.

The district court held that the March 4, 1980, agreement barred the members of the second group of employees in this case, the Burlington group, from receiving any benefits under the Milwaukee Act. The RLEA argues variously that the district judge had no jurisdiction to construe the agreement, that he misconstrued it, and that regardless of what the agreement provides these employees are entitled to statutory benefits.

The jurisdictional argument is as follows: The March 4 agreement was an agreement between management and labor over conditions of employment in the railroad industry and therefore a collective bargaining agreement within the meaning of the Railway Labor Act, 45 U.S.C. §§ 151 et seq.; that Act has been interpreted to give the arbitration tribunals set up under it exclusive jurisdiction to interpret such agreements, see, e.g., Andrews v. Louisville & Nashville R.R., 406 U.S. 320, 322-24, 92 S.Ct. 1562, 1564-65, 32 L.Ed.2d 95 (1972) (with an exception for so-called “major disputes” that is irrelevant to this case); therefore the district court had no power to interpret the agreement and ought to have stayed its proceedings while the parties submitted their dispute to arbitration.

There is authority for classifying a labor-protection arrangement that is incidental to a transaction regulated by the Interstate Commerce Commission as a collective bargaining agreement under the Railway Labor Act. See, e.g., O’Mar a v. Erie Lackawanna R.R., 407 F.2d 674, 677-78 (2d Cir. 1969), aff’d on other grounds sub nom. Czosek v. O’M ara, 397 U.S. 25, 90 S.Ct. 770, 25 L.Ed.2d 21 (1970). But the March 4 agreement states that it is based on the Milwaukee Act as well as on the Railway Labor Act; the agreement was expressly authorized by section 9 of the Milwaukee Act, even though that section refers only to employees of the Milwaukee; and it is therefore a mixture of a section 9 agreement and a Railway Labor Act agreement. We must decide whether Congress, in enacting section 9 without making any reference to the Railway Labor Act, wanted any dispute over a section 9 (or a mixed section 9-Rail-way Labor Act) agreement to be decided by arbitration, subject to the extremely narrow judicial review of railroad arbitration awards provided in 45 U.S.C. § 153 First (p). See, e.g., Union Pac. R.R. v. Sheehan, 439 U.S. 89, 93-94, 99 S.Ct. 399, 402, 58 L.Ed.2d 354 (1978) (per curiam).

That seems unlikely. Cf. Brotherhood of Locomotive Engineers v. Chicago & North Western Ry., 314 F.2d 424 (8th Cir.1963), and cases discussed there. Section 5(b)(1) of the Milwaukee Act required the district court to impose labor-protection conditions before approving the sale of any of the Milwaukee’s lines to the Burlington. The Burlington argued to the court that the March 4 agreement barred protective conditions for any of its own employees not taken care of in that agreement. To determine the merit of this argument the district court had first to determine whether the parties to the agreement intended it to be exclusive. Although the court may have had the power to refer this preliminary question to one of the arbitration boards set up under the Railway Labor Act (a question we need not decide), to have done so would have delayed the reorganization, thus thwarting a major purpose of the Milwaukee Act — to enable the reorganization court to dispose speedily of the Milwaukee’s surplus lines and thereby prevent the complete shutdown of the railroad. The Milwaukee Act does provide for reference to the Interstate Commerce Commission, see section 5(b)(2), but under stringent time limits. Yet the RLEA contends there is a duty that is without limit of time to refer any dispute over the meaning of a section 9 agreement to arbitration, so that even though the RLEA did not request such a reference until after it had asked the district court to interpret the agreement and had been disappointed by the court’s interpretation, it had an absolute right, founded on the district court’s lack of jurisdiction, to the reference. But it is unlikely that Congress would have wanted to allow these reorganization proceedings to be strung out this way, cf. International Ass’n of Machinists & Aerospace Workers v. Northeast Airlines, Inc., 473 F.2d 549, 559-60 (1st Cir.1972), or that, having displaced much of the ICC’s traditional authority over a bankrupt railroad, Congress was concerned with preserving undiminished the authority of the arbitration tribunals created under the Railway Labor Act.

Moreover, the main reason for giving arbitrators exclusive jurisdiction to interpret rail labor agreements — that they are better at the task than courts — has little realism under a statute that designates a particular district court as the court for the reorganization of a particular railroad. The district judge assigned to sit as that court will learn as much about the labor relations of that railroad as any arbitrator. True, there is some role for arbitration under the Milwaukee Act: employees claiming benefits under section 9 are required by section 9(c)(2) to file their claims with the National Mediation Board, which is one of the arbitral bodies set up by the Railway Labor Act; and presumably the Board would, as an adjunct to its claim-paying function, resolve any disputes over the coverage of the Act. But the employees in the Burlington group are not claiming under the March 4 agreement or any other section 9 agreement (and could not, because section 9 is limited to employees of the Milwaukee) but under section 5(b)(1) itself.

This is another reason why the Burlington group’s jurisdictional argument must fail. By claiming that only an arbitrator can determine whether the agreement is a bar to obtaining statutory benefits, the group is not invoking the exclusive jurisdiction doctrine of the Andrews line of cases, which confine to the arbitration panels set up under the Railway Labor Act all claims that are within the scope of a railroad collective bargaining agreement but not claims, like these employees’ claims, that are not based on the agreement and that the employees contend are not even affected by it. The Burlington group is invoking the distinct doctrine of primary jurisdiction, which comes into play when in the course of a lawsuit properly brought in court an issue arises that is within the competence of some administrative agency to decide, and the suit is stayed while the parties go ask the agency to resolve the issue. See, e.g., City of Peoria v. General Elec. Cablevision Corp., 690 F.2d 116, 120-21 (7th Cir.1982).

The doctrine of primary jurisdiction has been used, in fact though not in name, to require a reorganization court to refer a question of interpretation of a collective bargaining contract to an arbitral tribunal set up under the Railway Labor Act. See Order of Ry. Conductors v. Pitney, 326 U.S. 561, 566-67, 66 S.Ct. 322, 324-25, 90 L.Ed. 318 (1946); cf. Air Line Pilots Ass’n v. Northwest Airlines, Inc., 627 F.2d 272 (D.C.Cir.1980). And Congress’s desire that such a tribunal rather than a court decide an issue might be so strong that the court would be required to refer the issue on its own initiative even if neither party had requested a reference in timely fashion. But even if Congress in enacting the Milwaukee Act had wanted the reorganization court to refer any issue involving interpretation of a collective bargaining agreement to the tribunals set up under the Railway Labor Act — a doubtful proposition as we have seen — we are quite sure it would not have wanted to prevent the parties to the reorganization proceeding from waiving reference. Otherwise, since there is no statutory deadline on reference (indeed, no statutory mention of reference), reorganization could be indefinitely delayed by a tardy demand for reference. We conclude that a failure to make a prompt request for reference is a waiver of any claim that an issue is within the arbitration tribunals’ primary jurisdiction. That claim was waived here by the RLEA’s failure to raise the issue till after it had twice requested the reorganization court to construe the contract. The delay was inexcusable, especially given the importance of expediting the reorganization proceeding.

Having concluded that we have jurisdiction to interpret the March 4 agreement, we must next determine whether the agreement was intended to bar a claim to statutory benefits by any Burlington employee not entitled to benefits under the agreement. The agreement itself is ambiguous. On the one hand it is labeled an agreement “between railroads parties hereto involved in midwest rail restructuring and employees of such railroads,” and thus would seem to have an intended scope going beyond the Milwaukee’s own employees. And it not only makes provision for other railroads’ employees, including employees of the Burlington, which was one of the parties to the agreement, but implies in one passage that that provision is exclusive: “A purchasing carrier [such as the Burlington] will provide a monthly compensation guarantee, as hereafter provided, only to bankrupt carrier employees hired by the purchasing carrier pursuant to this agreement and to its own employees who are (1) working in the same seniority district in the zone or working district of the acquired property and (2) are in active service on the date that interim operation is begun or purchase completed, whichever first occurs.” On the other hand, the preamble of the agreement states that its “scope and purpose . .. are to provide ... a fair equitable and complete arrangement for protection of Milwaukee ... workers,” and later the agreement states that its provisions “shall constitute the complete labor protection obligation of a purchasing carrier to the bankrupt carrier employees who are taken into its employ because of a transaction.” The employees in the Burlington group are not employees of the Milwaukee, the bankrupt carrier.

In an attempt to disambiguate the agreement, the Burlington begins, unpromisingly, with two arguments that are inconsistent— though of course one could be right. The first is that the RLEA has made no showing that any employee not covered by the agreement has been hurt by the Burlington’s purchase of lines from the Milwaukee. The basis of this argument is that the Burlington group consists of employees who do not work in the immediate geographical areas of the acquired lines. The Burlington’s second argument is that it would never have signed the agreement had it been exposed to potentially crushing liability for statutory benefits to employees not covered by it. But if no one is hurt, why would liability — a purely theoretical liability, having no practical consequences — have deterred the Burlington from signing the agreement? In any event, labor-protection arrangements are supposed to be in place before consummation of the purchase, to take care of workers who may be hurt by it. Because the district court had and exercised authority to allow the Burlington to become an interim operator of the Milwaukee lines that it wanted to purchase, there has been some experience with the actual effects of the transfer on Burlington workers in areas remote from the acquired lines, but not enough to be able to say that no such worker could possibly be hurt by the transfer. The March 4 agreement contemplates a merging of seniority districts of the Burlington and the Milwaukee in the areas where the acquired lines are located. As a result of this merger some Burlington workers may become redundant and be laid off, and their collective bargaining agreement with the Burlington may entitle them to relocate elsewhere in the Burlington system, thus bouncing workers there who have less seniority. It is those bounced workers who could be hurt by the transaction yet who would have no protection under the March 4 agreement. It is too soon to tell whether there will be any such workers.

The Burlington’s second argument is more persuasive. If the March 4 agreement was not intended to be exclusive, it would not have solved the great obstacle that the Burlington saw to purchasing lines from the Milwaukee: having to assume vast liabilities for severance pay. Since most of the workers made redundant by the purchase would be in areas where the lines to be purchased were located, a failure to provide for all possible labor-protection claims would not expose the railroads to catastrophic liability. But you do not need mass lay-offs to make labor protection expensive for a railroad. With statutory benefits set at 100 percent of a worker’s wage for six years, one severance can easily cost the railroad $100,000 or more. Since hundreds of Burlington workers are potentially affected by the transaction besides those in the immediate neighborhood of the acquired lines, there is potential liability in the tens of millions of dollars if the March 4 agreement is found to have the gap that the RLEA claims it does. It is unlikely that the parties agreed to such a gap. True, in St. Louis-Southwestern By.-Purchase, 363 I.C.C. 320, 380 (1980), the Commission interpreted the March 4 agreement (which applies to acquisitions of Rock Island as well as Milwaukee lines) as not barring out-of-district workers of the Southern Pacific from getting statutory benefits. But, according to the record of that case, Southern Pacific had agreed to protect these workers before the March 4 agreement was signed, and the agreement expressly entitles workers covered by it to choose between it and any preexisting protective arrangement in their favor.

It is also relevant to note that the benefits agreed upon in the March 4 agreement are less than half as large as the statutory benefits. The only thing that made them attractive to the Milwaukee’s employees was that they were to be paid immediately, not deferred till the end of the reorganization proceeding. But the deferral feature would not apply to employees of the Burlington, a solvent carrier that would have to pay any labor-protection benefits as soon as they become due, that is, as soon as a worker entitled to them was laid off. This means, though, that under the RLEA’s view of the agreement, adversely affected Burlington employees in areas remote from the acquired lines are to receive twice as much money at roughly the same time as Burlington employees in the area of the acquired lines, who are covered by the March 4 agreement. No reason for such a perverse disparity in treatment has been suggested and we doubt the parties intended the agreement to create it.

The remaining question is whether, even if the March 4 agreement was intended to extinguish any claims to statutory benefits by these employees, the employees can claim those benefits anyway, since section 5(b)(1) requires the Commission to make a fair arrangement that will give the workers at least the benefits they would have under 49 U.S.C. § 11347. They cannot get to first base with this argument unless they are employees within the meaning of the Milwaukee Act, which defines “employees” to include any employees of the Milwaukee Railroad “who worked on a line of such railroad the sale of which became effective on October 1, 1979,” but to exclude certain executive officers. 45 U.S.C. § 902(4). Although this definition may not be comprehensive, every specific reference in the Act to “employee” is to an employee of the Milwaukee or occasionally of some other bankrupt railroad; section 9, for example, is explicitly limited to agreements with employees of the Milwaukee. Still, it has long been the practice in railroad acquisitions to impose protective conditions for the benefit of workers of the purchasing as well as of the purchased line, and given the solicitude for labor that suffuses the Act it is implausible that Congress meant to deny the workers of the acquiring railroads any statutory protection.

So the question is whether the March 4 agreement as interpreted to exclude these Burlington workers from any protection from adverse consequences of the acquisition can still be deemed a “fair arrangement” at least as protective of the interests of the employees as is required by 49 U.S.C. § 11347. That section, after also defining minimum labor-protection conditions in terms of a “fair arrangement,” states: “Notwithstanding this subtitle, the arrangement may be made by the rail carrier and the authorized representative of its employees.” This implies at least some deference to voluntarily negotiated labor-management agreements such as the March 4 agreement. And what is “fair” under section 5(b)(1) must have some reference to the background and purposes of the Milwaukee Act. The overriding purpose was to avert a complete shutdown of the railroad by providing affected workers with immediate payments in substitution for the overly generous statutory benefits to which they would otherwise have been entitled but which might actually have been completely worthless to them because, if claimed, they would have plunged the railroad into the abyss. See 658 F.2d at 1156 n. 9. Although the main concern was with benefits for the Milwaukee’s own employees, it would have been difficult to get the purchasing railroads to agree to purchase the Milwaukee’s surplus lines if they could not have bought off all labor-protection claims at once — not only the Milwaukee’s workers’ claims but their own workers’ claims. An arrangement that accomplishes this is “fair” even though some workers, rather remotely connected to the transaction, get no protection.

This conclusion is not inconsistent with Norfolk & Western Ry. v. Nemitz, 404 U.S. 37, 92 S.Ct. 185, 30 L.Ed.2d 198 (1971), where the Supreme Court, interpreting a materially identical predecessor of section 11347, held that a labor-protection agreement incidental to a merger did not override the workers’ statutory entitlements. The railroad in that case relied on the sentence in what is now section 11347 that authorizes voluntary agreements. The Commission had read this sentence to allow it to delegate the making of adequate protective arrangements to parties negotiating such an agreement. The Court refused to allow the Commission to do this, no doubt fearing lest the labor representatives might not protect the interests of all of the workers though obligated to do so by the Railway Labor Act as interpreted in such cases as Steele v. Louisville & Nashville R.R., 323 U.S. 192, 199, 65 S.Ct. 226, 230, 89 L.Ed. 173 (1944). But in this case there was no automatic deference to the parties. The March 4 agreement was reviewed and approved by both the ICC and the district court as a fair arrangement. And the crisis that brought on the agreement in this case had no parallel in Nemitz and makes this agreement a reasonable one in the circumstance even though it entailed the potential sacrifice of the interests of a small number of the purchasing railroads’ workers to the interests of the thousands of Milwaukee employees whose jobs were in imminent jeopardy of disappearing with no severance pay at all. The fact that the sacrifice was only potential further distinguishes Nemitz: it is uncertain that workers so tenuously related to the transaction have to be given any protection for the arrangement to be fair under section 11347. See Clemens v. Central R.R. of New Jersey, 264 F.Supp. 551, 565-68 (E.D.Pa.1967), rev’d on other grounds, 399 F.2d 825 (3d Cir.1968); Latumer v. Burlington Northern, Inc., 501 F.2d 593, 608 n. 37 (9th Cir.1974) (dictum); cf. American Train Dispatchers Ass’n v. ICC, 578 F.2d 412 (D.C.Cir.1978).

Affirmed.