Matter of Valuation Proceedings, Etc.

531 F.Supp. 1191 (1981) In the Matter of The VALUATION PROCEEDINGS UNDER SECTIONS 303(c) AND 306 OF the REGIONAL RAIL REORGANIZATION ACT OF 1973. Misc. No. 76-1. Special Court, Regional Rail Reorganization Act. November 24, 1981. On Petitions for Reconsideration January 25, 1982. *1192 *1193 *1194 *1195 *1196 *1197 William C. Lieblich, Washington, D. C., and Thomas B. Radom, Southfield, Mich. (Thomas S. Warrick and Pierson, Semmes, Crolius & Finley, Washington, D. C., David W. Berry, and Joseph S. Radom, P.C., Southfield, Mich., of counsel), for Trustee of the Ann Arbor R. Co. Stanley Weiss, Newark, N. J. (Dean R. May and Carpenter, Bennett & Morrissey, Newark, N. J., of counsel), for Central Jersey Industries, Inc. John N. Malyska, Newark, N. J., for Delaware and Bound Brook R. Co. *1198 Harry G. Silleck, Jr., John L. Altieri, Jr., and Howard W. Goldstein, New York City (Peter W. Asher, Thomas G. Gallatin, Jr., Jack P. Adler, Harold G. Levison, Susan M. Schmedes, Frances E. Edwards, Douglas M. Parker, Frank T. Laznovsky, Robert A. Jaffe, Karen M. Cullen, Robert A. Cantor, and Mudge, Rose, Guthrie & Alexander, New York City, of counsel), for Trustees of the Erie Lackawanna Ry. Co. John Michael Clear, St. Louis, Mo. (Stuart R. Warren, and Bryan, Cave, McPheeters & McRoberts, St. Louis, Mo., Harry G. Silleck, Jr., Peter W. Asher, John L. Altieri, Jr., and Mudge, Rose, Guthrie & Alexander, New York City, David C. Toomey, Jared I. Roberts, Richard C. Unger, Jr., Lewis R. Olshin, and Duane, Morris & Heckscher, Howard H. Lewis, James A. Sox, and Obermayer, Rebmann, Maxwell & Hippel, Philadelphia, Pa., John Vanderstar, S. William Livingston, Jr., Carolyn F. Corwin, Arvid E. Roach II, and Covington & Burling, Washington, D. C., of counsel), for Joint Terminals Transferors. Jack M. Zackin, Newark, N. J., for Receiver of the Lehigh and New England Ry. Co. David C. Toomey, Philadelphia, Pa. (Jared I. Roberts, Richard C. Unger, Jr., Lewis R. Olshin, and Duane, Morris & Heckscher, Philadelphia, Pa., of counsel), for Trustee of the Lehigh Valley R. Co. Thomas E. Tyre, New York City, for North Pennsylvania R. Co. Charles I. Thompson, Jr., Philadelphia, Pa., for Philadelphia, Germantown & Norristown R. Co. Howard H. Lewis, Philadelphia, Pa. (James A. Sox, Richard W. Vitaris, Robert F. Stewart, Jr., John J. Ehlinger, Jr., and Obermayer, Rebmann, Maxwell & Hippel, Philadelphia, Pa., of counsel), for Reading Co. and Certain Affiliated Transferors. William R. Perlik, Michael S. Helfer, Roger M. Witten, David R. Johnson, Eric A. Von Salzen, and Stephen C. Rogers, Washington, D. C. (Peter F. Rousselot, Robert H. Kapp, Joe Chartoff, George W. Miller, Gail Starling Marshall, George W. Mayo, Jr., Walter A. Smith, Jr., V. Mark J. Policy, William J. Cassidy, Jr., Ronald E. Stauffer, Cathy A. Simon, Wallace A. Christensen, Stephen A. Goldberg, David C. Kohler, Karen L. McClearycale, Ann Morgan Vickery, Kevin J. Wolff, George H. Mernick, Paul W. Nolan, Steven B. Rotman, R. Verne Bradford, and Hogan & Hartson, Henry T. Rathbun, Arthur Z. Gardiner, Jr., A. Douglas Melamed, Richard W. Cass, David M. Becker, William J. Perlstein, Stephen P. Doyle, John B. McNeece, III, Maury J. Mechanick, Bruce E. Coolidge, Marie N. Doland, Lauren B. Homer, Sheila Bihary, Joan S. Powers, Judith Barry Wish, and Wilmer, Cutler & Pickering, Cary W. Dickieson, Washington, D. C., Frazer C. Hilder, Detroit, Mich., David H. Siehl, David L. Prestemon, Barbara E. Sosnick, Larry C. Malski, Jonathan E. Pansius, and Mark H. Sidman, Railway Association, Washington, D. C., of counsel), for United States Ry. Assn. Alice Daniel, Asst. Atty. Gen., John H. Broadley, Thomas S. Martin, and Jane A. Restani, Dept. of Justice, Washington, D. C., for United States of America. Before FRIENDLY, Presiding Judge, and WISDOM and THOMSEN, Judges. TABLE OF CONTENTS Introduction 1199 I. Procedural History 1201 II. Positions of the Parties 1204 A. Joint Terminals 1205 B. Penn Central 1205 C. Erie Lackawanna 1206 D. Lehigh Valley 1207 E. Reading 1208 F. Central Railroad of New Jersey 1209 G. Ann Arbor 1209 H. Lehigh & New England 1209 I. Government Parties 1209 III. Introductory Analysis 1210 IV. General Background Considerations 1214 V. Bargaining for Properties Having Earnings Value 1217 VI. The Effect of Inflation on Earnings Forecasts 1222 VII. Valuation Methodology 1225 A. DCF vs. EBIT 1225 1. Timing 1225 2. Taxes 1226 3. Capital Expenditures Needed to Maintain Income 1227 B. Discount Rate 1230 1. Debt-to-Equity Ratio 1231 2. Cost of Equity 1232 (a) Railroad Risk 1233 (b) Special Risk 1233 3. Cost of Debt 1235 4. Calculating AWCC 1236 VIII. Rehabilitation Costs and Maintenance of Way Expenses 1236 A. The Government Parties' Study 1237 1. Track 1238 (a) Rail Replacement 1238 (b) Turnouts and Crossings 1242 2. Bridges 1243 3. Yards 1244 4. The Government Parties' Defense 1245 5. Conclusion 1246 *1199 B. Erie Lackawanna's Study 1247 C. Lehigh Valley's Study 1251 D. Ann Arbor 1254 E. Annual Maintenance of Way Expenses 1257 F. Conclusion 1259 IX. Labor Protection Costs 1260 A. Introduction 1260 B. Level of Labor Protection 1261 1. Existing Agreements 1261 2. Action of the Reorganization Courts 1262 3. ICC Action 1263 4. The Negotiating Process 1264 C. Calculation of Labor Protection Costs 1270 1. Erie Lackawanna 1271 (a) Operating Employees 1271 (b) Nonoperating Employees 1282 2. Lehigh Valley 1286 3. The Effect of Bargaining and Other Factors on the Amount of the LPC Deductions 1290 4. Ann Arbor 1291 5. Lehigh & New England 1293 X. Timing and ICC Conditions 1293 A. The Parties' Positions 1294 B. Analysis 1298 1. Likely Congressional Initiatives in the Alternative Scenario 1298 2. Probable Actions of the ICC 1306 (a) Conditions Reducing Value 1306 (b) Early ICC Approval of Particular Transferors' Sale and Abandonment Applications 1308 (c) Necessary Transactions Following Approval 1310 C. Conclusion 1311 XI. Acquisition by Public Bodies 1311 A. The Joint Freight Terminals Case 1313 B. Predominantly Passenger Lines in Northern New Jersey 1316 1. Legal Capacity 1316 2. Public Acquisition 1317 3. Financial Capacity 1320 4. Pricing 1324 (a) Erie Lackawanna's Claim for a Premium in Excess of X 1324 (b) Government Parties' Claim for Deductions from X 1333 (c) Bargaining with Public Bodies 1334 (d) Magnitude of the Deductions 1337 5. Trackage Rights Fees 1342 XII. Proffers of Freight Lines at Capitalized Earnings — Issues Not Heretofore Considered 1346 A. The EL-LV Combination 1346 1. The Existence of Competitive Bidding 1347 2. The Earnings Projections for EL-LV 1351 (a) Laurie's and Adik's Traffic and Revenue Studies 1351 (b) Jennison's and Adik's Cost Studies 1353 (c) Shannon's Further Adjustments 1355 3. The Addition of $100,000,000 to Reflect a Grant by the State of New York 1359 4. Consequences to the EL-LV Earnings Case of Our Rejection of the Joint Terminals Concept 1361 B. Ann Arbor 1364 1. The Position of the State of Michigan 1365 2. Competitive Bidding 1366 3. The Extent to which Off-Segment Earnings Should be Taken into account in Determining Capitalized Earnings Value 1369 4. American Motors Jeeps 1370 5. The Cost of Handling Off-Segment Traffic 1371 6. Use of 1973 Traffic as the Base for Projecting Revenue 1372 C. Lehigh & New England 1373 XIII. Central Railroad of New Jersey 1374 A. Central Railroad of New Jersey's "Real Life" Sale to New Jersey 1375 B. Central Railroad of New Jersey's Alternative Scenario 1379 1. Purchase for Rail Use 1379 2. Timing 1379 3. Price 1382 Glossary PER CURIAM: We here set forth such conclusions as we have been able to reach with respect to the value the transferors (Ts) which have not settled their cases could have obtained for properties that would have continued in rail use. We state the scope of this long opinion thus modestly because of several factors that could not have been anticipated when our CMV Opinion, 445 F.Supp. 994, (Sp.Ct. R.R.R.A.1977), and our initial pre-trial orders were issued. During the course of the proceedings, although after the evidentiary hearings were concluded, Penn Central (PC), the largest T, representing some 80% of the case in terms of property value, settled with the Government Parties (GPs) — the United States and the United States Railway Association (USRA). This was quickly followed by settlements on the part of several small Ts — the Lehigh & Hudson, the New York & Long Branch, the Pennsylvania-Reading Seashore Lines, and the Dayton Union. After the case had been briefed and argued and this opinion had been partially drafted, another important T, Reading (R) and its subsidiaries, also settled. Shortly thereafter, the independent owners of rail property operated by R, the Philadelphia, Germantown *1200 & Norristown, the Delaware & Bound Brook, and the North Pennsylvania, reached settlements with the GPs. Then the Niagara Junction Railroad, jointly owned by PC, Lehigh Valley (LV), and Erie Lackawanna (EL); the Kalamazoo, Allegan & Grand Rapids Railroad, an independent line leased by PC; and the Norwich & Worcester agreed in principle to a settlement of each of their claims with the GPs. This left EL, LV, Central Railroad of New Jersey (CNJ), Ann Arbor (AA), and Lehigh & New England (L&NE) as the only Ts whose claims are still at issue. When this opinion had been substantially completed, save only for minor textual revision, final checking, and typing, EL and the GPs, on October 22, 1981, filed a motion for an order suspending litigation activities between them. The motion stated that on October 19 negotiators for the United States, USRA, and the EL Trustees had reached an agreement for a settlement subject to necessary approvals of their principals. Government counsel explained, at a conference with the Presiding Judge of this court, that the obtaining of all necessary Government approvals might take several weeks. Upon these being obtained, the EL Trustees and the GPs proposed to submit the settlement agreement to us for approval. An order to show cause why the order suspending litigation should not be entered was issued on October 22, 1981, Sp. Ct. Rep. N 37875; no objections having been filed, the order was entered. On November 17, 1981, the GPs and the EL Trustees petitioned for approval of the settlement, and this court has entered an order to show cause specifying the procedure for intervention and the filing of objections, and setting a hearing for January 25, 1982. As pointed out in a memorandum accompanying the motion to suspend litigation, our hearing on the approval of the settlement must take account of Order No. 1094 of the District Court for the Northern District of Ohio in the EL reorganization proceedings. In this order, dated September 9, 1981, Judge Krupansky, the reorganization judge, authorized the EL Trustees to enter into a settlement agreement with the GPs subject to our approval and without the necessity of further proceedings in the reorganization court respecting the terms of the settlement. This authorization was conditioned on the Trustees' giving "reasonable notice to all parties customarily notified in these [reorganization] proceedings of any hearing before the Special Court concerning approval of such settlement agreements and shall support any such party's right to be heard by the Special Court concerning any such settlement agreement". In the event that any party to the reorganization proceeding filed a timely petition to intervene in this court and we did not permit it to intervene and be heard, the authority granted the Trustees by Order No. 1094 was made subject to further order of the reorganization court at which such party would be given an opportunity to be heard. Our order to show cause with respect to the settlement takes account of Judge Krupansky's Order No. 1094. We have concluded that, despite these developments, we should not defer the filing of this opinion until after the approval hearings on the EL settlement have been concluded. As will appear hereafter, we cannot and do not here arrive at the figure which, in our view, EL would have obtained in these proceedings; the greatest unknown and at present unknowable is the value of the passenger lines in Northern New Jersey which we conclude that the State would have purchased despite their negative earnings value. Release of the opinion in advance of the settlement approval hearing should assist the parties to the reorganization proceeding and ourselves in appraising the fairness of the settlement without, however, precluding its approval as an opinion setting total value might have done. We say "might" since even if we had arrived at a total figure higher than the value of the settlement (including the significant benefit to EL of receiving the proceeds now rather than on December 31, 1987, see Regional Rail Reorganization Act of 1973, § 306(c)(1)), account would have had to be taken of the possibility of disagreement by the Supreme Court with some or even all of our conclusions that are favorable to EL. Moreover, we think justice requires that the *1201 railroads that have not settled should be advised of our conclusions as soon as we are ready to announce them, both because the scheduling of the nonrail use phase of the case is geared to such announcement, Sixth Pre-Trial Order, ¶ 7, Sp. Ct. Rep. at N 35967, and because the announcement may facilitate settlement by them. When EL and the GPs have deferred their settlement so long, we perceive no obligation on our part completely to rewrite the opinion at this late date so as to eliminate the lengthy discussion of EL's case, with the possibility of having to restore this if the settlement should be disapproved. One major obstacle has prevented our accomplishing as much as we had hoped in this opinion. In a good many instances Ts have argued that rail use properties would have been sold at a price equal to or at a premium above their value for nonrail use. Analysis has made clear that the parties entertain widely different views as to what nonrail use value is and, for reasons developed in Part III of this opinion, the determination of that issue must await the nonrail use phase of the case, although we do here rule on whether premiums would have been paid or deductions made. While, as we recognized long ago, CMV Opinion, 445 F.Supp., at 1044, no amount of evidence, argument, deliberation and study can produce a single demonstrably right figure, we had hoped to be able to produce in this phase of the case definite figures of value for rail use. Although we have not been able fully to achieve this goal, we have been able to decide a large number of issues, to set forth the general method we intend to pursue in completing our work, and to focus the parties' attention on certain problems that must be dealt with in the ensuing phase of the case dealing with nonrail use value. Conceivably, this may cause the remaining parties and the GPs to do what we recommended over four years ago, CMV Opinion, 445 F.Supp. at 1045-46, and what finally was done in the cases of PC, R, and EL as well as those of smaller railroads, namely, to arrive at reasonable settlements. We shall begin our opinion with two expository sections, Part I recounting the procedural history of the case since the CMV Opinion and Part II summarizing the positions of the parties. Part III contains a legal analysis of certain problems, particularly as to what we may or may not properly now decide, that have surfaced during the presentation of this phase of the case. This is followed by six parts dealing with "cut-across" issues relating to claims for values based on earnings power, to wit, the claims of the GPs that the Ts' assertion of earning power should be rejected out of hand because of the allegedly hopeless conflict between them and the history that led to enactment of the Rail Act (Part IV); the nature and effect of bargaining (Part V); the GPs' claim that purchasers would have sharply reduced the Ts' estimates of future earnings because of inflation (Part VI); the method for translating estimated future earnings into earnings value (Part VII); the deductions that a purchaser would have made because of the need to rehabilitate and maintain the properties (Part VIII); the deductions that a purchaser would have made to reflect the cost of labor protection conditions (Part IX); and the timing of projected sales and the conditions, if any, that the Interstate Commerce Commission (ICC) would have affixed (Part X). In Part XI we deal with a number of vexing issues concerning proposed acquisitions by public bodies. Part XII is devoted to problems not theretofore considered relating to the sales for earnings value proposed by EL, LV, AA, and L&NE. Finally, in Part XIII, we deal with special contentions made by CNJ. I. Procedural History The CMV Opinion, 445 F.Supp. at 1009, set out a schedule designed to initiate a phase of the case intended to determine values the Ts could have obtained for their properties for continued rail use in the absence of the Rail Act. Each T claiming values from sales for rail use was required to exchange its proposal with all other Ts by January 9, 1978, so that they could meet among themselves to facilitate resolution of inconsistencies and deliver their resulting proffers to the GPs by March 9, 1978. The GPs were given 30 days to indicate objections relating to inconsistencies and the Ts *1202 were to file their finally revised proffers with the court within 30 days after receiving the objections. The First Pre-Trial Order, dated April 18, 1978, Sp. Ct. Rep. N 10078, further required that each T specify the principal rail use issues it believed had to be tried, along with a description of the evidence it intended to offer to demonstrate what it would have been able to realize for transferred properties in the absence of the Rail Act. This information and other reports from the parties provided the basis for a pre-trial conference on June 21-22, 1978, that led to the Second Pre-Trial Order, dated June 23, 1978, Sp. Ct. Rep. N 11796. It established a schedule for the submission of direct evidence by the Ts, and by the GPs on "affirmative defenses".[1] The latter issues included, among others, the cost of rehabilitating the properties proposed for transfer, the abilities of private purchasers and public agencies to finance the transfers, and the impact of labor protection obligations upon the values the Ts would have received for their properties. Most importantly, all direct evidence was ordered to be presented in writing.[2] The court rejected the proposals of various Ts that there be separate trials on separate issues because a single trial encompassing all rail use issues was believed conducive to more rapid progress. We also directed each T to submit its direct rail use case in such form as would permit us to make necessary adjustments where we found some but not all of a T's contentions persuasive. The significance of this order and the failure, perhaps in part inevitable, fully to comply with it, will become apparent in later parts of this opinion. Finally, as developed in Part III, preparation for trial of issues related to the values of the Ts' properties in nonrail use was deferred. The Ts each submitted their direct evidence, accompanied by a memorandum summarizing the submission, in December 1978; the GPs' submission arrived as scheduled on January 31, 1979. The Third Pre-Trial Order, dated January 11, 1979, Sp. Ct. Rep. N 15287, provided that the parties should conduct discovery and cross-examination by means of the deposition process, with discovery to be conducted by the exchange of information outside of the formal discovery process to the fullest extent practicable. One result of the dual requirement that all testimony be in writing and cross-examination be by deposition was to avoid what had been expected to be the need for a multitude of special masters in this phase of the case. This also permitted the parties to make their own agreements on how the discovery and cross-examination process should be handled and to conduct as many as a half-dozen cross-examinations simultaneously. The remainder of the order limited the number of depositions to which each witness could be subjected, provided that depositions may be offered for admission as testimony, and established procedures for the expeditious processing of objections.[3] The Fourth Pre-Trial Order, dated May 16, 1979, Sp. Ct. Rep. N 17130, effected a more precise division between the rail and nonrail use phases of the case. We granted a motion of the GPs to defer consideration of evidence of nonrail use values or of appraisal methodologies for determining nonrail use values of any properties. However, evidence indicating that certain lines would *1203 have been kept in rail use for reasons other than their earnings power as well as evidence that purchasers of lines sold for rail use would have paid prices based in whole or in part with reference to nonrail use values was to be included in the rail use phase of the proceedings. All parties were directed to begin preparation of their nonrail use evidence at this time, although our consideration of that evidence was to be delayed until after this decision. The Fourth Pre-Trial Order also directed the GPs to file their remaining evidence on rail use issues by November 1, 1979. This demand was less than successful; on October 23, 1979, the GPs filed a motion for a 90-day extension of time, Sp. Ct. Rep. N 18191, which was opposed by the Ts. The court could discern no adequate basis for the GPs' waiting until October 23 to call the problem to the court's attention, and also considered there was no justification for the requested delay. Yet the public interest in fully developing the facts in this litigation required granting the motion of the GPs. The Fifth Pre-Trial Order, dated March 27, 1980, Sp. Ct. Rep. N 19675, scheduled rebuttal evidence and started the briefing schedule. The Ts were given until October 10, 1980, to file any additional rail use evidence by way of rebuttal. The GPs were ordered to file their schedule of depositions with respect to the Ts' rebuttal evidence by November 10, 1980. On October 24, 1980, the GPs moved to strike a substantial portion of the Ts' rebuttal evidence as not constituting proper rebuttal, Sp. Ct. Rep. N 22028. This motion was granted in part; we struck witnesses' repetition of initial testimony and other evidence that could and should have been presented years before, Sp. Ct. Rep. N 22621. In the interest of expedition we ordered two simultaneous exchanges of briefs rather than sequential filings. The initial briefs were prepared on the basis of the Ts' and GPs' direct rail use cases, thus avoiding the delay that would have resulted from awaiting completion of the taking of rebuttal evidence. On May 8, 1980, in a separate order, Sp. Ct. Rep. N 20410, we denied the Ts' request that the initial briefs be accompanied by separate proposed findings of fact and conclusions of law to which opposing parties must respond. Initial briefs were scheduled to be due by December 2, 1980. Both sides were ordered to include in their briefs discussion of evidence developed by the opposing parties in their direct cases. Given the many statements of position and extensive testimony, each party was sufficiently aware of its opponents' cases to permit effective coverage in the initial brief subject to only minor supplementation later. The parties were also directed to include in their briefs the extent to which any claims of value to one T depended upon particular outcomes for other Ts. The proceedings in this case were briefly interrupted when PC and its affiliated Ts, including the New York & Long Branch, 50 percent of which was owned by CNJ, and the GPs entered into a settlement agreement on November 16, 1980.[4] The net liquidation value of all rail properties transferred and conveyed by PC and its affiliated Ts was stipulated to be $1.46 billion. The parties further stipulated the value of other benefits and compensable unconstitutional erosion (CUE), see CUE Opinion, 439 F.Supp. 1351, 1355 (Sp.Ct.R.R.R.A.1977), applicable to PC and its affiliated Ts to be zero. Each group of settling parties released essentially all its claims arising under the Rail Act against the other. The agreement provided that the Conrail securities, stipulated to have been without fair market value, to be received by the PC and its affiliated Ts were to be assigned to USRA. PC and its affiliated Ts thus received Certificates of Value with April 1, 1976, principal amounts equal to the stipulated net liquidation value, which were to be redeemed immediately. One frivolous objection was filed in response to the court's November 17, 1980, order to show cause why the agreement should not be approved; after hearing, the objection was *1204 promptly overruled, and the settlement was approved on December 5, 1980, Sp. Ct. Rep. N 23314.[5] This settlement, and the consequent disappearance of PC as a litigant which had taken the lead role for the Ts, complicated the task of the parties in preparing their opening briefs. In our separate December 5, 1980, order, Sp. Ct. Rep. N 23420, we therefore extended the date for filing initial briefs until January 12, 1981, and fixed April 20, 1981, as the date for reply briefs. The GPs' opening brief included 9 volumes of argument totaling almost 5000 pages, 49 maps, and a 32 volume supplemental appendix. The bankrupt Ts still in the case submitted 12 volumes of briefs accompanied by 52 volumes of appendices, in addition to various other submissions. Upon motion of the Ts, our March 24, 1981, order extended the due date for reply briefs until May 5, 1981. This time the GPs limited themselves to only 3 volumes and each T submitted only one, in addition to some additional volumes of appendices. Oral argument was set for June 8-10, 1981. We heard for a day each from the Ts and GPs, with the former presenting rebuttal argument on June 10, followed by a few brief comments from the GPs. Prior to the argument, we had indicated to the parties a number of issues upon which they should comment at argument. In addition, the court in constructing this opinion has had the benefit of written clarifications and elaborations of the parties' arguments and positions which it has requested, in each case affording opposing parties an opportunity to respond. On August 5, 1981, when considerable progress on the writing of this opinion had been made, the court received a petition of the GPs and the Reading Company, which had succeeded to the interest of the reorganization trustees of R, for approval of a settlement, Sp. Ct. Rep. N 36420. The agreement, which followed the general format of the PC settlement so far as practicable, settled the claims of R and five other companies for which it was (or expected soon to be) in a position to speak. The settlement for R and those five companies was for a base value of $87,497,489. With respect to several other lines which had not become bankrupts and in which R did not own a majority of the stock, the so-called NBIO's (non-bankrupt independent operators), the settlement contained a table of "Statutory Values", aggregating $33,475,095. These were sums at which the GPs were prepared to settle with each of these Ts. The Reading Company agreed to indemnify the GPs, subject to certain conditions, against any recovery by an NBIO in excess of its stipulated statutory value. A hearing with respect to approval of the settlement was had on September 2, 1981. Objection was made by North Pennsylvania Railroad Company (North Penn), the largest of the NBIO's, for which the settlement established a statutory value of $21,593,547. North Penn did not attack the reasonableness of the $120,973,484 fixed as the base value for the entire Reading System but sought an evidentiary hearing to determine a "fair and proper allocation" of that sum to North Penn. Concluding that North Penn had no right to such a hearing since it was left free to prosecute its own claim against the GPs, we entered orders approving the settlement and making certain provisions with respect to North Penn's prosecution of its claim, Sp. Ct. Rep. N 36847. Later North Penn decided to join the other NBIO's in entering into settlement agreements similar to R's at the "Statutory Values" set forth in the R agreement. No objection was made to these settlements.[6] II. Positions of the Parties Given the dependence of each T's prospects upon what would have become of *1205 other Ts in the absence of the Rail Act, as well as the substantial overlap of issues across the claims of various Ts, it is useful to present at the outset a summary of the major contentions of the parties. In this portion of the opinion we make no attempt at completeness and express no judgment on the merits. A. Joint Terminals Four Ts (PC, EL, R, and LV) developed a proposal, not envisioned in our earlier consideration of the case, that freight properties within what are described as the Philadelphia and Northern New Jersey/New York terminal areas would have been purchased and consolidated by public bodies in the absence of the Rail Act.[7] Public bodies allegedly would have been motivated by the necessity of keeping the freight terminals in operation and the desirability of rationalizing the freight transportation system through elimination of duplicative and obsolete facilities and coordination to reduce the facilities necessary to perform terminal services, and to enhance integration of freight operations with commuter and passenger service. These Joint Terminals (JTs) would have handled the switching, origination, and termination of all freight traffic within the area, interchanged cars delivered by main line trains, and provided service within the terminal areas, including service to and from the New York/New Jersey and Philadelphia ports. The Northern New Jersey/New York terminal (NJ/NYTerm) properties were located in New Jersey, and concentrated along the Hudson River and New York Harbor, but extended north as far as commuter service or the New York state line, west to the Pennsylvania state line, and south to the Raritan River, and also included lines going beyond that to Trenton, West Trenton, and Bay Head, New Jersey. The Philadelphia terminal (Philaterm) included PC's and R's freight properties in the Philadelphia area up to the Delaware and New Jersey state lines and extended west in Pennsylvania as far as the limits of the electrified commuter network. Both terminals had some operating responsibility beyond these boundaries. Under this scenario, the Joint Terminals Ts (JTTs) thought that public bodies would have purchased terminal freight properties for salvage value, with terminal operating costs to be recouped through switching charges covering the costs (other than capital costs) of services actually performed. Those JTTs with commuter lines argue in their individual cases that public bodies, not necessarily the same ones as those purchasing the JTs properties, would have acquired commuter and intercity passenger lines in the area at premiums in excess of nonrail use value. Both the GPs and the JTTs conducted JTs studies to determine how high the switching charges would have had to be to operate the terminals on a breakeven basis. The requisite switching charges calculated by the GPs are on the order of twice those presented by the JTTs. B. Penn Central Although PC is no longer in the case, it is still necessary to understand its proffers since the probable dispositions and earnings values of other Ts' lines are to some extent dependent upon what would have become of parts of PC. PC contended that its high-density freight lines that connected major metropolitan areas and provided interchanges with Western and Southern railroads would have been acquired by profitable railroads. Eight Intercity Freight Lines (IFLs) were to comprise this network: four major east-west routes — Chicago to New York (via Cleveland and Buffalo), Chicago to Pittsburgh, St. Louis to Cleveland, and St. Louis to Pittsburgh and Harrisburg with an extension to Perryville, Maryland, where it would join the "spine" of the Northeast Corridor[8] — and four shorter *1206 routes — Chicago to Detroit, Chicago to Cincinnati, Detroit to Cincinnati, and Youngstown to Ashtabula.[9] Individual Western and Southern railroads were identified as the most likely acquirers, each motivated by the fear of losing traffic if the IFLs were abandoned or purchased by competitors. Projections were also made for the possible acquisition of all the IFLs by a consortium of the Western and Southern railroads, as well as for purchase by the profitable Eastern railroads — Chessie and Norfolk & Western. PC contended that its Northeast Corridor properties would have been acquired by public bodies for continued rail use absent the Rail Act. Amtrak was identified as a likely buyer for the spine, the main line of the Northeast Corridor running between Boston and Washington, D.C. State, regional, and local bodies would supposedly have acquired commuter lines connecting to the spine, and, if the negotiations already begun with Amtrak did not come to fruition, portions of the spine as well. One aspect of PC's valuation case was that an acquisition agreement for the spine would have included free trackage rights for PC's freight traffic. Other Ts were affected by the destiny of the Northeast Corridor since some proffered scenarios involved freight traffic originating, terminating, or moving over some of these lines. Extensive evidence was taken in support of and in opposition to PC's claims. The issues are discussed at length in the GPs' opening brief, which was in an advanced stage of preparation at the time of the PC settlement, but the remaining Ts by and large have not addressed these questions. C. Erie Lackawanna In 1973, the EL system included 2,627 miles of track (exclusive of trackage rights) between Chicago and the NJ/NYTerm. Its principal line connecting the two points ran 970 miles; the system also included three feeder lines: from Marion, Ohio, to Dayton and Cincinnati; from Leavittsburg, Ohio, to Cleveland; and from Cuba and Hornell, New York, to Buffalo and Suspension Bridge, New York. Freight operations were conducted over the entire system, and commuter passenger traffic over 165 miles in Northern New Jersey. Two thousand ninety-five miles of EL's system were conveyed under the Rail Act. EL's major contention is that a substantial portion of its freight lines would have been sold to a profitable railroad for continued rail use. EL argued that the road's chief problem was a lack of cash, which purchase by a profitable carrier could remedy. Western railroads — Santa Fe, Burlington Northern, Missouri Pacific, Southern Pacific, and Union Pacific — needed to preserve their connections to the East and thus were principal candidates. EL claims that an acquisition package suiting a Western acquirer's needs would have included the EL main freight lines between Chicago and Jersey City, as well as feeder lines to Cleveland and Buffalo and heavy density branch lines — 1,741 miles in all, of which 1,455 were conveyed pursuant to the Rail Act. EL labelled this package the EL Western acquisition. EL contends, however, that the potential acquirers would have been more likely to purchase a combination of EL and LV lines. This second package, the EL-LV Western acquisition, would have included the LV main line between the NJ/NYTerm and its connection with EL's main line near Waverly, New York, in place of EL's line between NJ/NYTerm and Binghamton, New York, (via Scranton, Pennsylvania) — 1,538 miles of the EL system in all, of which 1,252 were conveyed. EL's development of the EL-LV combination assumed that both EL and LV freight lines in New Jersey would have been part of the NJ/NYTerm, although they also claimed that the earnings value of this combination would have been enhanced by inclusion of their individual *1207 New Jersey freight lines (other than those proposed for abandonment by EL) had the JTs not come into existence. EL argues that Chessie and Norfolk & Western, concerned by an invasion of the East by Western railroads, also would have had an interest in purchasing its east-west lines. The EL Chessie acquisition package includes the same EL properties as the EL Western acquisition except for the EL lines west of Sterling, Ohio, the junction between EL's and Chessie's main line to Chicago — a total of 1,382 miles of EL track, of which 1,315 were conveyed. A fourth package featured an EL-LV Chessie acquisition, with a configuration similar to the EL Chessie acquisition — 1,145 miles of EL's lines, of which 1,078 were conveyed. Similar configurations for acquisitions by Norfolk & Western were proffered, although not specifically developed in EL's financial submissions. The Norfolk & Western packages would have included EL's lines east of Huntington, Indiana, the junction with Norfolk & Western's system. EL's analysis shows the following earnings values for these lines: Acquisition Earnings Value (in $1000's) Western (EL) $225,908 Chessie (EL) $244,985 Western (EL-LV)/JT $278,077[*] Chessie (EL-LV)/JT $251,561[*] Western (EL-LV) $294,660[*] Chessie (EL-LV) $293,693[*] EL argues, however, that opportunity costs, including traffic lost if EL and other bankrupts were to shut down, traffic lost if competitors made the acquisition, reductions in car hire deficits, and other factors, see 1 ELF at 108-09, and New York's willingness because of the importance of EL's line into and through New York State to make $100 million available to a private acquirer to facilitate the purchase, would have induced an acquirer to pay more than earnings value. In particular, EL asserted that it would have received a higher price based on "net salvage value",[10] which it computed for its properties as follows: Acquisition Salvage Value (in $1000's) Western (EL) $370,653 Chessie (EL) $333,540 Western (EL-LV) $310,854[*] Chessie (EL-LV) $273,741[*] EL also contends that New Jersey would have purchased EL commuter lines (in addition to public acquisition of its JT area freight lines) for continued rail use due to their central role in New Jersey's commuter rail system. EL claims that the acquisition price would have included a 10 percent premium over net salvage value, reflecting a combination of $5.7 million for the difference between net and gross liquidation value and $4.9 million for "going concern value". The State supposedly would have been willing to pay these additional amounts based upon a fear that a condemnation award would be even higher and a desire to avoid the litigation costs of condemnation proceedings. D. Lehigh Valley The LV had operated a 430-mile main route between Buffalo/Niagara Falls and *1208 Oak Island Yard near Newark, and 425 miles of branch lines. Five hundred thirty-nine miles of these lines were conveyed pursuant to the Rail Act. In addition, LV leased 128 miles of the Lehigh & Susquehanna that paralleled LV's main line in Pennsylvania on the opposite side of the Lehigh River. These lines were also conveyed, and are included as part of LV's rail use proffers. LV suggested three possible rail use packages in which its properties would have been purchased in the absence of the Rail Act. The first involves sale as part of the EL-LV combination. This would have included LV's conveyed lines in Pennsylvania that continue from NJ/NYTerm and connect with EL just outside of Waverly, New York, near Sayre, Pennsylvania, as well as its lines from Sayre northward through Ithaca to Ludlowville, New York, and its terminal properties in the Buffalo area — 382 miles, plus 113 miles in the NJ/NYTerm area and the leased L&S lines in Pennsylvania. The earnings values of the EL-LV combination in the Western and Chessie acquisitions were included above in our discussion of EL's proffers. The net salvage values for LV's properties are claimed to be $39.6 million (the same lines are involved for either acquirer), in addition to $34.9 million for its lines that it claims would be included in NJ/NYTerm. At oral argument EL and LV agreed that an award for an EL-LV combination based on earnings value should be divided in accordance with estimated salvage values. Tr. at 67, 107. In addition to the EL-LV acquisition, LV claims that another possible outcome would have been the acquisition of its rail use properties, totalling 641 miles including the Lehigh and Susquehanna and LV's New Jersey properties (thus assuming the absence of a JT purchase by public bodies), as a single package by a large railroad capable of carrying freight between LV's lines and more southern and western gateways. This package would yield an earnings value of approximately $75 million in early 1974 and $95 million as of April 1, 1976.[11] Finally, LV argues that if none of the possibilities for private acquisition had materialized, 609 miles of its track would have been acquired at "non-earnings appraisal value" for the most part by the states[12] where its lines were located. The salvage value claimed for the New Jersey lines is the $34.9 million noted above in connection with valuation of the EL-LV combination. The lines in Pennsylvania were suggested to have a salvage value of $20 million and the New York lines of nearly $3.5 million. This proffer does not include LV's Buffalo area properties that it contends would have been kept in rail use as part of a Buffalo terminal system. Although included in the EL-LV capitalized earnings studies, as public acquisitions they would be valued as "specialized properties", a determination that has been deferred to a later phase in the case. Second Pre-Trial Order ¶ 12, Sp. Ct. Rep. at N 11802.[13] E. Reading As in the case of PC, despite the settlement of claims of all components of the Reading System, it will be useful to summarize R's case because of its bearing on those of other Ts. In 1973, R's lines covered 1160 miles, mostly in eastern Pennsylvania with some lines extending into Delaware and New Jersey. Its principal freight lines connected Reading, Pennsylvania, with Lurgan, Pennsylvania, (via Harrisburg), Reading with Allentown, Reading with Philadelphia, and Philadelphia and Allentown (the latter using *1209 trackage rights) with Bound Brook, New Jersey. R's passenger services were in the Philadelphia area. Eighty-eight percent of its lines were conveyed under the Rail Act. R claimed that 347 miles of its freight properties, designated "class (a)", would have been sold to a private purchaser in the absence of the Rail Act, at an earnings value of $200 million. Chessie was identified as the most likely purchaser; Norfolk & Western and the acquirer of PC's lines west of Harrisburg would also have been interested. "Class (b)" properties — 265 miles of freight and passenger properties — it alleged would have been purchased by unspecified public bodies: $67 million for the freight properties (as part of the JTs submission) and $61 million for its passenger properties. Memorandum Accompanying Dec. 31, 1978 Evidentiary Submission at 10-11. The latter claim included a 50 percent premium over net salvage value, which was predicated upon estimates based upon examinations of past negotiated acquisitions between public and private bodies, the social importance of the property, and arguments that public bodies would pay to avoid condemnation risks and costs. Finally, although R concurred in the latter portion of LV's arguments which R claims indicate the likelihood that its remaining conveyed properties, labelled "class (c)", would have been purchased for rail use by public authorities or groups of shippers, it does not contend that the acquisitions would have been at prices in excess of nonrail use value. F. Central Railroad of New Jersey In 1973, the CNJ system included 825 miles of freight and passenger lines within New Jersey, 756 of which were conveyed under the Rail Act. CNJ contends that it would have sold all its New Jersey rail assets (including those not conveyed) to the State of New Jersey for $100 million in the absence of the Rail Act. The CNJ's case emphasizes its value as perceived by a prospective purchaser, the State of New Jersey, as opposed to calculations of hypothesized earnings or scrap value performed after the fact by external sources. Its $100 million estimate is based upon discussions it had with state officials that were terminated as a result of our decision in the 180-Day Appeals, 384 F.Supp. 895 (Sp.Ct.R.R.R.A. 1974), holding that the CNJ should be reorganized under the Rail Act. The CNJ claims further support for this valuation based upon appraisals of its scrap value that ranged from $80 to $121 million. G. Ann Arbor The AA had operated approximately 290 miles of lines from Toledo, Ohio, just south of the Michigan border, through Ann Arbor to Frankfort, Michigan, where car ferries operated to two points in Wisconsin. Two segments of its lines were conveyed. AA claims that the Ann Arbor-Toledo segment, approximately 54 miles, would have been purchased by the State of Michigan for $9.7 million based on the line's earning value. With respect to the conveyed segment running 131.2 miles between Durand and Cadillac, Michigan, AA does not claim more than its value for nonrail use. H. Lehigh & New England In 1973, the L&NE consisted of two segments of line in Pennsylvania: the first, a 22-mile line from a connection with R and LV at Bethlehem to Uhler, Pennsylvania, interchanging with EL at Bath; the second, a 5-mile line connecting with R at Tamaqua, Pennsylvania. L&NE had been a subsidiary of CNJ and, while not abandoned when CNJ withdrew from most of its operations in Pennsylvania, no longer connected physically with any CNJ line. L&NE contends that all its property would have been acquired for continued rail use based upon its earnings value to a solvent carrier acquiring the above-mentioned connecting lines. The earnings value of the Bethlehem segment was estimated to range from $1.05 to $3.2 million, and of the Tamaqua segment from $1.2 to $2.45 million, depending on the acquirer. I. Government Parties The GPs advance a broad general denial of the Ts' claims. Their basic position is that, whether or not particular segments would have continued in rail use, the Ts would have realized no more than nonrail use values for their assets. Earnings values *1210 are claimed to be either negative or insufficient to justify offers in excess of scrap value, and public bodies would not have been compelled to offer premiums to acquire any properties. Moreover, the GPs proffer a "Master Liquidation Plan" wherein all the properties of the Ts would have been sold for scrap in a short period of time, with the resulting glut depressing the values of their properties. This so-called "benchmark" value for all Ts including PC was originally stated to be $630 million. At oral argument the GPs indicated that this would be revised upward to approximately $1.26 billion and that roughly 20 percent of this, equivalent to approximately $252 million, would provide "some notion" of the then nonsettling Ts' share. Tr. at 563. During oral argument, the GPs continued to emphasize their claim that the Ts still had not discharged their burden of showing non-speculative solutions to their predicament in the absence of the Rail Act that would have yielded values better than offered by the Master Liquidation Plan. Since the GPs' various contentions will be discussed at length in the course of our opinion, there is no need to detail them further here.[14] III. Introductory Analysis The CMV Opinion established that the Ts are entitled to compensation for the properties conveyed by them under the USRA's Final System Plan (FSP) for whatever they could have realized for them in the absence of the Rail Act.[15] Determination of that amount is the issue we must ultimately decide in this case. In this phase of the proceeding the issue is the more limited one stated in paragraphs (1), (2), and (3) of the Second Pre-trial Order, to wit, "the values that would have been received for disposition of ... [the Ts'] properties for continued rail use in the absence of the Rail Act". In so phrasing the issue for this phase of the case, the court had in mind a model, developed in considerable part from earlier proceedings, somewhat as follows: (1) The Ts would endeavor to prove that certain freight lines would have continued in rail use and would have had a realizable positive earnings value exceeding "scrap value", however "scrap value" might ultimately be determined. (2) The Ts would endeavor to prove that other lines, primarily passenger lines, although without a positive and even with a decidedly negative earnings value, would have continued in rail use as a result of purchase by public bodies and that these purchases would have been made at more than scrap value. (3) The GPs would contest the Ts' assertions as to the amount of mileage that would have remained in rail use. (4) In other instances and, as a secondary position in the cases described in (3), the GPs would contest the Ts' ability to have realized more than scrap value. (5) In cases where the court sustained the GPs' arguments described in (3) or (4), it would reserve the determination of scrap value to the second phase of the *1211 proceeding, which would also include the valuation of properties that the Ts conceded would not have continued in rail use. The proceeding did not develop that way. One complicating factor was the introduction of the concept of the JTs, with public bodies assumed to have purchased the predominantly freight lines at "net salvage value" and the predominantly passenger lines within the JTs at a premium above this. Another was the proposal of several Ts that lines would be sold for continued rail use at "net salvage value". This is a concept the components of which have not yet been fully explored and as to which different Ts may hold different views. Clearly, however, there is a wide gap between the Ts' concept of "net salvage value" and the GPs' concept of scrap value as embodied in the Master Liquidation Plan.[16] As a result of these factors and perhaps others, the vast record and the exhaustive briefs contain little discussion of which of the proffered lines would have been continued in rail use even if a T's estimate of earnings value were not realized.[17] In order to be sure that the court was not misapprehending the GPs' apparent unwillingness to challenge the Ts' proffers on the ground that lines without substantial earnings value would not have continued in rail use, the court, on May 26, 1981, addressed a letter to the Chairmen of the GPs' and the Ts' Steering Committees, Sp. Ct. Rep. N 36201. After noting that its reading of the briefs indicated that "the debate appears to be not so much over what lines would be continued in rail use but as to what the transferors would have obtained for them", the letter said that the court would be assisted by the GPs stating prior to the oral argument "which, if any, of the proffered lines the Government Parties contend would not *1212 have been continued in rail use regardless of sales price." The GPs' answer of June 3, 1981, Sp. Ct. Rep. N 35829, essentially accepting the court's characterization of what the debate had been, declined to furnish such a list. They stated that they "cannot and do not contend that any particular proffered line (indeed, any conveyed line) would certainly not have been continued in rail use regardless of sales price." The letter went on to say that this was not a concession "that the proffered lines would have been likely to continue in rail use in the absence of the Rail Act". It proceeded to argue that the GPs' evidence showed that "even the most valuable of the transferors' lines would have had negative or very low earnings values"; that accordingly "sales to private purchasers may well have been impossible for virtually all of the transferors' properties"; that, in contrast, "sales of some lines to public bodies could not be entirely ruled out"; but that this possibility did not matter because public bodies would not pay more than the Ts' best private marketplace alternative, to wit, sale at scrap value. In the absence of any adversarial position by the GPs with respect to continued rail use of any particular line, the court would be justified in considering that all proffered lines would have been continued in rail use except in those instances where the development of a T's own case demonstrated that if a particular acquisition were posited, portions of its railroad would have been abandoned.[18] This course would involve no procedural unfairness to the GPs;[19] the court's May 26 letter gave them an opportunity to argue which proffered lines would not have continued in rail use at any price and they chose not to make the gambit since in their view of the law it made no difference.[20] Despite this, the court is not inclined to abandon all inquiry into whether a particular line would have continued in rail use. The GPs did offer extensive evidence that public bodies would not have acquired proffered properties at any price, not because these were insufficiently important from a transportation standpoint but because the public bodies lacked legal power or financial ability to make the acquisition. We shall never be in a better position to pass on such arguments than we now are. Moreover, despite the GPs' unwillingness to engage in debate over whether lines would or would not have continued in rail use, we cannot ignore the Ts' argument that in weighing the likelihood or unlikelihood of their proposed dispositions, the court must consider that the discontinuance of essential rail service was the most unlikely outcome of all, *1213 with the consequence that in the absence of the Rail Act some basically nonfederal solution to the problems of rail service in the Northeast would have been found. The foregoing and other developments have caused the court to be increasingly concerned with what it may properly decide in this opinion. This concern led us to begin the final day of the argument by addressing a series of questions on this subject to counsel for both sides. Tr. at 408-16. Our questioning made clear that the parties were using terms sounding like scrap value in many different senses. We shall start at the bottom, with the GPs' Master Liquidation Plan, which would price all the Ts' properties on the theory that all properties of all Ts, including those that would have continued in rail use, and even some of those so continued because of high earnings value if such there were, would have been sold for scrap at substantially the same time. Another model, characterized by CNJ as "the residual scrap model", CNJRB at 30, "would price the property remaining in rail use by looking at a market affected by the simultaneous scrapping of all properties that would be withdrawn from rail use".[21] A third is what CNJ calls the "stand-alone" model, wherein the scrap value of any T's property would be computed without regard to the fate of the property of other Ts. Presumably this could encompass pricing either on the basis of a scrap sale of all properties or of only those that would in fact have been withdrawn from rail use. Highest on the scale is the net salvage value theory, already discussed, see note 16, supra, and accompanying text, which deducts only some of the costs of sale. One point that is quite clear to us and, indeed, seems to be agreed upon by all the parties except CNJ is that at this stage of the proceeding we cannot properly decide among these theories or endeavor to apply the chosen one or ones. In ¶ 1 of the Fourth Pre-Trial Order, filed May 16, 1979, we granted the motion of the GPs "to the extent it seeks to defer the consideration of evidence with respect to determining non-rail use values of any properties or appraisal methodologies for determining non-rail use values of any properties", Sp. Ct. Rep. at N 17130-31. Properly relying on this, the GPs did not cross-examine or rebut statements by the Ts' witnesses relating to the Ts' net salvage value, appraisal value, etc. On the other hand, in the same paragraph of the Fourth Pre-Trial Order we directed that "evidence that purchasers of lines sold for continued rail use would have paid prices established in whole or in part with reference to various different types of non-rail use values shall be presented during the rail use phase of these proceedings". Much such evidence was presented, briefed, and argued.[22] We believe that judicial economy would be promoted by deciding, at this stage of the proceeding, which properties would have been acquired at prices related to their nonrail use value and whether or not the acquirer would have paid premiums above or negotiated deductions from scrap value. For simplicity we shall sometimes refer to the values that we will determine in the second phase of the proceeding simply as X.[23] We shall expect the parties to give us maximum assistance in expert testimony, briefs and arguments on what now seems the most important "cut-across" issue in the nonrail use phase, namely, what properties should be considered *1214 in determining "glut" for a particular T. See Sixth Pre-Trial Order, ¶ 7, Sp. Ct. Rep. at N 35968. We recognize that this solution does not provide as neat a distinction between the first and second phases of this proceeding as we had anticipated, but it seems dictated by the manner in which the case has unfolded. IV. General Background Considerations The court is confronted in this case with a task of almost insuperable difficulty. That is to determine what action would have been taken by interests of all kinds — the Ts, other railroads, state and local public bodies, the United States,[24] the ICC, labor, and shippers, in order to avoid a shutdown of the Ts (including PC) which would have entailed disaster for all. As witness Charles Shannon[25] said, testifying for the PC, "any attempt to determine how the Northeast rail crisis might have been resolved if Congress had not enacted the Rail Act in 1973 is an essentially unrealistic enterprise", since "[t]here was never any realistic likelihood that the federal government would allow market forces to play a prominent role in the way in which rail service in the Northeast would be restructured". Shannon (S) (PC) (Dec. 31, 1978) at 2. The GPs come close to suggesting that this unreality should foreclose our consideration of the Ts' proposals concerning what might have occurred. They repeatedly call our attention to the statement in Olson v. United States, 292 U.S. 246, 257, 54 S.Ct. 704, 709, 78 L.Ed. 1236 (1934), that [e]lements affecting value that depend upon events or occurrences which, while within the realm of possibility, are not fairly shown to be reasonably probable, should be excluded from consideration .... However, that sentence must be read along with the previous statements that in respect of each item of taken property the court must estimate market value, deemed to be the sum which, considering all the circumstances, could have been obtained for it; that is, the amount that in all probability would have been arrived at by fair negotiations between an owner willing to sell and a purchaser desiring to buy; that [i]n making that estimate there should be taken into account all considerations that fairly might be brought forward and reasonably be given substantial weight in such bargaining; and finally that [t]he determination is to be made in the light of all facts affecting the market value that are shown by the evidence taken in connection with those of such general notoriety as not to require proof. Moreover, Olson involved a taking for a new use which the Court concluded no person other than the condemnor could accomplish and the statement relied upon by the GPs was addressed to that situation. While we agree with the GPs that the Ts bear the burden of proof, this does not require them to show with something like mathematical certainty that hypothetical transactions would have occurred and hypothesized prices exactly realized. Their burden is simply to show by a preponderance of the evidence that transactions such as those *1215 proposed by them were more likely to have occurred than what the GPs argue was the Ts' next best alternative — abandonment and subsequent dismantlement. When they have satisfied that burden, the court must ultimately determine, as best it can, what the price would have been.[26] The GPs begin their assault on the Ts' proposals, more particularly their freight line proposals, by painting a dismal picture of the conditions of the railroad industry, especially in the Northeast[27] as this appeared in 1973. A summary of this will suffice. For the facts the GPs relied heavily on a staff report of the Senate Committee on Commerce, dated December 1972, entitled The Penn Central and Other Railroads. They cite the report, 1 GP at A 176-77, as concluding that railroads in this area suffered from [s]teadily declining freight volume, particularly in traditionally rail-carried bulk commodities such as coal, steel and forest products; outdated terminals that could not efficiently handle current volume, let alone any hoped-for increased volume; service problems, including an unbelievably high incidence of damage to shipments and inability to meet delivery schedules; steadily increasing competition from the more efficient motor and water carriers; deteriorated physical plants; outdated work rules and large job protection commitments; concerns of shippers, local governments and unions that would inevitably delay abandonments of unprofitable lines; and continued losses from fulfilling obligations to provide passenger service. A significant source of the industry's problems was the relative decline in share of the gross national product represented by bulk items as distinguished from services. Hence railroad shipments and revenues could fall even when the general economy was growing. Shifts in population away from the Northeast and the relocation of industrial plants into suburban industrial complexes and to points closer to the new population centers operated further to the disadvantage of the Northeastern railroads, many of which were left with substantial amounts of little used freight track into cities. For short and even for longer hauls of manufactured products, the railroads, because of lesser reliability and greater proneness to loss and damage, were unable to compete successfully with trucks. The GPs cite a gloomy report of the PC Trustees dated September 17, 1971, to the PC reorganization judge on the basic problems of the Northeastern carriers; on July 24, 1972, the LV Trustees rendered an even gloomier report to the same judge, claiming that, however bad PC's situation might be, PC was managing to maintain its strength at the expense of its smaller competitors. Declining earnings had led to deferral of maintenance, which led in turn to a further decline in service and revenue. The GPs' brief extensively recites statements by PC trustees as to the devastating effect of the number and expense of the freight terminals it was obliged to operate. On the other hand efforts in 1972 and early 1973 on the part of the R, LV, and CNJ Trustees to consolidate some operations in Northern New Jersey came to nothing, in part because representatives of the State were concerned over the impact of immediate large scale layoffs on organized labor. Antiquated work rules constituted a heavy burden, and the threat of labor protection payments clouded proposed shrinkage and mergers. Relatively little progress had been made in the way of abandonments. For many of the Ts, including EL and CNJ, as well as *1216 PC and R, passenger service created a further burden since state and local authorities were unwilling to subsidize more than avoidable costs. The results of these developments were summarized in two tables in the GPs' opening brief, 1 GP at A 204, 206, comparing net railway operating income and freight operating ratios for five of the Ts for 1966 and 1972. Net Railway Operating Income 1966 1972 EL $17 million ($7.7 million)[*] R $7.7 million ($12.1 million) LV ($1.4 million) ($8.9 million) AA ($288 thousand) ($1.1 million) CNJ ($4.9 million) ($1.2 million) Freight Operating Ratios 1966 1972 EL 72.2% 83.3% R 77.4% 89.5% LV 84.2% 93.5% AA 86.3% 93.1% CNJ 89.1% 92.4% The decline in net railway operating income was reflected in a decrease in cash flow and, with the accompaniment of inflation which made depreciation charges on equipment inadequate to provide for the cost of new equipment, required the railroads to incur even larger obligations, whether in the form of debt or leases, for the acquisition of new equipment. The average rate of return for railroads in the Eastern District, which includes profitable railroads such as Chessie and Norfolk & Western, from 1969 to 1973 was a mere 0.14 percent. Smith (S) (Jan. 30, 1980) Ex. 17 at 16. The GPs also call our attention to reports of trustees of the Ts to their respective reorganization courts in 1972 and 1973, many of which are cited in our CUE Opinion, 439 F.Supp. at 1359-62 (PC), 1381-82 (CNJ), 1383 (R), 1384 (LV), 1386-87 (EL), as to imminent cashlessness and the consequent need to shut down the railroad. Citing from our CMV Opinion, 445 F.Supp. at 1088, the GPs conclude, 1 GP at A 211-12: It is difficult to envision against the picture painted in this Section — largely from published data from rail sources and from other contemporaneous public documents — that profitable railroads would have found it "in their `own self interest'" to have paid large sums for the privilege of standing in the shoes of the transferors. It was clear in the early days of 1973 that none of the transferors could be reorganized on a conventional private income basis. They were faced with outworn physical facilities, heavy debt, negative earnings, inability to raise astronomical sums of essential capital, adverse economic trends beyond their control and operating problems which had been unsolved for decades. They argue from this that "no rail use values inhered in the transferors' properties" and that they "are entitled to recover no more in these proceedings than the realizable nonrail use value of their properties". The Ts do not deny that the 1973 picture for the Northeastern railroads was sombre. Their answer to the GPs is rather, to use Mr. Justice Holmes' phrase in his dissent in Lochner v. New York, 198 U.S. 45, 76, 25 S.Ct. 539, 547, 49 L.Ed. 937 (1905), "General propositions do not decide concrete cases". As the case now stands, the only instances where we must determine the possibility of sale on a capitalized earnings basis are greatly pared-down versions of both EL and LV; the short Ann Arbor-Toledo segment of AA; and L&NE. EL urges, 1 ELF at 27-34, ELRB at 14-22, that in 1973 it had one of the highest density ratings in the nation and the longest average haul of any carrier in the East; that its lines were in relatively good condition; that it had been steadily increasing its share of the eastern market, was well managed, and enjoyed excellent relations with shippers and labor. EL had rebounded from the adverse effect of the PC merger and produced freight net railway operating income of $11 million in 1971. It entered bankruptcy only in 1972, in consequence of the devastation wrought to it and many of its shippers by Hurricane Agnes. In an order entered on April 3, 1974, pursuant to § 207(b) of the Rail Act, the judge in charge of EL's reorganization found that *1217 EL was reorganizable on an income basis pursuant to § 77 of the Bankruptcy Act. He relied upon a statement of position by the EL trustees "supported by comprehensive evidence, financial projections and economic feasibility studies by the highly respected Salomon Bros., banking and financial consultants, and Wyer, Dick & Co., railroad consultants". Plans for EL included the rebuilding and rehabilitation of its yard at Marion, Ohio, which was expected to pay for itself in two years and for which Salomon Brothers had already obtained financial commitments aggregating 80% of the cost, and also its yard at Croxton, New Jersey. EL contends that the only weakness in the plan it presented to the reorganization court was "a dangerously tight cash position in 1974 and 1975". 1 ELF at 33. This weakness became fatal as a result of the recession of 1974-75, accompanied by continued inflation. In an order dated January 18, 1975, the reorganization judge carefully reviewed these unexpected developments and concluded that EL no longer had the ability to continue reorganization on an income basis. However, EL contends, these adverse developments would not necessarily have precluded a solvent railroad from believing in 1973 and later years that EL, while perhaps unable to survive as an independent railroad, would have been attractive as part of a larger system if shorn of the burden of passenger deficits and unprofitable branch lines, and, as we shall see, EL submitted substantial evidence to that effect. LV likewise contends that its admittedly bad condition prior to the Rail Act is not inconsistent with its claim that the proffered portion of its lines would had have earnings value as part of a larger system. It points out that from 1962 until PC's and its own bankruptcies in 1970 it was under the control of the Pennsylvania Railroad and, after the 1968 Pennsylvania-New York Central merger, of PC, both of which were competitors; that, as predicted by the hearing examiner on the Pennsylvania-New York Central merger, this caused substantial diversion from LV; and that its Trustees had had significant success in rebuilding its traffic base. Beyond this many of LV's problems were attributable to its relatively short mileage, its longest possible haul being 443 miles between Jersey City, New Jersey, and Buffalo, New York, with the result that it received a small proportion of total revenues while bearing heavy terminal expense. This problem would have been alleviated by LV's inclusion in a larger system. Moreover, the portion of LV conveyed pursuant to the Rail Act, most of which was here proffered for continued rail use, was, in LV's words, LV at 16, "a very different system than the one the Trustees operated before conveyance". The FSP did not designate LV's main line west of Van Etten Jct., New York, or various branch lines in New York and Pennsylvania which "produced substantial costs and virtually no income", id. AA claims, somewhat similarly, that the unprofitability of its 298-mile rail system as a whole casts little light on the earnings value of the 54.2-mile Ann Arbor-Toledo segment, which is all that is at issue here. Doubtless because of the special nature of its case, discussed in Part XIII of this opinion, CNJ makes no particularized reply to the general arguments of the GPs here under consideration. The upshot of all this is that while the general arguments of the GPs being here considered suffice to create a considerable degree of scepticism over the Ts' assertions of earnings value, they do not relieve this court of the onerous task of detailed analysis of the Ts' proposals. If a T has presented a set of figures produced by competent witnesses which show significant capitalized earnings value, we cannot simply reject the proposition that the T might have obtained more than X from a solvent railroad, no matter how dismal the record of the past had been. We therefore turn from the background quite properly presented by the GPs to the specific claims advanced by the Ts. V. Bargaining for Properties Having Earnings Value Establishment by a T of a capitalized earnings value higher than X does not necessarily mean that the T would have been *1218 able to realize this in the marketplace. It might not have been able to realize that much; conceivably, if the "earnings" that were capitalized were simply those on the purchased line and other elements of value to the purchaser were not taken into account, it could have gotten more. As indicated in our CMV Opinion, 445 F.Supp. at 1044, the amount would have depended on a process of bargaining — a process which, the GPs insist, would have included not only the buyer and the seller of the particular property but many other railroads and public bodies on whose purchase of losing properties the earnings value of others would depend. The parties produced two distinguished academics as witnesses on the subject of bargaining. The GPs called Donald Jacobs, Dean of the Graduate School of Management at Northwestern University; his direct testimony is contained in a volume entitled "The Nature and Consequences of Likely Bargaining among Interested Participants in the Alternative Scenario". PC called in rebuttal Thomas Schelling, Professor of Economics at Harvard University and Professor of Political Economy at its John F. Kennedy School of Government. Although his testimony was addressed primarily to the IFLs, many of his observations apply also to the properties here before us. In addition Shannon's testimony for EL and PC and Professor Alain Kornhauser's[28] testimony for PC entitled "Opportunity Costs for Potential Purchasers of the Intercity Freight Lines", contain a considerable amount of discussion about bargaining. Professor Schelling did not disagree with some fundamental principles stated by Dean Jacobs, which we find it useful to set out. The most basic of these were: 27. The fundamental principle of bargaining is that each party seeks through agreement to improve his position over that which it would occupy in the absence of agreement. A corollary is that parties assess the motivations and bargaining strengths of opposing parties with reference to those parties' alternatives. Parties will reach agreement if they believe that doing so will leave each of them at least as well off, and at least some of them better off, than they would have been in the absence of agreement. * * * * * * 29. In bargaining between a buyer and seller regarding a proposed sale of property, agreements are reached only if the property is at least as valuable to the buyer as it is to the seller. Only in such instances would a rational buyer be willing to pay enough to induce the seller to part with the property. If the seller is rational, it would not be willing to sell the property to any buyer for less than the value it could otherwise realize (by keeping the property or by selling it to another buyer). By the same token, a rational buyer would never pay more for a property than the value of that property to it (or more than it would have to pay to buy a substitute property of equivalent value to it). The maximum amount that the buyer would rationally pay for the property is called the buyer's "reservation price," and the minimum amount that the seller would rationally accept for the property is called the seller's "reservation price." Thus, if the buyer and the seller are rational, they will not agree on a sale of the property if the buyer's reservation price is less than the seller's reservation price. Jacobs (S) (Nature and Consequences) (Nov. 1, 1979) at 11-13. Dean Jacobs called the excess of the buyer's reservation price over the seller's reservation price "excess value". He testified also that account must be taken of the ability of third parties to postpone, prevent, or have disadvantageous conditions placed on the sale. Amounts required to be spent, whether by the buyer or the seller, to obtain the assent of such third *1219 parties would reduce the amount otherwise realizable by the seller. It is worth noting at this point that ambiguities as serious as those to which we have already referred, see p. 1211 & n.16, 1213, supra, with respect to the parties' use of such terms as scrap value, salvage value, etc., which would constitute the seller's reservation price since the seller could obtain this amount without making the sale, are also present in the parties' references to earnings value. We find scattered through the record the following concepts of the earnings to be capitalized or otherwise taken into account; we have arranged these generally in ascending order: (1) The assumed future earnings of a line conceived as carrying essentially the same traffic as it did before the acquisition, increased or decreased only in accordance with estimates of general economic conditions. (2) The earnings reflected in (1) plus cost savings resulting from the acquisition.[29] (3) The earnings reflected in (1) and (2) plus earnings the buyer would have lost if the seller had ceased operation.[30] (4) The earnings reflected in (1) and (2) plus earnings the buyer would have lost if all the bankrupt carriers had ceased operation.[31] (5) The earnings reflected in some or all of the preceding plus earnings on traffic from or to points on the acquired road which the acquisition would permit the buyer to divert to its own line.[32] (6) The earnings reflected in some or all of the preceding, plus earnings on traffic which in the absence of the acquisition *1220 would not have moved by rail at all.[33] A capitalization of "earnings" that includes items at the upper end of this scale will produce a higher buyer's reservation price and consequent excess value than those at the lower, just as use of the Master Liquidation Plan's approach to scrap value rather than the Ts' various salvage value approaches will produce a lower seller's reservation price. Yet apparently neither Dean Jacobs nor Professor Schelling had been made aware of the ambiguities in what they conceived as the reservations prices of the two parties. Their conclusions, more particularly some of Dean Jacobs', must be read with this in mind. Thus, while one may readily accept that a buyer would not pay more than capitalized earnings value if that term included all the elements in the preceding scale, it is by no means clear that a buyer would not go beyond earnings value computed as outlined in (1) or (2). Despite this we are inclined generally to accept Dean Jacobs' conclusion (although not all his reasons for it) that the seller would realize a relatively small share of "excess value" if there is only one potential buyer. The only leverage possessed by such a seller would be the threat to close down operations. We do not minimize this. Table B to Shannon's EL statement showed, for example, that assuming a 70 percent loss to Santa Fe of the 1973 traffic handled between Santa Fe and EL by a shutdown of EL alone and a 60 percent incremental cost ratio for handling the traffic, Santa Fe would have suffered an annual loss of $5,780,516 in pretax net income.[34] However, the 70 percent estimate appears to be on the high side, especially if one or more of PC's lines east of Chicago remained in operation.[35] Moreover, a buyer offering something more than X, particularly if this were X in one of its higher ranges, would find the seller's threat to refuse this and close down operations lacking in credibility. For one thing, if the offer were high enough in terms of X, the ICC would have refused to permit abandonment (in the sense of dismantlement) pursuant to its established policy, see note 16, supra. Apart from this the trustees of few railroads of any size or the judges presiding over their reorganizations would have been perceived as having the moral stamina to force an actual shutdown, thereby foregoing any chance ever to realize more than had been offered[36] and becoming the subject of obloquy from public officials, shippers, and labor, when they had on the negotiating table an offer for something more than could be obtained by scrapping the road. Our view that in the single-bidder situation the seller would not receive much over X must be qualified, however, by an important caveat, namely, that neither buyer nor seller would have known just what X would be. This would not have been so much for lack of information; several, perhaps all, of the reorganization courts had directed appraisals of the bankrupts to be made, and many of these were available in 1973. It was rather that no one knew how these appraisals would translate themselves into reality. What counts is what the buyer would have perceived the seller's likely receipts from dismantling or condemnation to be. Saying that in the absence of a competing bid the seller would not have received much of the excess value does not *1221 mean therefore that he might not have obtained considerably more than the prices contained in the Master Liquidation Plan. Before considering what the division of the excess value would have been under competitive bidding, it is necessary to address the question whether such bidding would ever have existed. The GPs argue that it would not, since the interests of prospective bidders, especially if these were railroads, would have led them to join in bidding as a consortium. The motivation for this would have been not only to buy at a cheaper price but to avoid the disruption of traffic patterns that would have resulted, e.g., if different Western roads had purchased various IFLs and EL. While any such purchases would have afforded opportunity for diversion, the benefit of this might have been decreased or nullified by similar opportunities afforded by similar purchases by others, see note 32, supra. We ourselves introduced the notion of purchase by a consortium when, in the CMV Opinion, we said, "we do not at all exclude such possibilities as joint action by the transferors in the Northeast Region or joint purchases by profitable railroads to the west and south which ordinarily might have been deemed objectionable under the antitrust laws", 445 F.Supp. at 1010. Although the prime thought behind our suggestion of a joint purchase was to overcome problems of lack of adequate financial resources,[37] we also had in mind the elimination of the opposition before the ICC and the endless battles over protective conditions and inclusion that have spun out the length of many railroad merger proceedings — sometimes to the point of killing the merger. Clearly § 5(11) of the Interstate Commerce Act, 49 U.S.C. § 11343, authorizes the Commission to confer antitrust immunity on carriers who have joined in purchasing another railroad, and the Supreme Court has approved its doing so, Minneapolis & St. Louis Ry. v. United States, 361 U.S. 173, 80 S.Ct. 229, 4 L.Ed.2d 223 (1959).[38] Despite this we have the gravest doubt whether the Commission would have authorized a combination of all prospective bidders to purchase all the lines here proffered. No one has contended that the potential bidders would not have had resources adequate to effect the proposed purchases.[39] Hence approval of a consortium purchase could not be defended on the basis that it would not depress the market since otherwise there would have been none. Insofar as an objective of the consortium would have been to lower the price the Ts would have received, we do not see why it would not violate the antitrust laws just as does bid-rigging which results in the purchaser having to pay more, see United States v. Koppers Co., 652 F.2d 290 (2nd Cir. 1981); the antitrust laws are as concerned with sellers as with buyers. To be sure, § 5(11) empowers the Commission to grant exemption from the antitrust laws, and it can be argued that the Commission would have been predisposed to do this since its primary interest lay in the continuation of rail service at the lowest cost. As against this the Commission would have realized that investment in railroad securities would hardly be promoted by its endorsing a policy that if a railroad has to be liquidated, prospective competing buyers can permissibly join forces to depress the purchase price. As to the alleged pro-competitive factors in a consortium acquisition through preserving established traffic patterns, we would think that, however forceful these considerations may have been in the case of a short-line "bridge carrier", such as the Toledo, Peoria & Western, which had observed and was *1222 pledged to continue to observe a policy of "strict neutrality between all connections", Minneapolis & St. Louis Ry., supra, 361 U.S. at 189, 80 S.Ct. at 239, the situation would appear quite different if a half dozen or more carriers were buying the properties at issue here. Individual acquisition of various lines would open up possibilities of more aggressive competition, not simply by more energetic sales effort but because of the possibility of more efficient single line service; this is hardly to be disregarded because competition for shipments which had been and would continue to be handled interline between the acquirer and property acquired by another might be disadvantaged. Beyond all this we do not readily foresee the feasibility of a six or eight party management practicing a policy of strict neutrality over a system as large as this — a situation altogether different from the common joint ownership of local terminal railroads. The possibilities for friction would be endless; management might have to spend as much time quieting the sensibilities of the owners as running the railroad. In particular we perceive no likelihood of a consortium consisting of both Western and Eastern lines, since their objectives would have been directly opposed. In sum, while we cannot categorically say that the ICC would not have approved acquisition by a consortium, we think the odds would have been overwhelmingly against this outcome.[40] For these reasons we find it unnecessary to consider whether introduction of the concept of a consortium would be inconsistent with basic notions of eminent domain law with respect to a free market. Once the hypothetical transaction is visualized in terms of competitive bidding, the basis for Dean Jacobs' prophecy that little excess value would find its way into the hands of the Ts largely breaks down. We see no force in the GPs' contention that a railroad having an interest in acquisition would not enter the contest if it perceived that another would be likely to bid more. This assumes a degree of information about the intentions of other parties which normally does not exist. Beyond that, a railroad, even if not expecting to be successful, might have an interest in driving up the price a competitor would have to pay. The GPs correctly point out that we must consider not only competition among buyers but also competition among sellers. While no one can do more than guess, we would suppose that in a two-bidder situation, after making deduction for payments that would be required to be made to third parties, the road having the lesser interest would have bid well up toward its reservation price and the road with the greater interest would have topped this by a relatively small amount. VI. The Effect of Inflation on Earnings Forecasts The GPs, without quantifying the amount of the adjustment urged to be required, attack the Ts' projections because of their failure to allow for the allegedly detrimental effect of inflation on earnings. The GPs presented their evidence through a qualified witness, Russell Murphy.[41] *1223 Murphy's thesis was that between 1967 and April 1, 1976, three separate influences had caused railroad earnings to suffer as a result of inflation and would be expected to continue to do so. These were as follows: (1) The railroad industry voluntarily chose not to petition the ICC to recover the entirety of its inflated costs. (2) The ICC awarded the industry less rate relief than the industry ultimately sought — partly in the interest of shippers and partly because it believed the requested increases would cause such loss of traffic as to be counterproductive. See, e.g., Increased Freight Rates and Charges, 1975, Nationwide, Ex parte 310, 349 I.C.C. 555, 575 (1975); Increased Freight Rates and Charges, 1973, Nationwide, Ex parte 295, 344 I.C.C. 589, 637 (1973). (3) The industry, primarily because of competition from other modes of transportation, did not implement fully the rate increases that had been granted.[42] Examining the 14 general rate increase proceedings conducted by the ICC between 1967 and April 1, 1976, he found that although the roads reported inflated costs of $10.614 billion, the roads had sought increases of only $8.255 billion, and had been granted increases of only $8.229 billion. However, because of a variety of factors — fears of loss of traffic to other modes of transportation, delay in obtaining permission to increase intrastate rates, and "accounting delay", the railroads realized only some 82 percent of the increases authorized.[43] This picture, however, had somewhat improved after 1973. In the four general rate increase authorizations for which data were reported by the industry before April 1, 1976, an average revenue yield of 92.5 percent of authorized increases was realized. R. Murphy (S) (Jan. 30, 1980) at 35-39. The portion of EL's opening brief dealing with this subject largely contented itself with relying on testimony of James White, Chessie's Director of Cost and Economic Analysis and author of the Chessie study of EL and R, that Chessie used constant dollars because "we generally found from studies that freight revenues generally inflate, in some cases, slightly greater, but not significantly greater than do costs", White (D) (July 20, 1977) at 365-66, EL App. K at 146-47; of John Darling, Director of Cost Analysis and Research for Santa Fe, who said that use of constant dollars was "an appropriate method of financial analysis in the railroad industry", Darling (D) (July 31, 1980) at 28, EL App. K at 150; and Frederick Smith,[44] the GPs' chief valuation witness, who acknowledged that he had performed a constant dollar analysis for his employer, First Boston Corporation.[45] See 2 ELF at 410-12. *1224 In its reply brief EL made a more extensive answer, ELRB at 172-85. This starts with an assertion that Murphy's data with respect to increased costs were based on a summary prepared in 1979 by the Association of American Railroads (AAR), which seems to have relied on the railroads' submissions in the increased rate cases. In testing his model Murphy compiled actual cost data for the railroad industry for the period 1966-76. This allegedly demonstrated that for the period preceding the Ts' proposed 1974 disposition date, 1966-73, the actual cumulative cost increases experienced by the railroad industry totaled $5,373 million, as compared to the $6,965 million reported by the AAR. The cumulative revenue increases authorized by the ICC for the same period totaled $5,573 million, so that these more than recovered the railroad's increased costs.[46] EL supports this by citations to various ICC proceedings in which the Commission found the railroad's cost claims, which were those in the AAR data, to be overstated.[47] It goes on to make three criticisms of the portion of Murphy's study concerning the relationship between increased costs and rate increases, the most important being that the study underestimated total rate increases because it took into account only general rate increases achieved in nationwide Ex parte proceedings, ignoring annual volume rates which are frequently subject to automatic escalation clauses and other specific tariff filings. EL contends that Murphy's conclusion that the roads did not obtain rate relief equal to their increased costs thus cannot stand. Turning to Murphy's final point, the alleged failure of the railroads to realize the full rate relief accorded by the Commission, EL contends that the Commission took this into account in granting rate relief. Admitting that the ICC has concluded that the railroads realize only some 82 percent of the rate relief authorized,[48] EL says that the Commission grants increases such that 82 percent realization would meet increased costs, see Ex Parte 265 & Ex Parte 267, supra, 339 I.C.C. at 177-78; Ex Parte 357, 359 I.C.C. 740 (1978). In addition EL called attention to a report of a USRA panel, of which Murphy was named as a member, which minimized the fear of diversion to trucks as a factor that would prevent Conrail from raising its rates to meet inflation by saying that "the effect of inflation on the cost levels of the two modes was essentially the same". R. Murphy (D) (May 19, 1980) Ex. 52 at 4, EL App. O at 640; ELRB at 182. We believe that EL has satisfactorily answered Murphy's claims.[49] It is clear that, at least in the cases which a buyer would have regarded as most relevant, the ICC granted increases in a nominal amount greater than the railroads' increased costs because of its perception that only a fraction, say 82 percent would be realized. Taking, for example, the figures in note 46, the yield on authorized rate increases would be 82 percent of $8,229, or $6,748 million leaving a shortfall of $725 million. This is not a stupendous amount for the entire railroad industry over a ten year period. Even this estimate is undoubtedly too high since it does not include any allowance for specific tariff filings and automatic escalation clauses. Finally, Murphy does not and cannot answer the testimony of White and *1225 Darling, executives of railroads who were potential acquirers of EL and LV, to the effect that inflation was not a serious concern. VII. Valuation Methodology There are two components to be considered in valuing an asset: the amount of cash it generates and the discount rate. The first component is obviously important — an investment yielding $100 a year is worth twice as much as one yielding $50 a year. The second component is equally important — $100 now is worth more than $100 ten years from now. All the parties agree that these are the two relevant concerns. Where they disagree is on how to measure them. A. DCF vs. EBIT The parties used essentially two different methods for valuing the earnings claimed to have been generated by the Ts' properties. The GPs, PC, and, with modification, EL determined the present discounted value of the distributable cash flow (DCF) generated by the enterprise. Under this approach, one first projects DCF for future years. DCF is essentially earnings plus after-tax interest, principal payments, and depreciation minus taxes and capital expenditures needed to maintain earnings.[50] As its name implies, DCF represents the funds that can be distributed to all investors, including bondholders, without impairing the earning power of the enterprise. The next step is to discount DCF for future years to present discounted value (PDV) using an appropriate discount rate.[51] AA capitalized earnings before interest and taxes (EBIT) for a normalized, or average, year. L&NE capitalized earnings before interest but after taxes, assuming a 50 percent tax rate. EBIT represents what its name implies — earnings with interest, principal payments, and taxes added back in. This figure is then capitalized using an appropriate interest rate on the assumption that the enterprise will generate normalized (an average year's) EBIT in perpetuity.[52] On balance, the court believes that the approach presented by the GPs' expert witness Smith[53] is preferable to AA's and L&NE's method. The advantages of using DCF are highlighted by examining the three key differences between the two methods: the effect of timing, the impact of taxes, and the treatment and estimation of capital expenditures needed to maintain income. 1. Timing The first major difference is that DCF relies on a series of annual projections while EBIT uses a single average year. Smith argues that by using annual projections rather than a single average year, it is easier to assess the effect of the forces that affect earnings. While we are not convinced that trying to assess the impact of these trends on twenty numbers is necessarily *1226 easier than assessing them on one, the use of annual projections does have the major advantage of allowing one to take account of the effect of time on present value. As a rule, the earlier one realizes income the better. The use of a single normalized year blurs the effect of the timing of the receipt of income on the capitalized value of an enterprise by assuming that earnings will be realized in a constant flow. If one is able to project that some years will be better than others, while the average earnings for two companies could be the same, the capitalized value of their earnings streams could be quite different. An illustration of this is given in the margin.[54] Of course, the GPs' witness did not assume that an enterprise would disappear at the end of twenty years. Smith assumed that the final year's cash flow would represent the expected normalized year's experience and calculated the capitalized value of that figure. The result, called the terminal value, is then discounted to PDV.[55] The total value of the enterprise is equal to the PDV of the annual cash flow projections plus the PDV of the terminal value. While the GPs' witness thought that twenty or thirty-year cash flow projections would be optimal, no GPs' witness projected detailed annual cash flows for EL, LV, AA, or L&NE. Instead, the GPs relied on the projections made by the Ts themselves. EL relied on DCF for the first four years and after-tax earnings for the fifth year. EL hypothesized that its fifth-year earnings would represent a normalized year and calculated the terminal value of its proffer from that figure. AA and L&NE did not present any annual projections of their earnings and relied solely on an estimate of a single year's normalized earnings. 2. Taxes The second major difference between DCF and EBIT is the treatment of taxes: DCF is an after-tax figure, EBIT is a before-tax figure. The GPs' expert testified that since investors are generally interested in the amount of money left over after every one else, including federal and state governments, has taken his cut, after-tax figures should be used in valuing an enterprise. We do not believe the Ts to dispute him on this point. Rather, AA, the only T other than LV[56] that used before tax figures, argued that the GPs' proposed adjustment to EBIT in their particular cases is inappropriate. AA's expert, James Laurie,[57] argued that *1227 AA should be valued on a pre-tax basis since one of the bidders for its property, the Detroit, Toledo & Ironton (D,T&I), was tax exempt because of its significant loss carry-forwards. But as the GPs point out, even if one accepts AA's contention that D,T&I would have enjoyed a tax-free status forever,[58] AA would not have gotten any benefit from that status. Any investment made by D,T&I would be worth more as a result of D,T&I's tax position. Rather than pay over the value of its tax position to AA, D,T&I would forego the purchase of AA and make other investments where it would be able to keep the entire value of its tax benefit to itself.[59] In order to adjust AA's projection, Smith reduced AA's DCF by 50 percent, representing a 48 percent federal tax rate and an average 4 percent state tax with a credit against federal taxes. The GPs added back 4.3 percent of this after-tax figure to account for the investment tax credit. This figure was derived from the projections for the normalized operation of the IFLs and AA does not challenge its propriety. 3. Capital Expenditures Needed to Maintain Income The final and perhaps most significant difference between DCF and EBIT is that DCF explicitly takes account of the fact that a company may have to make annual capital expenditures in order to maintain earnings. The GPs present a simple example illustrating this point. Suppose there are two companies, A and B, with annual earnings of $100 a year and with no depreciation charges. Capitalizing earnings alone at 10 percent would yield a value of $1000 for both firms. Suppose, however, that company A consumes $30 a year in capital (leaving $70 in DCF) and that company B consumes $80 a year in capital (leaving $20 in DCF) — that is, in order to maintain a $100 a year in income each company has to reinvest $30 and $80 a year respectively. By taking this factor into account, it becomes clear that use of a capitalized earnings approach yields an overly high valuation. Capitalized DCF would be only $700 for company A and $200 for company B. The illustration in the GPs' brief overstates the importance of this factor, however, since it assumes there is no depreciation. In the real world, depreciation charges against income can serve as a proxy for capital expenditures needed to maintain income.[60] EBIT is an after-depreciation figure. In addition, EBIT includes maintenance of way, another item that can properly be viewed as a capital expenditure required to maintain income. Indeed, it is conceivable that use of EBIT could result in the undervaluation of an enterprise if depreciation and maintenance of way charges *1228 exceed capital expenditures needed to maintain income.[61] The advantage of using DCF over EBIT then is not that EBIT ignores the fact that capital expenditures may be required to maintain income but that DCF explicitly includes the factor while EBIT at best includes it indirectly.[62] While we thus agree with Smith that it is desirable to adjust EBIT to take account of capital expenditures needed to maintain income, we do not completely accept the validity of the adjustments Smith proposes. Specifically, we think that for AA and L&NE, Smith's reduction of EBIT by one-third was inappropriate. Smith derived this figure from the projected normalized earnings statements for the IFLs. Smith found that DCF equalled two-thirds of EBIT for the IFLs and assumed that a similar relationship would hold for AA and L&NE. This assumption was unjustified. Smith testified that the bulk of his one-third adjustment was accounted for by capital expenditures, the largest part of which is made up of rolling stock. Smith (D) (Aug. 25, 1980) at 2022-34, 19 GP Supp. App. at 374-86; 2 GP at F 19 n.9. AA based its earnings study, however, on the assumption that, save for locomotives, its acquirer would have leased all the equipment it would have needed. And as Smith himself admitted at his deposition, if the "income statement correctly reflected the total amount of per diems payable, then there would be double counting" if the one-third deduction were also made. Smith (D) (Aug. 25, 1980) at 2028, 19 GP Supp.App. at 380. This is because the shortfall in allowable depreciation "will be passed on" by the owner of the car in the rental charges. Smith (D) (Aug. 26, 1980) at 2120, 19 GP Supp.App. at 399. It was improper, therefore, for Smith to reduce AA's EBIT to account for capital expenditures needed to maintain income.[63] Smith's one-third adjustment is similarly overstated for L&NE. L&NE's witness, Jack Storm,[64] assumed that all new equipment needed by L&NE would be rented. Storm (D) (Mar. 28, 1979) at 46, L&NE App. Accordingly Smith should not have deducted one-third from L&NE's EBIT to arrive at DCF. EL stands on a different footing. It posited that the acquirer of its lines would have purchased new equipment and would have financed those purchases by issuing equipment trust certificates.[65] EL deducted interest and principal payments on these certificates to derive what it calls levered cash flow. The GPs' expert testified that the acquisition cost of new equipment should be charged against cash inflow in the year that *1229 it was acquired. In order to convert EL's figures, Smith performed a two-step adjustment. First, he added interest and principal payments on new equipment back into the EL's figures. Smith next subtracted the acquisition cost of new equipment. Smith (S) (Jan. 30, 1980) at 249: id. at Ex. GPS-Smith (II)-31. This yields what the GPs call an unlevered cash flow. At times it has seemed to the court that the parties have argued that one or the other approach was the "correct" one. See, e.g., 2 ELF at 441-42; Smith (S) (Jan. 30, 1980) at 249. The GPs' reply brief and oral argument make it clear, however, that both sides agree that it is proper to use either a levered or unlevered cash flow in valuing an enterprise as long as a debt-to-equity ratio and cost of debt consistent with that assumption is used and that either approach, properly applied, will yield the same result. 1 GPRB at F 56-58; Tr. at 70-73 (EL), 282-86 (GPs). There will be further discussion of the debt-to-equity ratio and cost of debt at pp. 1231-1232, 1235, infra. Suffice it to say here that the parties agree that if the effect of equipment financing is accounted for in the cash flow, then the debt-to-equity ratio and cost of debt should exclude equipment debt. Similarly, if an unlevered cash flow is used, equipment debt should be included in the ratio and cost of debt. The choice between levered and unlevered cash flows then boils down to one over the proper debt-to-equity ratio and cost of debt. We opt for the use of unlevered cash flows. This is because, as our full discussion of debt-to-equity ratios will reveal, we accept the GPs' approach to choosing the proper ratio, and the ratios the GPs ask us to consider include equipment obligations. Rather than adjust those ratios, we think it is easier to adopt the GPs' methodology. The other difference between the approaches of the GPs and EL is that EL capitalized fifth-year earnings while the GPs capitalized fifth-year unlevered cash flow. As already discussed, we think it is preferable to use DCF. EL argues, however, that Smith's use of EL's own projected fifth-year cash flow was improper. Smith calculated the PDV of the first four years' projected unlevered cash flows and added to that the PDV of the capitalized value of the fifth-year's unlevered cash flow. Implicit in this approach is the assumption that fifth-year cash flows represented cash flows for a normalized year. EL argues that this assumption was unfounded. Tr. at 77-78. Shannon testified that fifth-year earnings for EL were normal because by that time the acquirer of EL's lines would have completed its rehabilitation program and would have fully realized the projected savings from merged operations. Shannon made no judgment as to whether fifth-year cash flow, as opposed to fifth-year earnings, represented a normal year's experience. EL claimed that projected fifth-year capital expenditures did not represent the normalized experience because Shannon's replacement schedule for EL's car fleet was based on a 23-year replacement cycle whereas the useful life of this equipment is 28 to 30 years. EL concluded that fifth-year equipment acquisition levels are higher than normal and that, therefore, the fifth-year cash flow understates the "normalized" cash flow. Id. Since this argument had not been fully addressed in the briefs, the court in a letter dated June 25, 1981, asked counsel for EL to submit a statement giving more details of and record support for their contentions. EL submitted a statement to the court on July 10, 1981, Sp. Ct. Rep. N 36191, and the GPs responded on July 17, 1981, Sp. Ct. Rep. N 36300. An examination of these statements and the accompanying workpapers reveals that Shannon did not assume an accelerated replacement program. Instead, he explicitly assumed that a normal schedule would require the annual replacement of 4 percent of the EL's total car fleet. This translates into a 25-year replacement cycle, a schedule comparable to the projections made by other witnesses. Smith was perfectly justified, therefore, in treating the fifth-year cash flow as representing a normal year's experience. *1230 B. Discount Rate In the previous discussion, we have frequently alluded to an "appropriate interest rate". Not surprisingly, the parties differ as to what that rate should be. The GPs' lead witness on the subject, Smith, opined that the proper rate would be 10.66 percent assuming a July 1, 1974, disposition and 10.45 percent assuming an April 1, 1976, disposition. EL's witness, Percy Young, Jr.,[66] applied a 6 percent discount rate to the first four years' DCF and capitalized fifth-year earnings at 10 percent.[67] AA's witness, Laurie, used a 7.5 percent capitalization rate[68] and Storm, L&NE's witness, used a 10 percent rate. Neither Laurie nor Storm, however, gives any justification for his choice of interest rate.[69] The witnesses who did give detailed consideration to this question, Young for EL, Smith, Dr. Richard Roll,[70] and Dr. Richard Meyer[71] for the GPs, and Paul Hallingby, Jr.[72] and Dr. Stewart Myers[73] for PC agreed that the appropriate interest rate for discounting DCF could be derived by calculating the average weighted cost of capital (AWCC). AWCC is made up of three components: cost of equity, cost of debt, and debt-to-equity ratio. Calculation of AWCC requires an average of the cost of debt and cost of equity, weighted by the percentage of debt and equity in the capital structure. So, for example, if the cost of debt is 5 percent and cost of equity is 10 percent and debt comprises 50 percent of the capital structure, AWCC is 7.5% - (.5 × .05) + (.5 × .10).[74] Cost of debt and cost of equity are both made up of three factors — the risk-free rate of interest, an inflation component, and a premium compensating the investor for risk. The risk component is the only one unique to each enterprise, however. Risk is essentially an assessment of the likelihood that the true outcome will differ from the expected outcome. So, for example, if there are two investments that are both expected to generate a DCF of $10 a year on average, but one is expected to have a constant return and the other a variable return, the second investment would be considered riskier. Given a choice between the two investments with the same price, a rational investor would chose the one with the more predictable yield. In order to induce an investor to purchase the riskier asset, its purchase price must be lower than that of less risky assets with the same DCF. A lower purchase price translates into a higher rate of return on the riskier asset. The degree of risk also affects the debt-to-equity ratio — the greater the risk, the less debt the company can carry and still maintain the same cost of debt. *1231 The Ts' witnesses testified that a purchaser of the proffered lines would view the investment as having the same level of risk as an investment in an average risk security. The GPs' witnesses testified that an investment in the Ts' properties would involve a greater than average amount of risk. On balance, we consider that the conclusion is almost inescapable that a purchaser of the Ts' lines would have viewed the investment as having a greater than average amount of risk. We believe, however, that the GPs' witnesses overstate the risk premium associated with an investment in the Ts' lines. Accordingly, we have no choice but to discuss each of the three components that make up AWCC and derive our own discount rate. 1. Debt-to-Equity Ratio The first component we consider is the debt-to-equity ratio because the cost of debt and the cost of equity both depend on that number. The relationship between the debt-to-equity ratio and the cost of debt and equity is a simple one. Debt is generally cheaper than equity for two reasons. First, because bondholders have first claim on the assets of the company, there is less risk that, in liquidation or reorganization, bondholders will not recoup their investments with interest. Second, because interest payments are tax deductible, the effective cost of debt is reduced to the company. One might think, then, that the optimal strategy would be for a company to reduce its total cost of capital by weighting its debt-to-equity ratio heavily in favor of debt. There are two limits to this however. First, as the percentage of debt in the capital structure increases, so does the probability that the company will not be able to meet its interest and principal payments. The probability that bondholders will not recoup their investment in the event of liquidation or reorganization also increases. The increased riskiness of the debt translates in the marketplace to a higher interest rate so the cost advantages of debt tend to disappear as the proportion of debt in the capital structure increases. The riskiness of equity investment also increases as the debt-to-equity ratio goes up. Since a stockholder's interest is in the cash generated by the enterprise after bondholders have been paid, the greater the percentage of debt the greater the volatility in the stockholder's interest. To illustrate, assume two companies are expected to generate DCF of $100 a year. Company A has no debt, Company B is 80 percent debt financed, and pays $80 a year to bondholders. If actual earnings are 10 percent less than expected, the shortfall to an equity investor in Company A from his expected return is also 10 percent. But for an equity investor in Company B, the shortfall is 50 percent — he expected $20 and received only $10. The same phenomenon occurs on the upside as well — an unexpected 10 percent increase in earnings results in an unexpected 50 percent gain to equity investors in Company B. In selecting a debt-to-equity ratio, the witnesses uniformly start with the average ratio for railroads as a group. Roll, the GPs' witness, testified that based on book value for all five samples of railroads he examined, railroads in the mid-1970's were financed with roughly 40 percent debt and 60 percent equity. Roll (S) (Jan. 30, 1980) at 112. Smith determined that railroads that had A-rated senior mortgage debt from 1966-75 generally maintained ratios of debt to total capitalization of 30-35 percent. Smith (S) (Jan. 30, 1980) at 80-81. Stewart Myers, a witness for PC, estimated that the average debt ratio for ten large railroads making up the S&P 10-railroad index from 1969-74 was about 45 percent. Myers, however, calculated this debt ratio using estimated market values of equity and debt rather than book values. All the witnesses included equipment debt in their ratios. Our first step, then, is to choose between book-and market-valued ratios. Stewart Myers states flatly that "the relevant measures of the debt and equity proportions are the market values of debt and equity." Myers (S) (Dec. 1, 1978) at App. C-1. No further justification is given. However, Smith tells us that bond rating agencies primarily rely on book values. It seems likely, therefore, that an acquirer of the Ts' lines would have made its capitalization *1232 decision on the basis of book valued debt ratios of comparable lines. Smith (S) (Jan. 31, 1979) at 36.[75] Assuming that book-valued ratios are the relevant figures, we must still choose a number. We believe that Smith made a convincing case for the use of a 30 percent debt ratio. Smith based his argument on the assumption that an acquirer would have sought an A rating on its senior mortgage debt. Smith thought that there would be two related advantages to an acquirer in seeking an A rating. First, he noted that debt rated at less than an A rating was far more difficult to market than higher rated bonds at the relevant times. In turn, this difficulty became reflected in the interest rates investors demanded. See Smith (S) (Jan. 30, 1980) at 143-44. Since the A-rated railroads that Smith examined had a book valued debt ratio of 30-35 percent including equipment obligations, and since the Ts' lines would have been viewed as riskier than a typical A-rated road, an assumption with which we agree, see pp. 1233-1235, infra, Smith concluded that an acquirer would have used a debt ratio at the low end of this range. See Smith (S) (Jan. 30, 1980) at 80-83. For the purposes of our analysis then, we will assume that acquirers of the Ts' lines would have ultimately financed their acquisitions using 30 percent debt and 70 percent equity.[76] 2. Cost of Equity In estimating the cost of equity three factors must be considered: the rate of return demanded by investors for average-risk stocks, the riskiness of railroad stocks relative to average-risk stocks, and the riskiness of an equity investment in the Ts' lines relative to railroad stocks. We will refer to the two additional risk factors as railroad risk and special risk. There is rough agreement among the experts on the rate of return on average-risk investments. The GPs' witness, Smith (S) (Jan. 30, 1980) at 61, and PC's witness, Myers (S) (Dec. 1, 1978) at 29-34, both relied on a study performed by Ibbotson and Sinquefield[77] which found that, for the period from 1926 to 1976, the real rate of return for the Standard & Poor's index of 500 stocks was 9.2 percent. The experts disagree over whether an investment in the Ts' lines would be viewed by a purchaser as entailing a greater than average risk. Smith concluded that railroad stocks as a group involve greater than average risk and that an investment in the Ts' lines in particular would involve still greater risk. He stated that the Ts' lines should be assigned nominal costs of equity ranging from 20 to 25 percent.[78] For valuation purposes, he settled on the average, 22.5 percent, as the nominal cost of equity as of April 1, 1976, which translated into a nominal cost of equity of 21.98 percent as of July 1, 1974.[79] The Ts' witnesses testified *1233 that investment in the Ts' properties would not be viewed as entailing any additional risk. Smith did not quantify how much of his additional risk premium consisted of railroad and how much of special risk. Nonetheless, by examining the arguments and statistical evidence offered in support of Smith's conclusion that an investment in the Ts' properties would involve both railroad and special risk we can come to a judgment as to the reasonableness of Smith's conclusion and as to what the proper discount rate should be. (a) Railroad Risk Smith enumerated a long list of factors supporting his conclusion that investment in the railroad industry is riskier than investment in an average-risk industry. He stated that: the returns earned by railroads are impacted to a greater extent than most businesses by a combination of factors beyond their control, including economic cycles, inflation, government transportation policy and regulation ... competitive considerations, capital intensiveness, labor intensiveness, and adverse weather. Smith (S) (Jan. 30, 1980) at 45-46. Whatever the merit of Smith's judgments as to the respects in which railroads are riskier than average,[80] Smith also admitted that in numerous respects railroads were less risky than average. For example, Smith admitted that railroads are essentially immune from the risks associated with dealing directly with consumers or with being in a field with rapid technological change. See 2 ELF at 420-21. Simply as a matter of judgment, then, it would seem to be a very difficult task to weigh all the relevant factors and state confidently that an industry is more or less risky than average. This is one area, however, where the court need not rely solely on an expert's personal judgment. Statistical analysis is available and was offered by the parties to determine whether railroad stocks as a group have historically yielded higher rates of return to compensate investors for higher than average-risk. The parties have generally avoided heavy reliance on this material in their briefs for fear of burdening the court with the detailed and admittedly technical debate among three experts which spans four volumes of testimony. Likewise, we will not encumber this opinion with a lengthy discussion of this evidence. Briefly, PC's witness, Stewart Myers, estimated two commonly used measures of risk, standard deviation of the annual rate of return and beta, for a representative sample of large railroads. Beta measures the amount of risk which a stock contributes to a portfolio made up of a large number of stocks. Both measures showed that railroad stocks were average-risk investments. The GPs' witnesses offered several criticisms and refinements of Myers' study. None of this evidence gives us any reason to question Myers' conclusions. The GPs' witness, Roll, proposed another measure of riskiness which he called "excess return", and estimated that railroad stocks commanded a 6.14 percent premium over average risk stocks for the 1969-74 period. In essence, excess return is the difference in risk premium between an average-risk stock and the stock or stocks being studied. But Roll's estimates of excess returns were not statistically significant — that is, one could not say to an acceptable degree of certainty that Roll's findings were not attributable to random fluctuations. And their values varied widely depending on the period studied, so much so that for the years 1964-69 excess returns were negative. In sum, despite their impressive efforts, Roll and Richard Meyer have failed to develop evidence which would lead us to reject Myers' initial conclusions. (b) Special Risk While we reject the GPs' contention that railroad stocks as a group are relatively *1234 risky investments, there can be no doubt that, as the GPs' brief puts it, "[b]oth the record and simple common sense", 1 GPRB at E 79, indicate that potential acquirers of the Ts' lines would have viewed such investments as being of above average risk. However, we believe that Smith overstated the special risks attending the Ts' properties. Smith initially made his risk assessment for the IFLs and went on to assume that the Ts' properties were equally risky, see note 78, supra. In reviewing the three major reasons he gave for the IFLs' special risk the court is convinced that, although the Ts' properties share the same characteristics that contributed to the IFLs' peculiar riskiness, Smith's blanket conclusion that the Ts' lines were as risky as the IFLs was unfounded. The first factor Smith identified is that the Ts' properties were all located in the Northeast. Smith argued with considerable force that: ... Class I railroads in the Northeast on average were characterized by less favorable operating margins, lower returns, less growth and less financial resiliency than observed elsewhere in the industry.... Rail operations in the Eastern District (roughly, the Northeast region) were particularly vulnerable to competition from the trucking industry, and were characterized by relatively higher terminal costs, shorter average lengths of haul, and higher labor costs than rail operations elsewhere in the country. These factors contributed to an increased volatility of railroad financial performance.... Smith (S) (Jan. 30, 1980) at 46-47. But, as the transcript of his deposition reveals, Smith (D) (Aug. 21, 1980) at 1624-27, EL App. J at 337-40, when Smith said "railroads in the Northeast", he meant bankrupt railroads in the Northeast. The financial performance of Norfolk & Western, C&O, and B&O all compare favorably with Western and Southern railroads. Smith argued, however, that the profitable Eastern carriers as a group were not comparable to the bankrupts. For example, he stated that the Norfolk & Western and C&O were heavily into transporting coal, a service which is virtually immune from trucking competition. Smith (D) (Aug. 21, 1980) at 1625, EL App. J at 335. And as Smith points out, there is no getting around the fact that at that time the major railroad bankruptcies had occurred in the same region and that this supports an inference that railroads operating in the Northeast shared common problems.[81] Saying this, however, does not mean that all the Ts would be viewed by potential acquirers as suffering from these problems to the same extent as PC. This difference is highlighted by the next factor Smith cites. He claims that potential acquirers would view investment in the Ts' properties as particularly risky because the railroad industry in the Northeast would be radically reconfigured in an attempt to overcome all of the problems previously discussed. Again, to a certain extent, Smith is correct. Even without the uncertainties introduced by the creation of the JTs, the transformation would have been striking. Approximately one-third of the combined trackage of EL and LV would have been abandoned. New owners, new routings and new competitive relationships would have all contributed to the ensuing uncertainty. But while the restructuring of EL's and LV's lines would have been considerable, it pales in comparison to that envisioned for PC. In the alternative scenario, PC posited that 16,000 miles out of its 26,000-mile system would have been abandoned. The scope of the adjustments necessitated by PC abandonments would have been far greater than those for the other Ts. Smith admitted at his deposition that "Erie was just, on a raw basis, in terms of mileage, not to be reconfigured to the same extent, as certainly the IFL." Smith (D) *1235 (Aug. 8, 1980) at 757-58, EL App. J at 165-66. And the complexities and uncertainties attendant on the restructuring of a 26,000-mile system cannot be compared to the simple single-line proffers made by AA and L&NE. To be sure, Smith correctly points out that wholly apart from uncertainties created by the restructuring of their own lines, the extensive abandonments and restructuring of PC's lines would have created problems for the purchasers of the other Ts. But this factor cuts both ways. Laurie, EL's witness on traffic, projected that while EL would lose 17,000 cars annually due to PC's abandonments, EL would gain over 50,000 new cars. Laurie (S) (Dec. 30, 1978) App. 1 at 4. Finally, Smith reasoned that potential acquirers would assign a higher discount rate to the Ts' lines because "property with which a potential purchaser railroad does not have significant experience or any historical reference point — will be treated with less confidence." 2 GP at F 54. Again, we accept Smith's general proposition but do not believe that the Ts' lines stand on the same footing as the IFLs. As EL and AA point out, in contrast to the IFL case their projections are based on verifiable historic traffic and cost data. While potential acquirers of the Ts may not have had significant first hand operating experience with those lines, this does not mean that all projections would have been viewed as having the same inherent unreliability.[82] One benchmark we have not yet considered is the costs of equity that Smith actually employed in his valuation of five other rail acquisitions conducted at various times from 1976 to early 1980. The average rates he has used ranged from 16.5 percent to 19 percent. Although EL argues that we should reject this testimony because of Smith's failure to give significant details about how those numbers were derived, see ELRB 159-62, keeping these numbers in mind gives a feel for what real world numbers are like. The court is left then with the difficult task of choosing a discount rate based on an informed judgment as to what factors to consider and as to the riskiness of the Ts' lines relative to the IFLs. The range we consider runs from 16 percent, the rate of return on average-risk equity investments, to 25 percent, the highest rate considered by Smith. We do not think it worthwhile to attempt to derive different rates for different Ts since, all things considered, we do not think we will find much to distinguish them. While AA and L&NE may have the simplest proffers, Smith points out, Smith (S) (Jan. 30, 1980) at 293, that their small size creates an entirely different set of risks. And relative to AA and L&NE, the EL-LV package has neither any particularly positive nor negative factors. To be sure, no number which we choose will be "correct" since we are making a subjective judgment as to what the subjective judgment of a potential acquirer would have been. On balance, however, we think that 20 percent is a reasonable estimate of the cost of equity for the Ts' lines as of January 1, 1974, assuming 30 percent debt financing.[83] 3. Cost of Debt The final component in determining AWCC is the cost of debt. As previously discussed, the 30 percent debt ratio that we have used assumes that senior mortgage debt would be A-rated. According to Smith, senior mortgage debt would make up one-third of the total debt. The other two-thirds would be made up of equipment *1236 trust certificates. The certificates would be Aa-rated. Equipment debt is generally rated one grade higher than senior mortgage debt because it is better secured and because it receives favorable treatment in bankruptcy. No witness provided us with estimates of what the yield on railroad bonds would have been.[84] However, we do have two tables prepared by the GPs' witness Roll, Roll (S) (Jan. 30, 1980) at 87, 97, which show the yields of various railroad bonds. The tables show that on January 15, 1974, Burlington Northern issued A-rated, non-convertible senior mortgage debt with a yield of 8.6 percent and that on the same date, Norfolk & Western issued Aa-rated equipment trust certificates yielding 7.9 percent. We believe that a prospective acquirer would have viewed these as reasonable estimates of their cost of debt capital. Taking a weighted average of these rates gives us an 8.13 percent pre-tax cost of debt as of January 1, 1974. Only one further adjustment needs to be made. The cost of debt which we have so far discussed is pre-tax. However, since interest payments are tax deductible, the relevant measure is the after-tax cost of debt. All of the witnesses who have considered this have assumed a 50 percent marginal tax rate and have accordingly reduced the cost of debt by one-half. 4. Calculating AWCC We now have all the necessary ingredients. The table below shows how AWCC for January 1, 1974 is computed. AWCC as of January 1, 1974 Weighted cost Weighted after-tax AWCC of equity cost of debt ________________________________________________________ (20 × .70) (4.067 × .30) + = 15.22 14 1.22 The AWCC which we have so far calculated is stated in nominal terms. The cash flow projections are stated in constant dollars. We therefore need to factor out the inflation component from the rates. Using Smith's assumption of a 5.58 percent long term inflation rate for 1974, Smith (S) (Jan. 30, 1980) at 54, we derive a real AWCC of 9.13 percent as of January 1, 1974. VIII. Rehabilitation Costs and Maintenance of Way Expenses The GPs and each T provided estimates for the rehabilitation costs (RC) and annual maintenance of way expenses (MOWE) necessary to restore and maintain the properties they proffered for continued rail use. The GPs obtained their estimates from a study conducted expressly for this litigation by Sverdrup & Parcel and Associates, Inc. (S&P), consulting engineers.[85] The Ts obtained their estimates from studies conducted in-house, studies by outside consultants, and studies by potential acquirers. That the methodologies of the various studies differ markedly is not surprising. What is surprising is the extent to which the results differ.[86] *1237 The differences in method between S&P's study and the various studies of the Ts are too substantial to allow us to compare the studies item by item and to extract a composite figure for RC and MOWE that melds the results of each. Rather, we have examined each study individually to determine its reliability in the light of the shortcomings pointed out by its opponents, with a view toward selecting the more reliable study and adjusting it for whatever errors are uncovered. For the reasons set out below, we conclude that the GPs' study does not withstand scrutiny. We therefore accept the studies submitted by EL, LV, and AA as the starting point for deciding the RC and MOWE expenses facing the potential acquirers of those Ts' lines. A. The Government Parties' Study The methodology employed by S&P in conducting its study is detailed extensively in two multivolume reports prepared by S&P and introduced in evidence by the GPs. See Rehabilitation Study for United States Railway Association, January 31, 1979 (1979 Study), and Rehabilitation and Maintenance Study for United States Railway Association, January 30, 1980 (1980 Study).[87] The two principal architects of the S&P Studies were H. Porter Morgan, who was in charge of all track related items, and Roy Gunderson, who was in charge of bridges and tunnels. Both Morgan and Gunderson had extensive experience in railroad maintenance.[88] Morgan and Gunderson began by dividing the Ts' rail facilities into four major categories — track, bridges, yards, and signals. Drawing upon a variety of sources, see 1979 Study at III, they developed standards and criteria to be used in defining the level to which these rail facilities should be restored and maintained. Morgan and Gunderson then set out to determine the condition of the Ts' lines as they existed in 1974. Because they did not begin their study until four years later, any new data they could have collected concerning the condition of the railroads would have had to be adjusted to account for the deterioration that occurred and the maintenance and rehabilitation that was undertaken in the intervening years. Rather than undertake this difficult and uncertain process, they obtained their raw data from a study, commonly known as the Bechtel Study, which was commissioned by the USRA during its preparation of the FSP and which was conducted in 1974 and 1975. The data generated for the Bechtel Study were obtained from inspections, interviews with key railroad personnel, maintenance charts, and other company records. Extrapolating from this Bechtel Study data, Morgan and Gunderson attempted to assess the condition of each T's lines. They determined how much work was necessary to restore the lines to the levels dictated by the standards they had developed, and defined these rehabilitation requirements in terms of units called "work functions".[89]*1238 Morgan and Gunderson also developed unit cost figures for each work function. Straightforward calculations based on their projections for the total number of units of work needed and the costs attributable to those units of work yielded RC figures for each proposed acquisition hypothesized by the Ts in the alternative scenario. A similar set of calculations based on projected annual maintenance needs yielded a corresponding set of annual MOWE figures. See H. Morgan Corrections (S) (June 6, 1980) at Tables VIII-2aA, VIII-2bA. The S&P methodology is not inherently weak. As applied, however, it yielded questionable results. In three of the four major areas of rehabilitation studied by the S&P team — track, bridges, and yards — sampling errors, incorrect use of standards, and inaccurate data undermined the reliability of S&P's results. 1. Track With one exception, S&P's estimates for track rehabilitation constituted the largest portion, in some cases over 50 percent, of every T's total RC. See 2 GP at E 10-11. The Ts catalogue numerous errors in the methods S&P used in making its estimates of track rehabilitation. The asserted errors relating to three portions of the track study — rail replacement, turnouts, and crossings — bear detailed discussion. (a) Rail Replacement Just as track rehabilitation generally constituted the largest portion of each T's total RC, rail replacement constituted the largest single item of track rehabilitation.[90] For the reasons explained below, Morgan overstated the rail replacement needs of the Ts. i. Vertical Headwear. Headwear refers to the wear to the surface of the rail that comes in contact with a train's wheels. Although this surface of the rail wears both horizontally, affecting the width — or gauge — of the rail, and vertically, the controversy over headwear in this case is confined to vertical headwear. The rate of headwear depends primarily on the speed and weight of the traffic that runs over it. When the headwear on rail used in heavy, high-speed traffic reaches certain levels, considerations of safety and economics demand that it be removed and relaid in areas of lighter use, such as secondary lines, branch lines, and yards.[91] In making his rail replacement decisions, Morgan relied on a table of headwear standards developed by the American Railway Engineering Association (AREA).[92] Morgan misapplied these AREA standards, however, in three important respects. *1239 Morgan's first error concerns his interpretation of the AREA figures. The AREA table sets out four classes of use for rail: main line (Class I), branch line (Class II), light branch line (Class III), and yards (Class IV). For each class, the table assigns figures for maximum allowable headwear, both vertically and horizontally, for various weights of rail.[93] For example, the table sets out the following figures for maximum vertical headwear on 140-pound rail: Class I ¼ inch Class II 3/8 inch Class III 5/8 inch Class IV ¾ inch 1979 Study at III 30. Morgan interpreted the AREA table to mean that as headwear on the Ts' rails in each class approached the allowable figures, it should be removed and relaid in the next lower class.[94] Thus, as Morgan interpreted the table, 140-pound rail in Class I use should be removed and relaid in Class II use when the headwear approaches ¼ inch; rail in Class II service should move to Class III service when headwear approaches 3/8 inch. This interpretation and application of the AREA standards is directly opposed to the views of other railroad engineers. When the Bechtel Study was completed, Bechtel had a group of six railroad engineers review and assess the work of the various engineering firms who had done the leg work for the study. Those engineers had this to say about the AREA table: [T]he table actually recommends the limits of rail wear for various classes of service at which it is economically feasible to remove bolted rail from track, weld it and replace it in track of the same class service rather than requiring removal from track to a lesser class of service. H. Morgan (D) (May 7, 1980) Ex. 109, EL App. E at 198. Under this interpretation of the table, even if the table is used as a guide for determining when to remove rail, a purpose for which it was not intended,[95] Class I rail can still be economically relaid in Class I service as wear approaches ¼ inch; it is not until wear exceeds ¼ inch that relaying in Class I service becomes uneconomical because its useful life would be too short. Indeed, if Class I rail is destined for Class II service, the table indicates only that for economic purposes the rail need not be removed from Class I service until wear approaches 3/8 inch. The GPs make no attempt to defend Morgan's interpretation of the AREA table. Nor do they quarrel with the Bechtel consultants' explanation of the proper use of the table. See 3 GPRB at App. A 41-42. Instead, the GPs' line of defense is that Morgan did not make rail replacement decisions solely on the basis of headwear; headwear was but one of a number of factors considered. Nevertheless, Morgan and one of the two assistants assigned to making rail replacement decisions admitted that headwear was a major factor in making those decisions.[96] See H. Morgan (D) (Apr. 14, *1240 1980) at 201-02, 216, 28 GP Supp.App. at 6-7, 14. By analyzing a major factor like headwear under standards that were more strict than those used in the industry, Morgan and his assistants would have replaced rail prematurely, thereby inflating the cost of rehabilitation. The second way in which Morgan misapplied the AREA table was by misclassifying the Ts' lines. The AREA table sets out four classes of track use. The table gives no guidance, however, in determining the appropriate classification for a particular rail line. Morgan used traffic density, measured in millions of gross tons per year (MGT), as the determinant. Using traffic density to classify railroad lines is an acceptable practice. The problem arises from the density figures Morgan chose. Morgan classified the Ts' lines as follows:[97] Class I 10 or MGT Class II 1 to 9.9 MGT Class III 0 to 0.9 MGT 1979 Study at III-32. In contrast, given the speed limits on the Ts' lines, the Bechtel Study used the following classification: Class I 20 or more MGT Class II 5 to 19.9 MGT Class III 0 to 4.9 MGT ELF at 163; H. Morgan (D) (May 7, 1980) Ex. 108, EL App. E at 187. As is obvious from a comparison of the two tables, Morgan's scheme places track between 10 and 19.9 MGT and track between 1 and 4.9 MGT in higher classes than they were placed in Bechtel's scheme. Since headwear standards become more strict as the class of service increases, Morgan's classification subjects all the Ts' lines within those density limits to more restrictive headwear standards than would be the case under Bechtel's classification. The inevitable result is that more rail replacement is called for under Morgan's approach than that of Bechtel. On the record before us, we cannot definitively say which is the correct classification scheme since the AREA does not set out the densities for its classifications. For several reasons, however, we conclude that Morgan's figures are too conservative. First, the Bechtel Study was performed by disinterested parties for the FSP. Second, the Bechtel classification scheme was specifically approved by Chessie engineers who made their own independent study of the rehabilitation needs of some of the Ts' lines. Chessie System Preliminary Engineering Plan for Selected Property Acquisitions of Erie Lackawanna, Penn Central, Reading (Chessie Study), EL App. B at 211. Morgan in contrast offered no support, other than his own expert judgment, for the reasonableness of his classifications. The third problem with Morgan's application of the AREA tables is his use of inaccurate data concerning the actual density of traffic on a number of the Ts' lines. For example, Morgan assigned excessive densities to 13 of the 27 EL lines in the Western acquisition for which he had inspection data.[98] See 1 ELF at 169 & n.132. According to the GPs' own estimates, more than 27 percent of the rail replacement in the EL Western acquisition in Morgan's study occurred in lines where densities were overstated. 3 GPRB at App. A 49. Morgan admitted that 280 of the 390 miles of LV track he would replace were on lines where correction of his density errors would have reduced the applicable AREA classification. H. Morgan (D) (Aug. 13, 1980) at 4026, EL App. D at 469. Although Morgan contended that correcting the densities would not have had a significant impact on his rail replacement decisions because headwear *1241 was but one factor, H. Morgan (D) (Aug. 13, 1980) at 4026-27, EL App. D at 469-70, he had no basis for that opinion because he did not know how many rail replacement decisions he would have had to change. Indeed, he could not know without reanalyzing those miles of track, H. Morgan (D) (May 5, 1980, Aug. 13, 1980) at 2571-72, 4027-28, EL App. D at 56-57, 470-71, something he did not do. By classifying substantial portions of the Ts' lines in higher classifications than they belonged, Morgan derived rail replacement figures that were too high. ii. End Batter and Mismatch. "End batter" and "mismatch" were two of the other factors Morgan and his assistants relied on in making rail replacement decisions. End batter is the deterioration of the ends of bolted rail, and mismatch is a difference in height between the ends of two abutting lengths of rail. Morgan adopted generally accepted standards published by the Federal Railroad Administration (FRA), which assign the maximum allowable limits for these two types of wear depending upon operating speed. Morgan's team would have replaced or repaired rail that exceeded one half these FRA limits. 1979 Study at III 40.[99] As with his density classifications, Morgan's FRA speed classifications for a number of the Ts' lines were incorrect. For example, over 900 miles of EL track in the Western acquisition were misclassified. See H. Morgan (D) (Aug. 13, 1980) at 3988-4008, 4014-15, EL App. D at 431-51, 457-58; 1 ELF at 196 & n.157. By the GPs' own calculations, almost 38 percent of the rail to be replaced in the EL Western acquisition was located in lines where the speed class was too high. Again, Morgan was unable to state how much his rail replacement decisions would have been affected had he classified the lines properly. Nor did he attempt to reanalyze those decisions after these errors were pointed out, H. Morgan (D) (Aug. 13, 1980) at 4016, EL App. D at 459, even though he made adjustments in some of his other calculations to take account of these errors. H. Morgan Corrections (S) (June 6, 1980) at 7. As in the case of the density errors, misclassifying the rail as to operating speed subjected the rail to higher standards than necessary and therefore inflated rail replacement needs. There were other errors in Morgan's study that increased rail replacement estimates. For example, his study called for the replacement of rail on nonexistent lines and on lines where maintenance was conducted by railroads other than the Ts. 1 ELF at 183-84. We could perhaps adjust those errors. The four major errors detailed above, however, cannot be corrected without redoing the study line by line. Those errors substantially overstated the Ts' rail replacement needs.[100] A comparison of Morgan's estimates for rail replacement with the figures proposed by the Bechtel Study reinforces that conclusion. For EL and LV, Morgan called for the replacement of between 40 and 60 percent of the rail on the proffered lines; the Bechtel Study, on the other hand, called for replacement of only 7 percent. See 1 ELF at 157.[101] *1242 (b) Turnouts and Crossings Turnouts are switches that divert trains from one line of track to an adjoining line. Crossings are the points where a rail line intersects with either another rail line or a highway. S&P's estimates for the cost of rehabilitating the Ts' turnouts and crossings are not as large as its estimates of rail replacement costs. The figures are nevertheless substantial.[102] The methodological shortcoming that undermines these figures is the use of questionable sampling techniques for obtaining data on the condition of these items. The sampling technique recommended by the Bechtel Study called for obtaining data from 10 percent of all the Ts' turnouts except yard turnouts. The sample turnouts were to be randomly selected and evenly distributed over the Ts' lines. See 1 1979 Study App. at III-1-15. Morgan admitted, however, that the Bechtel data he relied on did not meet the recommended 10 percent sampling rate. On the EL and LV acquisitions, his samples ranged from only 3.3 to 5.9 percent. H. Morgan (D) (Apr. 30, 1980) at 2405-09, EL App. D at 15-19. Moreover, the sampled turnouts were not evenly distributed over all EL and LV lines; some lines were entirely unrepresented, while as many as 33 percent of the turnouts on other lines were used in the sample. H. Morgan (D) (Apr. 30, 1980) at 2409-17, EL App. D at 19-27. The S&P RC estimate for the highway crossings of at least one T, LV, is based on an even more questionable sampling technique. Data were available on only three *1243 of an estimated 329 highway crossings on LV lines in one of the EL-LV acquisitions. All three were located on a branch line. The Bechtel Study had determined that they should be replaced. On the basis of that sample, and the Bechtel Study's recommendation, Morgan decided to replace all 329 crossings. See H. Morgan (D) (Apr. 30, 1980) Ex. 67, EL App. E at 90-91. Morgan later conceded that this was not a reliable sampling method. H. Morgan (D) (Apr. 30, 1980) at 2472-73, 28 GP Supp.App. at 228-29. The meager nature of these samples, both for turnouts and for highway crossings, as well as the willingness to rely on such limited data in assessing substantial rehabilitation sums to the Ts, does not inspire confidence in the accuracy of the S&P estimates, and indeed indicts the S&P Study in general. That these sampling problems introduced errors in favor of the GPs is confirmed by a comparison of S&P's estimate for these items with the numbers generated by a potential acquirer that had no motive to understate costs. Chessie estimated that turnout replacement on the EL lines in the Chessie acquisition would cost $326,100. Chessie Study, EL App. B at 221. In contrast, the S&P estimate was almost $6 million before salvage credit. H. Morgan (D) (Aug. 22, 1980) Ex. 280, EL App. E at 278-80 (Proffer Code 83). Chessie's estimate for highway crossings was $400,000, EL App. B at 221, while S&P's was over $3 million. H. Morgan (D) (Aug. 22, 1980) Ex. 280, EL App. E at 280. 2. Bridges Gunderson headed the S&P team in charge of producing bridge RC estimates on the Ts' lines. The Ts launch a broad attack on the methods and results of these estimates, illuminating errors in all phases of the study. We find it necessary to focus on only one error in reaching our conclusion to reject the S&P estimates. Like Morgan, Gunderson obtained virtually all of his data from the Bechtel Study. Gunderson (S) (Jan. 30, 1980) at 24-25. The six engineering firms hired to do the Bechtel Study conducted inspections of sample bridges[103] and completed for each bridge a standardized form that listed 90 or more bridge features to check for needed repairs. See 1 1979 Study App. at IV-2-8 to 13. The inspectors also took extensive photographs and conducted interviews with railroad personnel but Gunderson and his team relied only on the information in the field forms. Gunderson (D) (May 14, 16, 1980) at 175-77, 578-79, EL App. F at 29-31, 70-71. The bridges in the sample were divided into three groups according to construction material: steel, masonry, and timber. The lengths of the bridges were added together to obtain the total number of bridge track feet (BTF) falling in each category. Relying on the information found in the field forms, Gunderson and his staff then placed each sample bridge into one of three possible repair classifications: no repair, light repair, or heavy repair. For example, Gunderson determined that, of the total BTF in steel bridges, 30 percent needed heavy repair, 65 percent needed light repair, and 5 percent needed no repair. 1979 Study at IV 25-27. To determine the bridge rehabilitation needs of each individual T, Gunderson took the total inventory of the T's BTF in each bridge category, and applied the repair percentages he had gleaned from the sample. Assuming EL had 1,000 BTF of steel bridges, 300 BTF would be deemed to need heavy repairs, 650 BTF to need light repairs, and 50 BTF to need no repairs. See generally 1979 Study at IV 20-27; 2 GP at E 57-87. To transform these repair percentages into dollar amounts, Gunderson developed a work function for each bridge category and repair classification. The work function was simply a standardized repair cost for each classification. For example, Gunderson *1244 developed a standard cost per BTF for light repairs on steel bridges, a standard cost per BTF for heavy repairs on masonry bridges, and so on. The final step was to multiply the total BTF in each repair category by the applicable cost figure. The fatal flaw in Gunderson's study is that the sample from which he drew his data lumped together bridges of all the Ts. Therefore, his data and conclusions reflect, if anything, the average condition of all the bridges conveyed by the Ts. For the purposes of deciding the rehabilitation needs of any individual T, however, the statistics are meaningless without proof that the bridges of all the Ts had fallen into the same state of disrepair. There is no credible evidence to support that assumption. A valid study of this nature should have been based on separate samples for each T with separate repair percentages for each T. Indeed, this was the original approach adopted by the S&P team, but it had to be abandoned because of a scarcity of data for some of the Ts' lines. Gunderson (D) (May 19, 1980) Ex. 27, EL App. G at 203-04; id. at 672-76, EL App. F at 72-76; 3 GPRB at App. A 30-31. In defense of his decision to lump all the Ts' bridges together rather than obtain repair percentages for each separately, Gunderson cited several reasons to support the assumption that the bridges of all the T's had fallen into the same state of disrepair. He pointed out that all the Ts were bankrupt, that the bridges were all subject to comparable climatic conditions, and that Ts all carried comparable commodities and loads on their lines. In addition, he stated that his impression that all the Ts' bridges were in comparable states of disrepair was partly based on conversations he had had at conventions with professional associates familiar with the Ts' maintenance practices. None of these justifications offer very firm ground for his assumption. That all the Ts were bankrupt offers only weak support for Gunderson's assumption, and even that support is undercut because the Ts went bankrupt at different times — some as early as 1967, others as late as 1972. If bankruptcy is indeed an indicator that maintenance is being deferred, then the length of time a railroad is in bankruptcy would indicate how long, and consequently how much, maintenance has been deferred. Gunderson himself admitted that, as to the EL, which went bankrupt in 1972, the fact of bankruptcy would not weigh very heavily as an indicator that EL's bridges were in the same state of disrepair as the other Ts' bridges. Gunderson (D) (May 21, 1980) at 1107-10, EL App. F at 187-90. The second and third reasons — comparable climatic conditions and comparable loads — may have something to do with the rate at which railroad bridges deteriorate with use, but have little to do with whether, and at what rate, a particular railroad chooses to defer maintenance on its lines. Casual conversations at conventions could give one a reasonable basis for assuming that the Ts' bridges were falling into disrepair, but they are not a basis for concluding that all the Ts' bridges had fallen into the same state of disrepair. Not only is there a paucity of support for Gunderson's assumption concerning the similarity in condition of the Ts' bridges, but the data in the Bechtel Study indicate that bridge conditions among the Ts differed substantially. For example, 93.4 percent of the CNJ steel bridges in the sample needed heavy repair, compared with only 7.1 percent of R's steel bridges; the repair percentages for the other Ts' steel bridges fell at various points between those two figures. Gunderson (D) (May 23, 1980) Ex. 76, EL App. G at 303. Similarly, wide disparities exist among the Ts in the repair percentages for masonry bridges. See Gunderson (D) (May 23, 1980) Ex. 76, EL App. G at 302. We must, of course, be careful to refrain from attaching too much importance to these figures for the individual Ts because the samples for each T are not large enough to give a reliable picture of bridge conditions throughout each T's system. Nevertheless, to the extent these figures provide at least some hard evidence of the comparative condition of the Ts' bridges, they undercut Gunderson's position. The Ts assert that the S&P Study contained numerous other errors including the *1245 misinterpretation of Bechtel Study field forms, the failure to make use of other Bechtel Study data such as photographs and field interviews, and the use of inaccurate bridge inventories for some of the Ts. See generally 1 ELF 230-80; 2 RF at 384-434. Although some of the Ts' objections may have merit, particularly the criticism of Gunderson's interpretation of the data on the field forms,[104] we find it unnecessary to resolve these arguments. Because the GPs have not proved that the bridges of all the Ts were in the same state of disrepair, averages reflecting the condition of the bridges of all the Ts lumped together cannot yield reliable results when applied to the bridges of each T individually. It would be pure happenstance if that sort of statistical technique yielded accurate results. 3. Yards S&P's estimates for the RC of the Ts' yards are open to the same objection as are the bridge estimates: the data from which repair percentages were developed came from a sample that lumped the yards of all the Ts except PC together.[105] Again, this statistical technique rests on the unproved assumption that all the Ts' yards had fallen into the same state of disrepair. Moreover, the sample was not representative of the relative size of the various Ts. Of the 27 yards in the sample, only three were owned by EL, 1 ELF at 209; H. Morgan (D) (Apr. 23, 1980) Ex. 18, EL App. E at 73-75, even though EL was the largest T in the sample, proffering an average of over 2,000 miles of mainline track. See 1980 Study at VIII 5.[106] Finally, the S&P Study did not even approach the sampling frequencies specified by the Bechtel Study, nor was any consideration given to how large a sample had to be to give reliable results. See H. Morgan (D) (Apr. 23, 1980) at 1540, 1584-87, EL App. C at 205, 228-31. The S&P yard study used other questionable statistical techniques, as illustrated by the following two examples. The study divided yards into three classifications — small, medium, and large. For the small yards, there were inspection results for only nine turnouts. Six of these were located in a single CNJ yard and all were deemed to need replacement. On the basis of those figures, the S&P Study concluded that 67 percent of all turnouts in all of the small yards owned by the Ts except PC needed to be replaced.[107] See H. Morgan (D) (Apr. 23, 1980) Ex. 18, EL App. E at 75; H. Morgan (D) (Apr. 23, 1980) at 1582-84, EL App. C at 226-28. Turnout RC estimates for large yards were not much sounder. Not a single EL or LV turnout was inspected. EL was nevertheless charged with RC (before salvage credit) for turnouts in its large yards ranging from $16.9 million to $26.5 million *1246 on its four proffers; LV was charged $7.4 million and $8.7 million on its two proffers. 1 ELF at 213; H. Morgan (D) (Aug. 20, 1980) Ex. 248, EL App. E at 258-64. 4. The Government Parties' Defense In defense of the S&P Study, the GPs point to several studies and sets of statistics, developed by third parties, that purportedly confirm the accuracy of S&P's estimates. They cite the testimony of a PC witness, Clarence Jackman, who performed a RC and MOWE study on the PC lines. His figure for total RC and MOWE on the IFLs over an eleven-and-a-half-year period was about ten percent lower than S&P's figure. The GPs contend that the closeness of Jackman's results to their own confirms S&P's estimates for the other Ts. EL has a convincing answer: the methodology used to determine PC's needs was not the same as that used for EL and LV. Morgan stated that he used track charts to aid in determining the condition of the Ts' lines. He admitted, however, that track charts were not used in his studies of EL and LV. H. Morgan (D) (Apr. 22, 1980) at 1366-69, EL App. C at 156-59. Similarly, although Morgan used 1976 and 1977 Conrail maintenance information in assessing PC, he had little such information for EL and none for LV. H. Morgan (D) (Apr. 22, 23, 28, 1980) 1369-73, 1465-66, 2063-64, EL App. C at 159-63, 187-88, 427-28. Nor did Morgan have any data regarding interim maintenance work ("section 215 maintenance")[108] done on EL and LV lines pending conveyance of properties to Conrail. He did, however, use such data for PC. H. Morgan (D) (Apr. 22, 1980) at 1369-70, EL App. C at 159-60. Although S&P's line speed and density assumptions were, as stated earlier, demonstrated to be too high in a number of instances relating to EL and LV, its PC data on these items was apparently more accurate, for PC did not challenge Morgan on this issue. ELRB at 56. As to bridges, PC bridges dominated the sample — almost 85 percent of steel bridges and almost 70 percent of masonry bridges were on the IFLs and the Northeast Corridor lines. See Gunderson (D) (May 23, 1980) Ex. 76, 77, EL App. G at 302-03. It is likely therefore that the sample more nearly reflected the condition of PC's bridges than the condition of the other Ts' bridges. Finally, to assess PC's yard rehabilitation needs, the S&P team used a separate sample made up exclusively of PC yards. H. Morgan (D) (Apr. 16, 1980) Ex. 7, EL App. O at 73-75. In contrast, the sample used for the remaining Ts lumps together yards of all the Ts. As this list of methodological differences shows, the S&P team used more complete data and more reliable samples in its study of PC than it used for its study of the other Ts at a number of stages. It is not surprising, therefore, that S&P's overall estimate for rehabilitating and maintaining the IFLs is only 10 percent higher than PC's own estimate. Had S&P used the same procedures for the rest of the Ts, there would be some force to the argument that Jackman's PC study validates S&P's methods, and therefore the reliability of S&P's estimates for the remaining Ts. Since S&P's methods were not the same for the rest of the Ts, however, the reliability of the PC estimates only serves to highlight the doubtfulness of S&P's estimates for the remaining Ts. The GPs also cite statistics for actual rail replacement done by Conrail on the main IFLs between 1976 and 1979 as proof of the reasonableness of Morgan's rail replacement decisions. According to Morgan's testimony, Conrail replaced slightly more rail in those years than Morgan estimated would be required for the same period. H. Morgan (S) (Jan. 30, 1980) at 73-74. Again, because Morgan had considerable data concerning the PC lines that he did not have for the rest of the Ts' lines, his accuracy in predicting rail replacement on the IFLs does not prove the accuracy of his rail replacement estimates for the remaining Ts. As a final argument in support of Morgan's rail replacement decisions, the GPs point to the results of a study conducted by AA on its Toledo-Ann Arbor line. The AA *1247 study estimated that about 18 miles of rail needed replacement, a figure almost the same as Morgan's. See 1 GPRB at E 13 & n.50. This, the GPs argue, is further proof of the accuracy of Morgan's rail replacement estimates. This segment is so small in relation to the thousands of miles covered in the S&P study, see 1980 Study at VIII 5, and the difference between Morgan's figures and the figures of other independent studies on other lines is so large, that we cannot attach much significance to this single agreement between AA's study and Morgan's study. Gunderson points to two independently performed studies as proof of the validity of his estimates. The first is a 1977 Conrail Study of the paint condition on its steel bridges. In Gunderson's expert opinion, paint condition is often a good indicator of the overall condition of a bridge. The results of the paint condition study, when reclassified into the repair categories Gunderson used in his own study (heavy repair, light repair, and no repair) were similar. The Conrail Study, like Gunderson's, is a system-wide study encompassing bridges conveyed by all the Ts. It therefore corroborates, if anything, only that Gunderson's system wide percentages are reasonably accurate. As to each T, however, it confirms nothing because neither Gunderson nor the Conrail Study used separate samples for each T. The second study Gunderson cites for support is one performed by the engineering firm of DeLeuw-Novick concerning bridge RC on Northeast Corridor lines. DeLeuw-Novick's estimate for bridge RC on these lines was slightly larger than Gunderson's. This study confirms only the accuracy of Gunderson's estimates with respect to PC lines, an unsurprising result since his sample is dominated by PC bridges and is therefore likely to yield results representative of the condition of PC's bridges. Gunderson (D) (May 23, 1980) Ex. 76, 77, EL App. G at 302-03. Because there is no proof that the bridges of the other Ts were in the same condition as PC's bridges, DeLeuw-Novick's estimate proves nothing about the accuracy of Gunderson's estimates for the remaining Ts. 5. Conclusion We reject as unreliable the figures generated by the S&P Study. The methods concerning three major items of rehabilitation — track, bridges, and yards — introduced uncorrectable errors into the results. Questionable sampling techniques, in two instances based on the unproved assumption that the rail facilities of all the Ts had fallen into the same state of disrepair, as well as a paucity of data, affect various parts of the Study. The incorrect interpretation of the AREA table and the misclassification of a number of the Ts' lines as to density and speed undermine Morgan's track RC estimates. There are large, inexplicable differences between S&P's results and the results of studies conducted by independent third parties. All these factors direct us to the conclusion that we cannot accept S&P's figures. B. Erie Lackawanna's Study The EL Study was done by Wyer, Dick & Co., a consulting firm, under the direction of the firm's Chairman and Chief Executive Officer, Charles Shannon.[109] Wyer Dick conducted the study in 1975 with the goal of determining the rehabilitation needed to restore EL's lines to efficient operation at their historical timetable speeds.[110] Visual inspections of most of EL's track mileage provided the data for the Wyer Dick Study. The inspections were made by EL division engineers who prepared reports of needed rehabilitation work. Shannon (D) (May 25, 1979) Ex. 19, EL App. B at 36-93. Shannon and his team developed *1248 unit cost figures for doing the various kinds of work using EL's usual repair practices and production rates and 1975 material and labor costs. The estimate for EL's total system, including both proffered and unproffered lines, was $50 million. See Shannon (D) (May 25, 1979) 714-31, 742-43, EL App. B at 94-108, 148-49; id. Exs. 28, 31, EL App. B at 109-47, 155-56. Shannon then made three adjustments to determine the RC for each of the four acquisitions proffered by EL. First, for each acquisition he excluded the cost of work assigned to lines not included in the acquisition. Second, to give a true picture of the condition of EL in 1973 (two years before his study), he made a deduction to account for the additional deterioration occurring in the intervening two years (the "shortfall" estimate). Finally, Shannon restated all costs in 1973 dollars. As support for the validity of Shannon's estimate, EL offers the results of a study conducted independently by Chessie at about the same time Shannon did his study. See Preliminary Engineering Plan for Selected Property Acquisitions of Erie Lackawanna, Penn Central, Reading (Chessie Study), Shannon (D) (July 10, 1979) Ex. 36, EL App. B at 199-239. The Chessie Study estimated the cost of rehabilitating the portion of EL's system that Chessie surveyed at about $28.1 million in 1976 dollars. Shannon (D) (July 10, 1979) Ex. 36, EL App. B at 220. Shannon's estimate for essentially the same properties was $36.5 million in 1975 dollars.[111] In conducting their study, Chessie's engineers were aware of, and specifically rejected, significantly higher estimates prepared by DeLeuw, Cather as part of the Bechtel Study. Chessie Study at 14-15, EL App. B at 210-11. The GPs open their attack on Shannon's estimates by pointing out that they did not include costs for rehabilitating bridges, yards, and signals. Although it is true that Shannon did not include those costs, they are accounted for elsewhere in the study. As to bridges and signals, Shannon determined that what deterioration had occurred would be rectified by the increase in expenditures he projected for routine annual maintenance.[112] Shannon (D) (July 9, 1979) at 983-84, EL App. B at 195-96. As to yards, Shannon concluded that, with the exception of two facilities, no substantial problems existed and that any needed work would be covered by the projected increases in routine maintenance. Shannon did make allowance in his rehabilitation budget for the two EL yards, Marion and Croxton, that he concluded were in need of rehabilitation. Chessie's RC figures support the conclusion that the yards, signals, and bridges needed no substantial repairs. Chessie allotted only $815,200 for yards, $980,800 for signals, and $3,000,000 for bridges. In view of how low these estimates are, it is plausible that increases in routine maintenance would cure the minor deterioration that had occurred. The GPs next argue that Shannon's estimates are too low because the study attempted to ascertain only the work necessary to meet minimum standards. They point to a directive issued by EL's chief maintenance of way engineer to the inspectors who gathered the data for Shannon's study. The inspectors were instructed to *1249 inspect only those lines not up to the 50 mph operating speed that was stated as the goal for EL's main lines. Shannon (D) (May 14, 1979) Ex. 15, 18 GP Supp.App. at 584. The GPs contend that this proves that the study was concerned not with making a complete evaluation of EL's system, but only with eliminating slow orders. Shannon agreed in his deposition that a proper rehabilitation study requires a more detailed inspection than the inspection ordered by Chessie's chief maintenance of way engineer. Shannon (D) (May 24, 1979) at 671-72, EL App. B at 150-51. He also testified, however, that he specifically countermanded that order. Id. Although he apparently did not countermand the instructions in writing, the reports filed by the inspectors show that they viewed most of EL's lines, not just those with slow orders. See EL App. B at 36-93; EL App.N pocket part. Moreover, even for those segments that were not inspected, significant track rehabilitation was ordered based on data obtained from maintenance records and the personal knowledge of EL's engineers. See ELRB at 69-70; EL App. N at 114-15. The GPs also ask us not to attach any importance to the estimates developed by the Chessie Study, contending it was merely a preliminary report and that Chessie officials undoubtedly contemplated further evaluation of EL's rehabilitation needs. Although the Chessie Study was not as detailed as either EL's or S&P's study, its methodology was reasonable. Chessie engineers obtained statistical data concerning the installation dates of EL's rail and nonrail facilities along with records of the maintenance performed on them, and confirmed this data by conducting visual inspections of portions of EL's lines. E. Johnson (D) (Mar. 16, 17, 1978) at 4-8, 42-44, EL App. B at 269-73, 32 GP Supp. App. at 306-308. Moreover, Chessie's senior assistant chief engineer in charge of the study testified that he and his team developed these figures with the idea that Chessie's financial planners would rely on the study in determining whether to purchase a portion of EL. E. Johnson (D) (Mar. 16, 1978) at 7-8, EL App. B at 272-73. Since our purpose is to determine how a potential acquirer would view EL's lines, the Chessie Study, on which Chessie was willing to rely in making its acquisition decision, is strong evidence of the reasonableness of EL's estimate. The GPs argue that if the Chessie Study is used to confirm the EL Study, both Chessie's RC estimate and its annual MOWE estimate must be considered together. They point out that although Chessie's RC estimate is somewhat lower than EL's, the Chessie's annual MOWE figure is significantly higher. The GPs therefore suggest that a true picture of the Chessie Study's results can be obtained only by adding together RC and ten years of annual MOWE. As calculated by the GPs, the total of RC and MOWE using Chessie's estimates yields a figure considerably higher than a comparable total using EL's estimates. In fact, the Chessie total falls about midway between the total using EL's estimates and the total using S&P's estimates. See 1 GPRB at E 26. The GPs would therefore have us conclude that the Chessie Study does not confirm EL's study any more than it confirms S&P's study. As EL points out, the GPs have introduced no evidence that combining RC and MOWE in the manner they propose is a valid technique for comparing the results of Chessie's and EL's studies. In fact, EL's expert, Shannon, testified that such a comparison could not be made. Shannon (D) (July 10, 1979) at 1157-60, EL App. N at 134-35, 18 GP Supp.App. at 358-359. No other witness gave contrary testimony. Shannon did agree with the GPs, however, that RC and annual MOWE are to some extent interchangeable, id., and this proposition is supported by other testimony as well.[113] B. Murphy (D) (June 20, 1980) at *1250 138-39, 2 LV App. at 262-63; see LV at 50. Indeed, both Shannon and Adik, LV's expert, implicitly rely on the notion of interchangeability in excluding certain items of repair — e.g., bridges and signals — from their RC estimates. Their rationale for doing so, which we have accepted, is that they provided for those repairs by increasing their annual MOWE estimates. Shannon (D) (July 9, 1979) at 983-84, EL App. B at 193-94; Adik (S) (EL) (Dec. 31, 1978) at 14. Although RC and annual MOWE are perhaps not interchangeable dollar for dollar, we conclude that there is validity to the GPs' argument and that at least a rough comparison can be made along the lines they suggest. Even though we agree with the GPs that the comparative technique they suggest is useful, the GPs have failed to convince us that the Chessie Study does not support EL's estimates. EL points out several flaws in the figures the GPs used to compute their 10-year totals.[114] After corrections for these errors are made, the total of combined RC and 10-year MOWE using Chessie's estimates moves considerably closer to the total using EL's estimates. As adjusted, the EL's 10-year total is 85 percent of Chessie's total.[115] In contrast, the S&P's total is 160 percent of Chessie's. The revised pro formas submitted by EL in response to our June 19, 1981 letter demonstrate how inconsequential the difference between Chessie's and EL's estimates actually is. EL substituted Chessie's RC and annual MOWE figures for its own in the financial statements relating to a Chessie acquisition of a combined EL-LV system. The difference in Annual Cash Flow on the Cash Flow Statement averaged less than $1 million per year. Compare Memorandum of the Erie Lackawanna Trustees in Response to the Court's Letter of June 19, 1981, (July 10, 1981) at second table following page 10, Sp. Ct. Rep. at N 36177, with id. at fourth table following page 7, Sp. Ct. Rep. N 36172. We do not think this difference is substantial enough to erode the corroboration the Chessie Study provides for EL's figures. The GPs make two final arguments relating to two of the adjustments Shannon made to his raw RC figures. First, they contend that Shannon erred in his calculations concerning the restatement of his estimates from 1975 dollars into 1973 dollars. In deriving costs in his study Shannon used wage rates prevailing on January 1, 1975, and materials prices prevailing on March 31, 1975. To adjust his estimate to 1973 dollars, Shannon relied on indices calculated by the Association of American Railroads (AAR) that provide a basis for comparing year-end wage and materials costs for a given year with the year-end costs from previous years. Shannon used the indices *1251 comparing year-end 1975 figures with year-end 1973 figures. The GPs argue that since the wage and materials costs he used in his study were those prevailing at the beginning of 1975 rather than at the end of 1975, it would have been more accurate to compare 1974 year-end figures with 1973 year-end figures; i.e., only a one-year adjustment was needed to restate Shannon's estimates in 1973 dollars. EL has a ready, if not entirely convincing answer. See ELRB at 67 n.59. As to labor prices, they contend that wage rates generally become effective on January 1 of any given year, and that a two-year adjustment is therefore necessary to get from 1975 labor rates to 1973 labor rates. Similarly, as to materials prices, materials used in 1973 maintenance would have been purchased in early 1973 before work actually got under way, so a two-year adjustment back from March 1975 to March 1973 was also appropriate. We are uncomfortable with EL's explanation concerning the two-year adjustment to labor costs, largely because the choice between a one-year and a two-year adjustment hinges on the time of year that labor rates actually increase, and we can find no support for the assertion that labor rates generally go up on January 1.[116] Absent such support, the year-end 1974 index for labor appears more appropriate than the 1975 index. There is some support for the argument concerning materials, however, because Shannon did testify that the normal maintenance season begins in March; Shannon (D) (May 24, 1979) at 658, 18 GP Supp.App. at 289. Most materials purchases, therefore, would be made in anticipation of the maintenance season. We conclude that as to labor costs only a one-year adjustment is proper, but that as to materials prices the two-year adjustment made by Shannon is correct. Second, the GPs point to an inconsistency in the calculation of the shortfall adjustment. Shannon made shortfall adjustments to account for the years 1974 and 1975. The GPs note, however, that a legend on a summary sheet identifies the EL figures as an estimate of "catch-up maintenance ... as of April 1, 1975". If that is correct, the GPs argue, no shortfall adjustment should have been made for 1975 because Shannon's figure did not reflect the effects of any further deterioration occurring in 1975. On the other hand, if Shannon's study actually reflected conditions as of the fall of 1975 when EL's lines were inspected, then the shortfall for 1975 was incorrectly computed because it did not account for the effect of § 215 funds provided to EL for maintenance during 1975.[117] There is little doubt that Shannon's study stated conditions as they were in the fall of 1975. Shannon (D) (May 25, 1979) at 714-15, 717, 726-27, EL App. B at 94-95, 97, 106-07. EL does not contest that the shortfall adjustment is incorrect for the reasons stated by the GPs, ELRB at 67 n.59, so we conclude that the shortfall adjustment must be corrected to account for the § 215 funds spent by EL in 1975. EL's RC estimates stand up under the GPs' attack. The GPs have pinpointed no serious errors in Shannon's methodology. The Chessie estimate, prepared by an independent party with no motive to understate costs, corroborates the EL figures. We conclude that EL's figures, as adjusted to correct the two errors identified by the GPs, provide reasonable estimates of the rehabilitation needs of EL's lines. C. Lehigh Valley's Study LV's RC and MOWE estimates were compiled by Stephen Adik, a LV Vice President with experience in a number of phases of railroad operations including maintenance. *1252 Adik-2 (S) (LV) (Dec. 31, 1978) at 2-5. Adik authored two studies of LV, one to estimate the value of all LV lines conveyed to Conrail, another to determine the RC and MOWE of only those lines included in the EL-LV combined systems. He used different techniques to estimate RC in each study. Because we accept the EL-LV acquisition as the most likely outcome in the alternative scenario, we analyze only the study dealing with the LV lines in the EL-LV acquisition.[118] Adik developed his RC estimate for those lines from information and work papers submitted to the ICC in the 1974 Ex parte 305 proceeding. In that proceeding, a number of railroads sought rate increases to provide funds for reducing deferred maintenance and for making capital improvements. To justify the increases the railroads presented their estimates of necessary rehabilitation. LV estimated that $28.1 million was required to rehabilitate its entire system. See LV at 78. Adik adjusted this figure to eliminate the costs attributable to unproffered lines and to restate the estimates in 1973 dollars. He also reduced the Ex parte 305 estimates to account for differences between the maintenance work practices of LV and the more efficient work practices of a potential acquirer.[119] Finally, Adik eliminated the costs attributable to bridges, tunnels, and signals because they were small enough to be absorbed in an expanded annual MOWE budget. After these adjustments, RC on LV lines in the EL-LV acquisitions totalled $7.8 million. Adik (S) (EL) (Dec. 31, 1978) at 13-15. Adik based his annual MOWE estimates on a study conducted by Wyer Dick for USRA, which examined the maintenance costs of railroads similar to LV over a 17-year period. That study produced unit cost figures for seven major types of track maintenance. USRA adjusted these figures to 1973 cost levels and made several other slight modifications. These adjusted figures provided the basis for Adik's estimate for those seven major types of track maintenance, and he extrapolated from that estimate to obtain estimates for other types of maintenance not covered in the Wyer Dick study. Adik's estimate for annual maintenance came to $9.8 million per year. Id. at 10-13. Before examining the GPs' major attack on Adik's study, we address two relatively minor contentions concerning adjustments Adik made to the Ex parte 305 submission. The GPs argue, first, that Adik should not have eliminated the costs of nontrack items from his study. Adik's reason for doing so was that the relatively minor deterioration in those items (he placed it at 1.2 years of deferred maintenance at LV's 1973 actual expenditure level, id. at 14) could be eliminated over several years by increased annual MOWE. In light of the small amount of deferred maintenance, about $1 million, and the size of Adik's annual MOWE estimate — his estimate is $2.5 million higher than S&P's — we find Adik's explanation reasonable. The second contention has more merit. The GPs argue that there is no basis for Adik's adjustment to account for more efficient work practices of a potential acquirer. See Adik (S) (EL) (Dec. 31, 1978) Schedule 14 at 1. Adik admitted that he lacked familiarity with the work practices of the Western railroads named as possible acquirers, and that he had only limited knowledge of Norfolk & Western. The only proposed acquirer with which he had any significant familiarity was Chessie. Adik (D) (May 9, 1979) at 72-73, 8 GP Supp.App. at 519-20. *1253 The main basis for Adik's adjustment was his observation that USRA equipment and work practices under the § 215 program undertaken in 1975 were more efficient than LV's and that such "equipment and practices, in all likelihood, [were] similar to those already in use by most potential purchasers in 1973-74". Adik (S) (EL) (Dec. 31, 1978) at 14. Adik did, however, testify that it would be impossible to predict exactly what work practices would apply after an acquisition, because that is a subject negotiated by labor and management. Adik (D) (May 9, 1979) at 76, 8 GP Supp. App. at 521. As we suggest below, however, see pp. 117-18, infra, labor would have been flexible in renegotiating inefficient work rules with the acquirers, and profitable railroads are likely to use equipment that approaches as nearly as possible the most efficient equipment available. Consequently, we believe that Adik's adjustment of $2.225 million was appropriate. The GPs' major attack on Adik's RC estimate revolves around the testimony of Bryan Murphy. Murphy is now a planning engineer for Conrail, but as LV's Chief Engineer in 1974 he was in charge of the study that produced the Ex parte 305 submission on which Adik based his estimate. Murphy testified that the Ex parte 305 estimate reflected the minimum amount of rehabilitation necessary; that it reflected a "short term `patchwork' approach to rehabilitation". B. Murphy (S) (Jan. 30, 1980) at 10. As such, he urged, the Ex parte 305 submission understates the true rehabilitation needs of LV.[120] Because of his position as Chief Engineer of LV, Murphy's testimony cannot be taken lightly. Nevertheless, we have reservations about it. The Ex parte 305 submission was made in the latter part of 1974. In April of that year, Murphy had produced an estimate for track rehabilitation that placed the cost of rehabilitating track in the entire LV system at $6.5 million. His first Ex parte 305 estimate placed track rehabilitation needs at $18.6 million, nearly three times as much as he had estimated only a few months earlier. His estimate for total RC including both track and nontrack items was $24 million. The President of LV, who was well acquainted with the railroad's condition, felt that even this estimate was too low, and asked Murphy to review and resubmit his report. Murphy's final figure, $28.1 million, is the figure Adik relied on in making his estimate. The sequence of events leading to that estimate indicates that Murphy is susceptible to suggestion and casts doubt on Murphy's assertion that the Ex parte 305 estimate was merely a minimum figure. Moreover, LV had no motive to understate that figure; it was trying to justify a rate increase. If anything, LV had a motive to overstate its rehabilitation needs. In short, we are not convinced that the Ex parte 305 estimate is as conservative as Murphy says it is. Fortunately, we need not resolve the issues whether and how much the Ex parte 305 submission understates LV's rehabilitation needs, because the overall results of Adik's study do not differ markedly from S&P's results. Although Adik's RC estimate is substantially lower than S&P's, his annual MOWE estimate is substantially higher.[121] Accepting that RC and annual *1254 MOWE are to some extent interchangeable and using the rough method of comparison suggested by the GPs in their argument concerning EL, the two studies yield results only seven percent apart. Using S&P's figures, the total of combined RC and ten-year MOWE is $113.36 million; with LV's figures the total is $105.44 million. See H. Morgan Corrections (S) (June 6, 1980) Tables VIII-2aA, VIII-2bA; Adik (S) (EL) (Dec. 31, 1978) Schedules 11, 14 at 1. Given the demonstrated unreliability of the S&P Study, and the tendency of that study to overstate RC, we have no reluctance in accepting Adik's slightly lower figures. Even with Adik's figures, the RC per track mile assessed against LV is substantially higher than the cost assessed against either EL or R.[122] D. Ann Arbor AA takes an approach different from EL and LV. AA maintains that its Toledo-Ann Arbor line would have needed no rehabilitation at all because its acquirer would have chosen to reduce the operating speed on the segment from 40 mph to 25 mph. To support this contention, AA offered the testimony of two witnesses — Michael Weisman, a financial officer for D,T&I which operated AA before adoption of the FSP, and Warren Hale, a consulting engineer with Thomas K. Dyer, Inc. Weisman stated that from his vantage point as a cost expert it would not make any sense to maintain the 40-mile stretch in conditions above that necessary to operate at 25 mph. Weisman (D) (June 28, 1979) at 29, 3 AA Supp.App. at 190. Hale corroborated that testimony by stating that because of the small size of the line and the light density of traffic running over it, the Toledo-Ann Arbor segment could be operated at 25 mph with little or no adverse impact on rail operations. Hale (S) (Oct. 15, 1980) at 8. The GPs attack this evidence as insufficient to prove that an acquirer would have reduced the line's operating speed. They contend that Weisman was not qualified to offer this sort of testimony because there was no showing that he had any expertise in railroad engineering, maintenance, or operations, areas the GPs contend are essential to the making of such a prediction. 3 GP at K 43-44. We disagree. As an expert in the financial aspects of railroad operations, Weisman is well suited to make judgments concerning the impact a reduction in speed would have on profitability; he need not have maintenance experience to determine that a line can operate as profitably at 25 mph as at 40 mph. Moreover, the testimony of Hale provides a corroborating view from a railroad maintenance engineer whose railroading credentials the GPs acknowledge. See 1 GPRB at E 55. Finally, the GPs have offered no evidence to rebut this contention. Although AA's evidence on this point could have been stronger, we deem it sufficient to establish that AA would have been operated at 25 mph. We cannot, however, accept AA's argument that no rehabilitation would be needed if the operating speed were dropped to 25 mph. AA offered no credible testimony to support that view. AA's witness Hale stated that some rail replacement would still be necessary on AA's line at the reduced speed, although not nearly as much as Morgan predicted. Hale (S) (Oct. 15, 1980) at 8. The single piece of evidence AA cites is a letter from AA's trustee to the FRA listing slow orders on AA. There were but two slow orders calling for speeds *1255 below 25 mph in existence in March 1975, and only one of the two was caused by track deterioration. AA argues that this absence of slow orders proves that no rehabilitation is needed to operate the line at 25 mph. However, the absence of slow orders does not mean that the line is not on the verge of requiring slow orders. Other evidence submitted by both AA and the GPs suggests that at least some rehabilitation would have been needed even at a lower operating speed. Operating speed has a direct impact on the rehabilitation standards applied to track, but, as the GPs point out, it does not affect those standards relating to bridges, yards, and signals. 3 GPRB at App. A 82-83 & n.171. See generally 1980 Study at IV 18-34. Indeed, Morgan testified for the GPs that reducing the speed would not require a reduction to the original S&P estimate of even 50 percent. H. Morgan (D) (Aug. 19, 1980) at 4574, 28 GP Supp.App. at 346. And as to track rehabilitation, where a reduction in speed would have its greatest impact in reducing rehabilitation, even AA's witness Hale would not go so far as to say that lowering the speed would eliminate all rail replacement. Both Hale and Morgan had originally called for about 18 miles of rail replacement if the operating speed remained at its historical level. See 1 GPRB at E 13. Hale stated that two of these miles would still require replacement at a 25 mph operating speed. Hale (S) (Oct. 15, 1980) at 8-9. And as the GPs point out, the remaining 16 miles not replaced would probably still require some less drastic forms of rehabilitation work.[123] We conclude that some rehabilitation would be required even if the line's operating speed were dropped to 25 mph. Our problem is to determine how much would have been needed. At the outset, we reject Morgan's statement that, at most, reducing AA's operating speed would have cut his estimate of RC in half. We have already detailed the problems in his initial estimate; adjusting a flawed initial estimate for speed variations would yield only a flawed result. Furthermore, a reading of the transcript of the deposition at which Morgan offered this judgment reveals that Morgan had no opportunity to consider seriously the effect of a reduction in speed.[124] Indeed, the GPs do not rely on Morgan's remark. See 3 GPRB at A 82. The only considered evidence on this point is Hale's rebuttal testimony. As noted earlier, Hale stated that the rail on two miles of track would have to be replaced in order to rehabilitate AA to 25 mph operating speed. At his deposition, Hale also stated that no surfacing beyond what would be provided for in normal maintenance would be required. Hale (D) (Feb. 12, 1981) at 130-31, 1 AA Supp.App. at 155-56. The GPs challenge Hale's assessment of AA's track rehabilitation needs because he relied on only two sources in making his judgments — track charts (showing the weight and date of installation of rail) and Sperry Defect Reports (annual reports of rail failures due to rail defects). 3 GPRB App. A 84. The GPs mischaracterize Hale's testimony, however. While these may have been the only written sources he used, Hale also testified that he relied on the knowledge he gained in preparing his original testimony. Hale (D) (Apr. 11, 1979) at 16, 25 GP Supp.App. at 471. Furthermore, the *1256 GPs have presented no evidence showing that Hale's reliance on these two sources was improper. Given that Hale was limiting himself to a relatively small length of track of which he had detailed knowledge and that the GPs admit that he is an expert in this area, see 1 GPRB at E 14, we see no reason why his judgment on rail replacement cannot be trusted. Hale made no attempt in his rebuttal testimony to estimate AA's yard, tie, signal, turnout, or bridge rehabilitation needs assuming a reduced operating speed. The only other evidence on AA's rehabilitation needs is Hale's original study. The GPs correctly identified a problem with relying on this testimony, however. Although most witnesses defined RC as the amount of work necessary to bring the Ts' lines up to good operating condition or the condition "at which the asset should be maintained if operations are to be continue [sic] over an indefinite period of time", Hale (D) (May 16, 1979) at 4, 11 GP Supp.App. at 783; see note 110, supra, Hale's study was designed to determine what expenditures would be necessary to keep AA running for three to five years. Hale (D) (May 16, 1979) at 3-4, 11 GP Supp.App. at 783-84. But while Hale may have understated the AA's rehabilitation needs in this respect, he clearly overstated them in another. Hale testified that, under the federal classification scheme, to operate AA at 45 mph, as he assumed, it would have to be maintained at the level of a 60 mph railroad. Thus Hale was applying much more stringent standards for track rehabilitation than he would have done had he assumed a 40 mph operating speed, as did Morgan and Gunderson. We suspect that Hale's use of excessively stringent standards offset his short-term perspective. Our impression is reinforced when we observe that Hale and Morgan reached strikingly similar results in determining the amount of track replacement necessary.[125] Our reasoning applies equally to all components of rehabilitation affected by operating speed, including mainline ties and turnouts. Since we have no basis for adjusting these numbers to reflect a 25 mph operating speed and since AA has the ultimate burden of proof on this point, we charge AA with the entire cost of these items as estimated by Hale. Several components of RC, however, are not affected by main line operating speed. For rehabilitation of these facilities — yards, signals, and bridges — we have more difficulty in accepting Hale's estimates because no countervailing error offset the short-term perspective of Hale's study. On bridges, however, neither can we adopt the GP's estimate. The S&P study concluded that RC for AA's bridges would amount to $480,000. H. Morgan Corrections (S) (June 6, 1980) at Table VIII-2aA. But, as AA points out, the S&P sample included only five AA bridges. AA at 162; Gunderson (D) (July 21, 1980) Ex. 146. Of the five AA bridges considered, none required anything more than light repair. See AA at 162. Thus, what little evidence the GPs adduced on AA's bridges tends to support Hale's conclusion that there is no need for rehabilitation work, and that is the conclusion we accept. With regard to signals and yards, we are confronted with a problem that does not lend itself to a similar resolution. S&P estimated that RC for signals would be $10,000 and that yard rehabilitation would cost $1,040,000. Hale, on the other hand, does not seem to have offered any figure for rehabilitation of signals, and concluded that yard rehabilitation could be accomplished for $45,000. We know of no way to adjust Hale's figure explicitly to correct for his short-term perspective or to adjust S&P's figure to account for the errors discussed earlier, and we doubt that any potential acquirer would have been able to make such adjustments. A potential acquirer confronted with these figures would have weighed the shortcomings of both estimates and arrived at an educated guess. On yards and signals, then, we find that a potential acquirer would have expected to incur costs of $250,000 in rehabilitation. *1257 The only factor left to be determined is how long the rehabilitation program would have taken to complete. Morgan and Gunderson projected that it would take two years — Hale said only that it would take "several". See Hale (S) (Oct. 15, 1980) at 7-8. Since we do not know how long "several" years is, we accept the GPs' evidence on this point. The following table reflects the rehabilitation costs which we believe an acquirer of AA's lines would have expected to incur: Item Amount Unit Cost Item Cost Main line rail 2 miles $75,000 $150,000 Main line ties[126] 6,288 20 125,760 400 25 10,000 Turnouts 8 7,500 60,000 Surfacing NA NA 0 Bridges NA NA 0 Yards and Signals NA NA 250,000 _________ Total 595,760 E. Annual Maintenance of Way Expenses None of the parties devoted a great deal of attention to annual MOWE in their briefs, choosing instead to focus on rehabilitation. Accordingly, most of the Ts can be dealt with briefly. As demonstrated above, LV's MOWE figure is higher than that of S&P and that fact was important in our decision to accept LV's rehabilitation estimate. We do not expect that the GPs will object to our acceptance of LV's MOWE estimate as well. As for AA, the GPs put annual MOWE at $609,000 while AA's witness estimates it at $450,000. H. Morgan Corrections (S) (June 6, 1980) at Table VIII-2bA; Laurie (S) (AA) (Dec. 31, 1978) Ex. 1 at Table 4. Since Laurie was the only witness who based his estimate on the assumption of a 25 mph operating speed, see id. at 6, and since the GPs offer no reason why that estimate understates true MOWE, we accept Laurie's figure. This leaves us with EL. As experts for both EL and the GPs have testified, RC and annual MOWE are to at least some extent interchangeable; one engineer may treat a particular repair item as rehabilitation, another may treat it as MOWE.[127] Any attempt to combine EL's RC estimates with S&P's MOWE estimates therefore incurs the risk of excluding some items of repair and duplicating others. We *1258 nevertheless deem it necessary to examine S&P's MOWE estimates to determine whether there is some good reason to substitute them for those of EL. We find none. To determine MOWE, S&P developed maintenance cycles for each type of rail facility. Each cycle described the amount of annual maintenance needed for a particular facility based on its expected life and maintenance needs. For example, S&P estimated that new 140-pound rail in use on lines with densities of 20-30 MGT needs replacement at a rate of 3.57% per year. 1980 Study at IV 13. How much maintenance each T needed of course depended on the size of its inventories of the various facilities studied. S&P also developed maintenance work functions, similar to the rehabilitation work functions, to provide unit costs for various kinds of maintenance work. Applying these unit costs to the maintenance needs defined for each T produced annual MOWE estimates. Comparing EL's MOWE figures with S&P's is difficult because EL did not separately calculate MOWE. Those expenses are accounted for as reductions to net operating income in EL's pro forma statements. According to its own calculations, EL estimates that its MOWE figure for the EL-LV Western Line acquisition is about $9 million lower than S&P's. EL contends that this difference can be explained by a number of errors in S&P's methods that tended to overstate MOWE. At least some of EL's contentions have merit. EL points first to inaccuracies in S&P's data concerning the amount of track mileage in the EL proffers. The primary error concerns Morgan's overstatement of yard track miles. He used a Conrail system-wide ratio of yard track to main track mileage to estimate the yard track mileage in EL's proffers. Ratios specific to EL were readily available, and EL contends that Morgan should have used them instead of Conrail's system-wide ratios. 1 ELF at 218. The GPs argue that there is no record support for that assertion. 3 GPRB at App. A 194. Common sense, however, indicates that a ratio specific to EL would give a more reliable estimate of EL's yard track mileage than would Conrail's system-wide ratio, which of course covers lines of all Ts. To the extent that the use of Conrail's ratio overstated EL's yard track mileage, S&P overstated yard maintenance.[128] EL also contends that in developing his MOWE figures, Morgan failed to take into account the effect of his rehabilitation program, resulting in a duplication of rail replacement. For example, Morgan replaced 1,070 miles (40%) of the 2,650 miles in the Western acquisition. See H. Morgan (D) (Aug. 14, 1980) Ex. 185, EL App. E at 238. According to the table in the 1980 Study at IV 13, the replacement rail had a useful life of at least 40 years.[129] Morgan did not exclude or otherwise adjust for this substantial amount of new rail on EL's lines in estimating annual rail replacement. He projected 56 miles of rail replacement per year, which, over a period of 40 years, amounts to 2,240 miles. See H. Morgan (D) (Aug. 20, 1980) Ex. 247, EL App. E at 256. At that rate of rail replacement, Morgan would replace over 600 miles of rail in EL's lines more than once within a 40-year period. *1259 We do not accept EL's argument that the replacement of more than all of the rail over the 40-year period proves that Morgan's MOWE estimates were excessive. On the contrary, Morgan assumed that the replacement that was part of the rehabilitation program had already occurred when he stated that relaying 2.5 percent of the rail each year would maintain the track in average condition. H. Morgan (D) (May 5, 1980) at 2608-20; EL App. D at 77-89; 3 GPRB at App. A 208-10. Thus, over 40 years beginning from the time rehabilitation was completed, Morgan's maintenance program would replace one hundred percent of the rail, which is exactly the result EL seems to be urging. Yet we do think Morgan's figures are inaccurate to an unquantifiable extent for a reason suggested at his deposition. Although it is reasonable to assume that all the rail would require replacement over a 40-year period, we see no reason to assume that the need for replacement would occur evenly over the 40 years. Indeed, given that 40 percent of the rail would have been newly laid during the rehabilitation years, it is probable that the need for replacement would arise disproportionately in the final years of the 40-year period. The present discounted value of a stream of payments made regularly at constant levels would differ substantially from the present value of a stream of initially small payments followed by larger payments, even though the sum of payments in the two streams was the same. See note 54, supra. Morgan testified that he was unfamiliar with discounted value and did not consider it. See H. Morgan (D) (May 6, 1980) at 2879-80, EL App. D at 244-45. EL also makes several convincing arguments concerning Morgan's overstatement of labor costs in certain maintenance work functions. Morgan assumed that productivity in yard maintenance work would be very low; only 40 to 60 percent of work time in yards was considered productive. See 1980 Study at V 1-5. This view is contradicted by the Bechtel Manual, which states that full productivity should be assumed in yards except for weather delays. See Morgan (D) (May 7, 1980) Ex. 105, EL App. E at 186. The large difference in figures suggests that Morgan understated yard work productivity. Morgan also overstated labor costs in some work functions by double counting the labor costs attributable to field mechanics, by including the labor costs of signal maintainers in yard maintenance functions even though yards generally do not have signals, by calculating twice as many man hours as necessary for distributing materials for rail replacement, and by overstating the costs of ballast cleaning. 1 ELF at 295-99. Neither Morgan nor the GPs contest these errors. The GPs effectively rebut most of the other errors asserted by EL. On the whole, S&P's MOWE estimates stand up better than its RC estimates, and that perhaps explains why the Ts' estimates and S&P's estimates are reasonably close to each other. EL has nevertheless convinced us that S&P overstated EL's MOWE costs to some unquantifiable extent. The GPs, on the other hand, have not separately attacked EL's MOWE figures or the methods used to estimate them. In these circumstances, we conclude that, having accepted EL's RC figure, we should accept its MOWE estimates as well. F. Conclusion The efforts of the Ts, and particularly EL, have exposed a number of errors in the S&P Study sponsored by the GPs. The methods and data employed in several parts of the study were either incorrect or so questionable as to negate any reliability of the estimates it produced. We are convinced that the S&P Study has overstated, to a substantial though unquantifiable degree, the rehabilitation needs of the Ts. Although S&P's MOWE estimates do not suffer the same shortcomings as its RC estimates, we are also convinced that, at least with respect to EL, S&P's MOWE estimate is too high. The GPs have similarly exposed errors in the estimates presented by the Ts. Unlike the errors in S&P's study, however, those in the studies relied on by two of the Ts — EL and LV — do not undermine the reliability of the studies as a whole. Furthermore, the errors are all quantifiable and can be corrected by relatively simple recalculations. *1260 We therefore decide that the RC and MOWE figures submitted by EL and LV, as corrected to account for the errors detailed in our discussion of those Ts' studies, provide reasonably accurate estimates of the costs that would have faced an acquirer of those lines. As to AA, neither side has provided us with an overall estimate of RC which we can credit. Nonetheless, we believe that the evidence is sufficient to allow us to determine what AA's rehabilitation needs would have been assuming a drop in operating speed. We do fully accept the AA's estimate of MOWE. IX. Labor Protection Costs A. Introduction There has been a long history of collective bargaining agreements, federal statutes, and ICC orders that have provided labor protection for railway workers by guaranteeing employment and compensatory financial assistance to workers adversely affected by various types of organizational changes in the railroad industry. An excellent discussion of the history of such labor protection has been provided by Judge Waterman in New York Dock Ry. v. United States, 609 F.2d 83, 86-90 (2nd Cir. 1979), and there is no need to repeat it here. The GPs argue that the cost of labor protection either would have been borne directly by the Ts — and therefore must be offset against the proceeds of rail use sales — or would have been assumed by the acquirers — thus reducing sales prices directly. 5 GP at S 1-3. The Ts do not contest this. Rather, they contend that the proffered transactions would have resulted in either minimal or no labor protection costs. ELRB at 254-55; LV at 98. The Ts' responses raise two principal issues — (1) the level of labor protection benefits that would have applied in the proffered transactions; and (2) the costs of this level of benefits — which are addressed in the following sections. At the outset we note that labor protection costs would have been a factor only in the case of rail use sales and, as later explained, see p. 1269 & n.142, infra, even in those only to the extent that the excess of sales price over scrap value would have been greater than the labor protection costs. The GPs do not contend that existing labor protection agreements would have been applicable in the event the Ts abandoned their lines and sold them for scrap, and as our discussion of these agreements below suggests, we would reject such an argument. See pp. 1261-1265, infra. Similarly, the GPs appear to concede that the ICC would not impose labor protection responsibilities on the Ts in the context of complete abandonment of the Ts' lines for scrap. 5 GP at S 147. In any event, the ICC's decisions under sections 1(18) and 5(2) of the Interstate Commerce Act, 49 U.S.C. §§ 10901, 11701, strongly suggest that no labor protection payments would have been required if a T had abandoned all its lines.[130] Moreover, we would have considerable doubt regarding the constitutionality of ICC imposition of labor protection conditions in the abandonment of a hopelessly losing railroad.[131] *1261 B. Level of Labor Protection 1. Existing Agreements Discussion of the levels of labor protection that would have accompanied the proffered transactions must begin with an examination of the agreements and statutes bearing on the subject. These would have been of obvious importance in litigation and would have been an important consideration in negotiations between labor and the Ts. Although the GPs contend that a number of existing labor protection agreements would have been applicable to the proffered transactions,[132] their arguments with respect to the Ts remaining in the case focus almost exclusively on two such agreements — the Washington Job Protection Agreement (WJPA) and the Erie Lackawanna Merger Protection Agreement (ELMPA). The WJPA, entered into in 1936 by representatives of most rail carriers and railway unions, is binding on all remaining Ts either by succession or adoption, and provides protections for all employees represented by labor organizations party to the agreement. Ex. GPS-Labor 10. The agreement guarantees protected employees a choice between a monthly dismissal allowance of up to 60 percent of a base period salary for a maximum of five years following dismissal or a lump-sum separation payment. Both types of benefits vary with the length of an employee's service with the railroad. The Ts contend that the WJPA would not have been applied in the circumstances of the proffered transactions. 2 ELF at 540-51; LV at 110-14. The WJPA is applicable to employees adversely affected by a "coordination", defined as: joint action by two or more carriers whereby they unify, consolidate, merge, or pool in whole or in part their separate railroad facilities or any of the operations or services previously performed by them through such facilities. *1262 WJPA § 2(a). The agreement also provides that: the provisions of this agreement are to be restricted to those changes in employment in the Railway Industry solely due to or resulting from ... coordination[s]. Id. § 1. The Ts argue that the proffered transactions would have fallen outside the WJPA's definition of "coordination". They reason that, by completely liquidating its properties and ending its corporate existence, a T would have escaped the WJPA's requirement of "joint action" that "unif[ies], consolidate[s], merge[s], or pool[s]" the properties of two or more carriers. Rather, they argue, the proffered transactions constitute unilateral liquidations, not involving the "joint action" of another carrier. The GPs respond that, insofar as the Ts' lines are to continue in rail use, the proffered transactions are indistinguishable from ordinary mergers, which typically involve the expiration of one party. 2 GPRB at S 110-11 & n.273. We find the terms "coordination", "joint action", and "unify, consolidate, merge, or pool" ambiguous and unilluminating. The parties have produced little evidence bearing on the meaning of these terms and neither they nor we have found precedent relevant to the question.[133] A literal reading of the language of the WJPA would provide some support for the GPs' claim that the agreement would have been applicable in the circumstances of the alternative scenario. We would be reluctant, however, to rest our holding on such a mechanical process of interpretation. The Ts' liquidations present situations unlike the typical merger with which the WJPA is principally concerned. See, e.g., Ables, The History of and Experience Under Railroad Employee Protection Plans 136-37; Ex. GPS-Labor 45. Most notably, ordinary mergers are seldom marked by the characteristics of involuntariness and economic hardship that distinguish liquidations. The important differences between the circumstances of the alternative scenario and those of the subjects ordinarily considered within the scope of the WJPA would make us reluctant to read the ambiguous language of that agreement as applying to the proffered transactions. However, we need not resolve the question of the proper interpretation of the WJPA. The difficult questions of legal interpretation raised by the WJPA would have confronted both the Ts and labor's representatives had the proffered transactions actually occurred. Because of the uncertainty — described above — created by these questions, the parties would have harbored little confidence that their interpretation of the agreement would have prevailed in arbitration. We reject the conclusory opinions of the GPs' witnesses, e.g., Lowry (S) (Jan. 31, 1979) at 37-39; Mahoney (S) (Jan. 31, 1979) at 43-50, that labor organizations would have insisted on arbitration of disputes regarding the WJPA. These opinions were premised on the notion that the WJPA provided clear support for labor's claims, a contention that we reject.[134] We find that the uncertainty that would have surrounded the question of the agreement's applicability in the context of the proffered transactions would have made both the Ts' and labor's representatives reluctant to seek arbitration of the issue. Neither party would have had any confidence regarding the ultimate outcome of arbitration, and an adverse, binding decision would have seriously weakened the loser's *1263 bargaining position. We conclude, therefore, that the parties would have preferred the known hazards of negotiation to the perceived risks of pursuing arbitration to its conclusion. And, as discussed below, we conclude that negotiations would ultimately have produced new labor protection agreements acceptable to all parties. We reach a similar conclusion with respect to the ELMPA, an agreement entered into by EL prior to its inclusion in the Norfolk & Western system in 1968.[135] The agreement incorporated Norfolk & Western's "Nickel Plate Agreement" and guaranteed both "attrition protection"[136] and the protections of the WJPA to all workers employed by EL as of April 1, 1968. Exs. GPS-Labor 30, 32. The ELMPA provided protection for any covered EL employee "deprived of employment or placed in a worse position with respect to compensation, rules, working conditions, fringe benefits, or rights and privileges pertaining thereto at any time during such employment." ELMPA, at App. A § 1(b), EL App. M at 94. EL argues that the ELMPA would have been held inapplicable to the proffered transactions by an arbitration panel, 2 ELF at 527-34, reasoning that the ELMPA was impliedly conditioned on EL's continuation as an operating line. The GPs, in response, contend that "the agreement expressly contemplated the contingency that Erie would be unable to provide work for protected employees at some time in the future and required payment of benefits in all such circumstances." 2 GPRB at S 106. It is far from clear to us that the ELMPA would have been applicable in the circumstances of the alternative scenario. The agreement was reached in the context of EL's inclusion in the Norfolk & Western system, and we would not lightly extend its sweeping protections to the quite different circumstances of EL's complete liquidation. Our view seems to be shared by the ICC which discussed the scope of the ELMPA's applicability in Norfolk & Western Ry. and New York, Chicago & St. Louis R.R. — Merger, 347 I.C.C. 506 (1974), aff'd sub nom. Patton v. United States, No. C-74-841 (N.D.Ohio 1975). EL sought suspension of its labor protection obligations during the course of its reorganization proceedings on the theory that the ELMPA was impliedly conditioned on EL remaining under Norfolk & Western's control — which ceased on EL's entering reorganization. Although the ICC rejected EL's contention, we think its reasoning indicated an appreciation of the limits on the ELMPA's applicability. The Commission stated that only when an employee was adversely affected as a result of "facility coordinations or other operational changes" that took place "pursuant to the inclusion of the E-L in the N&W system" was the ELMPA applicable. Id. at 513. In the Norfolk & Western case, which involved employees adversely affected by transfers and other changes in operating procedures occurring shortly after EL's inclusion in the Norfolk & Western system, this two-pronged test was clearly satisfied. Id. In the circumstances of the alternative scenario there would have been considerable doubt as to whether the ICC's Norfolk & Western standard was met. EL's employees would have been suffering adverse effects nearly a decade after inclusion in the Norfolk & Western system and for reasons unrelated to that inclusion. Moreover, EL's total liquidation and withdrawal from the railroad business bears little resemblance to the small-scale, internal operating changes that the Commission described in Norfolk & Western. In short, the terms "facility coordination[]" and "operational change[]" — particularly when viewed in the context of the Norfolk & Western case — cannot readily *1264 be applied to the quite different circumstances of the alternative scenario. We conclude that these considerations would have created very substantial doubts as to the applicability of the ELMPA in the proffered transactions. This uncertainty surrounding the ELMPA would have been heightened by other questions regarding its applicability. EL finds support for its position in the words "during such employment" in the language of the ELMPA quoted above. EL asserts, not implausibly, that "such employment" presupposes its continued existence as an employer, a position that provides textual support for the two-pronged standard articulated by the ICC in Norfolk & Western. Moreover, even if the ELMPA were deemed applicable in the circumstances of the alternative scenario, there would be a very real possibility that the agreement would have been altered by the ICC. In Norfolk & Western the Commission stated that it would suspend the requirement of attrition protection upon a "showing of substantial detriment to the public interest." Id. at 514. The Commission noted that its prior approval of the ELMPA's attrition protection provisions had been grounded on the fact that these provisions would not "impose a significant burden on any of the carriers involved". Id. at 514 n.3. In the circumstances of the alternative scenario, where EL was in a hopeless financial condition and its liability under the ELMPA might have reached $154 million, Kasher (S) (Nov. 1, 1979) at 43, the ICC could easily have concluded that attrition protection was not only "significant", but so onerous as to require suspension of the agreement. This would have been particularly true if labor's insistence on attrition protection appeared likely to threaten the consummation of the rail use sales. Although we find EL's arguments that the ELMPA would have been inapplicable to the proffered transactions persuasive, we are convinced that neither EL nor labor would have pursued arbitration to its conclusion. EL would have had little incentive to risk an adverse — and costly — decision from an arbitration panel, even though it likely would have regarded its legal position as relatively secure. Labor also would have been reluctant to seek arbitration, principally because it would have perceived the slight prospects it would have had of prevailing. 2. Action of the Reorganization Courts Our conclusions regarding the uncertainty that would have surrounded the applicability of both the WJPA and the ELMPA are supported by the possibility that, even if these agreements were held applicable to the proffered transactions, they could have been rejected in the Ts' reorganization plans. ELRB at 531-34. The Ts rely on § 77(b)(5) of the old Bankruptcy Act, which provided that a plan of reorganization "may reject contracts of the debtor which are executory in whole or in part". The GPs respond that § 77(n) of the Bankruptcy Act, which provided that "[n]o judge or trustee acting under this title shall change the wages or working conditions of railroad employees except in the manner prescribed in the Railway Labor Act", would have prevented the rejection of any of the existing labor protection agreements. 5 GP at S 95. Resolution of the evident conflict between these two provisions of the Bankruptcy Act would have been a difficult task. Although we do not address the question in detail, we do note a claimed disagreement on the issue between a respected authority and the House Report accompanying the Bankruptcy Code. The Ts refer us to a leading bankruptcy treatise's discussion of a provision of the new Bankruptcy Code, 11 U.S.C. § 1167 (Supp. III 1979), which adopts substantially unchanged the provisions of § 77(n): It should be noted that section 1166 [sic] does not preclude rejection of a collective bargaining agreement under section 365 [granting the general power to reject executory contracts] or under a plan. Rejection of a collective bargaining agreement terminates the employer-employee relationship and is not within the proscription of section 1167 which merely prohibits a change in wages or working conditions. *1265 5 Collier on Bankruptcy § 1167.01 (15th ed. 1980). The GPs contend that the unsupported statement in Collier is inconsistent with the language of § 1167, which provides: Notwithstanding section 365 of this title [providing for the rejection of executory contracts] neither the court nor the trustee may change the wages or working conditions of employees of the debtor established by a collective bargaining agreement that is subject to the Railway Labor Act. 11 U.S.C. § 1167 (Supp. III 1979) (emphasis added). Moreover, the House Report accompanying the Bankruptcy Code stated: Section 1167 is derived from present section 77(n). It provides that notwithstanding the general section governing the rejection of executory contracts (section 365), neither the court nor the trustee may change the wages or working conditions of the debtor established by a collective bargaining agreement that is subject to the Railway Labor Act, except in accordance with section 6 of that Act. The subject of railway labor is too delicate and has too long a history for this code to upset established relationships. The balance has been struck over the years. This provision continues that balance unchanged. H.R.Rep.No.595, 95th Cong., 1st Sess. 423 (1977), U.S.Code Cong. & Admin.News 1978, pp. 5787, 6379. Even after legislative efforts to clarify the relationship between §§ 77(n) and 77(b)(5), it is apparent that there remains room for considerable disagreement on the subject. Although our conclusion regarding the uncertainty surrounding the interpretation of the ELMPA and the WJPA makes it unnecessary for us to answer the question, the evident confusion as to the proper relationship between §§ 77(n) and 77(b)(5) would have further muddied the parties' conceptions of their legal rights under existing labor protection agreements. 3. ICC Action In addition to the possibility of labor protection obligations arising from existing agreements, the ICC might have imposed labor protection obligations on the Ts as a condition of its approval of the proffered transactions. The GPs contend that the ICC would have considered the Ts' dispositions under § 5(2) of the Interstate Commerce Act, which requires the Commission to impose labor protection conditions as a condition of approval of a transaction. The GPs also argue that, even if the transactions had been processed under § 1(18), the ICC would have exercised its discretion and imposed labor protection conditions as part of the proffered transactions. 5 GP at S 120-21. Section 5(2) applies, inter alia, to transactions wherein "two or more carriers ... consolidate or merge their properties or franchises, or any part thereof, into one corporation for the ownership, management, and operation of the properties theretofore in separate ownership". Section 1(18) is applicable to "the extension of ... [a carrier's] line of railroad, or the construction of a new line of railroad or ... [the acquisition or operation of] any line of railroad" and to any action whereby a carrier "abandon[s] all or any portion of a line of railroad". The Ts argue that these statutory provisions are illuminated by a series of ICC decisions involving entire line abandonments in which the Commission permitted the railroads to proceed under § 1(18) rather than § 5(2). Okmulgee Northern Ry. — Abandonment of Entire Line, 320 I.C.C. 637 (1964), exemplifies this line of cases. In Okmulgee, a railway applied for permission to abandon its entire twelve-mile line, while a second railroad applied simultaneously for permission to acquire and operate portions of the abandoned trackage. The ICC, rejecting an objection by labor representatives, held that the transactions fell outside the scope of § 5(2), and that § 5(2)(f)'s compulsory labor protection provisions therefore were inapplicable. Id. at 645-46. The ICC explained: This transaction does not involve a merger or consolidation in the sense contemplated by ... [§ 5(2)'s] congressional purpose. There is no consolidation of carrier *1266 interests, or joining of two or more carriers, with a consequent temporization of competition. There is merely the expiration of one carrier which, because of business adversity, is no longer able to successfully compete; and the willingness of another to move into the vacuum to the extent some need might remain for rail service. Id. at 639. The ICC also declined to exercise its discretion under § 1(18) to impose labor protection conditions. The Commission stated that, "the imposition of protective conditions in proceedings ... where a carrier proposes to abandon its entire line of railroad normally is not warranted", id. at 645, and observed that "a departure from this principle in particular cases must be supported by clear and convincing evidence", id. The Commission relied on similar reasoning in later decisions reaching the same conclusion.[137] The GPs argue that the Ts' reliance on the Okmulgee line of cases is misplaced in that "none of the cases on which the transferors ... rely involved a significant impact on regional or national rail transportation because each carrier was small and employed relatively few workers." 5 GP at S 147. The apparent thrust of the GPs' distinction is that there is comparatively little need for § 5(2)(f)'s mandatory labor protection provisions where only a few employees are adversely affected. Under these circumstances, dismissed and displaced employees can be absorbed into other rail systems with little difficulty, thereby reducing the risk of both hardship and labor unrest. Resolution of the question whether § 5(2) or § 1(18) would have been applicable in the alternative scenario presents difficult legal questions. Although we would lean slightly towards accepting the GPs' distinction, we note that in several cases the ICC refused to apply § 5(2)(f) where the track abandonments — and numbers of adversely effected employees — were fairly substantial. E.g., Tennessee Central Ry. — Abandonment of Operations, 333 I.C.C. 443, 444-45 (1968) (287 miles of track abandoned; 279 covered employees affected); Rutland Ry. — Abandonment of Entire Line, 317 I.C.C. 393, 394, 404 (1962) (332 miles of track abandoned; 489 workers affected). These decisions, however, did not occur, as our case would have, in the context of large-scale regional abandonments and reorganizations. Had all the Ts elected to deny their employees labor protection and seek ICC approval under § 1(18), thousands of railway workers would have been affected. Kasher (S) (Nov. 1, 1979) at 15. This situation would have demanded a response from the Commission entirely different from that required in the Okmulgee line of cases cited above. The unusual need for ICC imposition of labor protection that would have arisen in the proffered transactions if the Ts had attempted to avoid making protective payments to their employees might well have resulted in a decision by the Commission to reject the Okmulgee cases and require the Ts to proceed under § 5(2). Moreover, the acquiring lines might well have insisted that the proceeding take place under that section in order to obtain the benefit of the antitrust immunity afforded by § 5(11). Finally, even if the ICC had processed the transactions under § 1(18), we think it might have determined that the standard articulated in Okmulgee governing the discretionary imposition of labor protection conditions was satisfied. See p. 1266, supra. If faced with the task of attaching labor protection conditions to the proffered transactions, the ICC would likely have adopted a set of benefits similar to those set forth in its decisions in New Orleans Union Passenger Terminal Case, 282 I.C.C. 271, 281-82 (1952); Chicago, Burlington & Quincy R.R. — Abandonment, 257 I.C.C. 700, 704-07 (1944); Oklahoma Ry. Trustees — Abandonment of Operations, 257 I.C.C. 177, 197-201 (1944). The Commission relied on these decisions frequently in setting labor protection conditions under §§ 5(2) and 1(18) in the period between 1950 and 1976, and we think it likely that decisions in the alternative scenario would have reflected these *1267 precedents. Despite this conclusion, we doubt that the ICC would have had to decide the issue. Instead, we find that, in light of the uncertainties discussed above, negotiations between the parties would have produced an agreement, thus obviating the need for litigation before the ICC regarding the imposition of labor protection conditions. 4. The Negotiating Process We find considerable support in the views of the parties for our conclusion that the level of labor protection benefits would have been decided through negotiations, rather than litigation or arbitration. According to EL, "EL and the Government Parties agree that a new agreement would have been reached, albeit with substantially different terms." ELRB at 214. LV does not disagree with this statement, but contends that it would not have entered into a new labor protection agreement which afforded LV employees benefits higher than those provided by existing agreements. The GPs correctly observe that "virtually every witness who considered the question supports the conclusion that, if the alleged sales had occurred, the transferors, purchasers, and unions would have negotiated new, comprehensive agreements providing labor protection benefits for adversely affected employees". 2 GPRB at S 6 (footnote omitted). Cf. Norfolk & Western R.R. v. Nemitz, 404 U.S. 37, 42, 92 S.Ct. 185, 188, 30 L.Ed.2d 198 (1971) (rail consolidations are generally accompanied by new agreements regarding labor protection).[138] New labor protection agreements would have been attractive to the parties for a number of reasons: the Ts would have avoided the delays and risks of strikes that labor opposition would have involved; the acquirers would have avoided these dangers, and would have been able to resolve questions relating to the integration of the Ts' lines and employees into their systems; and unions could have obtained protections for their members without risking the disastrous loss of jobs that would have accompanied a complete shutdown of the Ts' lines. New agreements would have eliminated the uncertainties surrounding existing legal rights of the parties, and would have facilitated the rail use sales, which offered substantial benefits for all parties involved. Despite their consensus regarding the likelihood of new labor protection agreements, the parties draw strikingly different pictures of the bargaining process. The GPs, pointing to what they see as an historical trend towards increased labor protection, argue that existing agreements such as the ELMPA and the WJPA would have provided a floor for benefits in the new agreements, 5 GP at S 179, and that labor's efforts to expand upon these existing protections would have been successful. The GPs conclude that the new agreements would have provided the "attrition protection" of the ELMPA for most of EL's employees, and the "Amtrak C-1"[139] protections for all other employees. 2 GPRB at S 75. The Ts respond that the possibility of a shutdown of their lines, and the consequent loss of jobs, would have placed labor in a *1268 precarious bargaining position. 2 ELF at 554. They contend that the inapplicability of the WJPA and the ELMPA would have prevented these agreements from providing a floor for the levels of benefits in the new agreements. Accordingly, the Ts postulate a widespread retreat by labor from historical levels of labor protection.[140] We find neither description of the labor protection bargaining process persuasive. The historical trend towards increased labor protection is not so one-sided as the GPs suggest, as the concessions by labor in the subsequent liquidations of the Rock Island and Milwaukee railroads indicate.[141] More fundamentally, the bargaining position of labor would have been weaker in the circumstances of the proffered transactions than was historically the case. The unprecedented possibility that — without federal intervention — most railway jobs in the Northeast would have been lost would have substantially restricted labor's bargaining power; the consummation of the proffered transactions would have been as important to the unions as it was to the Ts. Shannon (D) (Jan. 14, 1981) at 67-38, EL App. N at 375-76; Maxwell (S) (EL) (Oct. 15, 1980) at 4-5. As discussed above, the uncertainty surrounding the existing agreements would have made labor disinclined to resort to arbitration. An adverse decision would have been accompanied by a loss of labor's bargaining power, both because of the loss of a negotiating "chip" and the inflexibility of a bargaining position grounded primarily upon the self-destructive threat of strikes. The Ts' contention that labor would have "panicked" in the proffered transactions, Shannon (D) (Jan. 15, 1981) at 652, 30 GP Supp.App. at 179, is equally overstated. Labor's ability to demand arbitration of the existing labor protection agreements, and the threat of labor unrest and strikes, are inconsistent with the notion of a widespread retreat on labor protection issues. As the parties recognize, strikes would have "virtually foreclosed" rail use sales because of the disastrous effect that even relatively short shutdowns would have had on traffic patterns and earnings values. 2 GPRB at S 17-18; 1 RF at 35; Shannon (D) (Jan. 13, 1981) at 204-05, 30 GP Supp.App. at 75-76. Although extremely costly to labor as well as to the Ts, labor's ability to deny the Ts the economic benefits of the rail use sales at prices exceeding X by striking would have prevented any wholesale abandonment by labor of historical levels of labor protection. We conclude that the Ts and labor would have occupied roughly equal bargaining positions. Each possessed the ability to inflict serious costs on the other, yet only by accepting comparable costs themselves. Neither had ironclad cases under existing laws and agreements, yet both could have credibly threatened resort to litigation. All parties would have derived substantial benefits from consummation of the rail use sales. These considerations would have resulted, in all likelihood, in a relatively evenhanded compromise based roughly on historical levels of labor protection. *1269 Before discussing the precise terms of the labor protection agreements that we find the parties would have agreed to, we return to a point noted briefly at the beginning of this Part. The amount of labor protection benefits that unions could obtain would have been limited by the amount by which the proceeds of the rail use sales would have exceeded proceeds from sales for scrap.[142] Since, however, neither labor nor the Ts would know exactly what this excess would be, we do not think the bargaining would have proceeded with this particular figure in mind. Rather the parties would have reached what seemed a reasonable position in the light of the considerations we have discussed, and a T would then have had to decide, on the basis of admittedly imperfect data, whether to sign an agreement embodying that position or to break off the negotiations and abandon or sell at scrap value, which is the most that a buyer unprotected by an agreement would pay. Although numerous considerations, none of which is easily quantified, would have affected the labor protection negotiations,[143] the bargaining process we have outlined above permits a very rough prediction of the likely terms of the new labor protection agreements. As we explain below, this bargaining process would have settled on levels of labor protection roughly equivalent to the New Orleans conditions that the ICC historically has imposed under § 5(2) of the Interstate Commerce Act.[144] Stated briefly, the New Orleans benefits provide for 100 percent of an employee's earnings (for both displaced and dismissed employees) for up to four years following the ICC order approving a transaction plus any residual benefits under the five-year term of the WJPA. Alternatively, there is provision for a lump sum separation payment for dismissed employees at the employee's option. Both dismissal allowances and lump sum separation payments vary in accordance with the length of service of particular employees. Several considerations support our choice of New Orleans benefits as the approximate level of labor protection that the new agreements would have required. As the GPs note, this level of labor protection was imposed by the ICC on numerous occasions between 1950 and 1976, albeit principally in contexts other than complete abandonments.[145] Thus, all parties would have been familiar with the New Orleans conditions. This, and the perception of fairness that would have accompanied the choice of ICC-sanctioned levels of protection, would have made the New Orleans benefits a palatable compromise for all parties. Moreover, none of the parties would have felt that acceptance of this commonly used set of conditions would set an adverse precedent for future labor protection negotiations. Lastly, benefits based on the New Orleans decision would have represented a genuine compromise — providing labor with substantial protections while relieving the Ts of the potential burden of attrition protection. A similar bargaining process would have governed discussions on other terms of the labor protection agreements. The Ts (and the acquirers) would have been anxious to obtain broad power to transfer employees, both to permit efficient use of labor resources and to minimize labor protection costs. The GPs, relying in part upon existing agreements limiting EL's right to transfer employees, see agreements cited in 2 *1270 GPRB at S 14 n.29, and in part upon limitations contained in labor protection agreements negotiated in other liquidations, see Milwaukee Agreement cited in ELRB at 217 n.225, argue that the Ts would have been unable to obtain unrestricted transfer rights in the new agreements. We find the GPs' reasoning persuasive. Although we do not attempt to reach a precise determination of the issue, we conclude that management would have obtained transfer rights subject to payment of relocation costs and reasonable geographical restrictions. We also conclude that the new agreements would have contained a compromise regarding protection for non-union employees. Labor would have had considerably less incentive to obtain benefits for this group than for union members. Non-union employees, however, had historically received labor protection benefits as a result of ICC conditioning. See, e.g., Norfolk & Western Ry. and New York, Chicago & St. Louis R.R. — Merger, 330 I.C.C. 780, 825 modified, 331 I.C.C. 22, aff'd sub nom. Erie Lackawanna R.R. v. United States, 279 F.Supp. 316 (S.D.N.Y.1967), aff'd sub nom. Pennsylvania Central Merger and Norfolk & Western Inclusion Cases, 389 U.S. 486, 88 S.Ct. 602, 19 L.Ed.2d 723 (1968) (affirming decision by ICC in which Commission extended labor protection to non-union employees). Nonetheless, protection typically was not extended to supervisory, professional, and executive employees. We conclude that the agreement reached by the parties would have been roughly along the lines of anticipated ICC actions: nonunion operating employees would have been protected and a substantial number of non-union nonoperating employees would not have been protected. C. Calculation of Labor Protection Costs As noted at the outset of this Part, the labor protection costs (LPCs) that would have arisen from the new agreements must be deducted from the rail use sales values realizable by the Ts. We think this deduction would have taken the form of a reduction in the purchase price the acquirers would have paid for the Ts' lines and an assumption by the acquirers of the Ts' labor protection obligations. The Ts would have preferred this arrangement because, by relieving their estates of ongoing LPCs, a speedier distribution of sales proceeds would have been possible. Labor and the acquirers would have desired the arrangement because, by removing the Ts from the picture, it would have simplified their dealings with one another. This reduction would have been arrived at through bargaining between the Ts and the acquirers, which would have been based on the parties' calculations of projected LPCs. Two aspects of this bargaining process make it difficult for us to arrive at an exact estimate of the reduction in purchase prices that would have resulted from an acquirer's assumption of LPCs. First, as noted earlier, note 143, supra, any bargaining involves numerous unquantifiable variables. Second, projecting the costs of labor protection is extremely difficult: Total employee protection costs can be only rough estimates, since in many cases the actual costs will vary depending upon future events, the state of the economy and decisions made by others. I FSP at 164. See Ables, supra, at 117. Despite these difficulties we think we can arrive at reasonable estimates of the LPC reductions in the purchase prices for the Ts' properties. The first step in our analysis requires us to determine what an acquirer's management would have projected its LPC liability resulting from the proffered transaction to be. Arriving at an estimate of future LPCs demands consideration of numerous variables — the number of employees adversely affected by the proffered transaction, the average length of service and yearly compensation of a T's employees, the attrition rates applicable to the Ts' workforces, the rate of rehiring of dismissed employees, and so forth. There is little agreement among the parties on these issues or on a methodology for calculating LPCs. In the following sections we attempt to determine what assumptions regarding these variables the acquirers would have made and what methodology they would have employed in projecting LPCs. We recognize that any one of numerous methods of calculating LPCs could be adopted, but think that the conclusions *1271 we reach below approximate the projections that an acquirer would have made. Our task has been made somewhat easier in that, for reasons appearing elsewhere, we need make such projections only for EL, LV, and AA. The second step in our determination of the reduction in acquisition price attributable to LPCs involves consideration of the effect of bargaining between a T and acquirers on these LPC estimates. We address this issue in the section following our discussion of the acquirers' LPC estimates. 1. Erie Lackawanna (a) Operating Employees (i.) Number of Adversely Affected Employees. The first step that an acquirer would take in calculating LPCs would be estimating the number of employees who would be adversely affected by the proffered transactions. We discuss the LPCs attributable to EL's operating employees separately from those attributable to non-operating employees because of the differences in the methodologies chosen by the parties. We first consider the GPs' contentions regarding this issue. The GPs' LPC study was prepared by Richard Kasher.[146] Kasher's study was based on Conrail's Title V Data Base and a Conrail computer model designed, in part, to estimate LPCs. The Data Base contains information relating to subjects such as seniority, age, date of entrance into service, and prior earnings, for every employee of the Ts. Kasher (S) (Nov. 1, 1979) at 16-17. Using the Title V Data Base, Kasher estimated the number of adversely affected employees in each of the T's seniority districts, by assuming that the number of such employees would have been directly related to the track and facility abandonments. Kasher (S) (Nov. 1, 1979) at 12-14. Based on these assumptions, Kasher calculated that 16,626 PC union employees, 3,213 EL union employees, 1,346 R union employees, and 482 LV union employees would have been dismissed in the proffered transactions. Id. at 15. Kasher also estimated that 60 percent of the Ts' "non-agreement employees"[147] would have been terminated.[148] We begin our analysis of the LPCs attributable to EL's operating employees by rejecting Kasher's method of computing the numbers of operating employees that would have been adversely affected by the proffered transactions. Kasher's assumption that there would have been a direct correlation between the track mileage abandoned and the employees terminated is flawed in several respects. Kasher explicitly concedes that this assumption is inaccurate in many instances, and notes that had he had "more time and resources" he would have considered other factors in his analysis. Id. at 13-14. The failure to take into account considerations such as general manning requirements for the various acquisitions proffered by EL, proposed rehabilitation and capital improvement programs, and traffic retention data substantially reduces the value of Kasher's estimates of the number of employees adversely affected in the alternative scenario. Maxwell (S) (Oct. 15, 1980) at 6; Grotz (S) (Oct. 15, 1980) at 15-16. Moreover, Kasher's study completely ignores the labor expenses reflected in EL's earnings projections. In order for any degree of consistency to exist between EL's LPCs and its rail use sales values — which are based on the levels of employment reflected in the pro formas — EL's projected operating expenses must serve at least as a starting point for estimating reductions in its workforces.[149] *1272 Even if we were to accept the conceptual basis of Kasher's method of estimating the number of adversely affected EL employees, we would be forced to reject the results of his study because of serious flaws in its execution. First, it was premised on what would have happened to EL in the PC's "refined case",[150] and not on the dispositions proffered by EL. Although the PC refined case was similar to the EL-LV Chessie acquisition, there were substantial differences between it and the EL-LV Western acquisition, which is the relevant acquisition for our purposes.[151] Second, the personnel supervised by Kasher interpreted his instructions concerning estimating the number of adversely affected EL employees in a variety of inconsistent fashions,[152] and misunderstood important aspects of their assignments.[153] We are unpersuaded by the GPs' attempt to characterize these occurrences as "refinements" that "only enhanced the accuracy" of Kasher's study. 5 GP at S 209 n.44. The ad hoc adjustments to Kasher's methodology were undertaken by individuals with no experience in labor protection studies, and without knowledge of EL's traffic department.[154] We find the combination of conceptual flaws and mistakes in execution that marked Kasher's study to be so serious that, as regards its estimates of the number of adversely affected employees, it cannot provide even a starting point for our attempt to duplicate an acquirer's calculation of EL's LPCs. EL's estimates of the number of adversely affected employees are derived from Charles Shannon's LPC study.[155] Shannon (S) (Oct. 15, 1980). The methodology of the Shannon study differed in a number of respects from that of the Kasher study. Based on a comparison of lines 1-13 and 19 of the pro formas for the EL Western acquisition with EL's Annual Report to the ICC (R-1 report) for 1973, 32 GP Supp.App. at 537, Shannon calculated the change in operating expenses of the system that the proffered transaction would cause. Shannon (D) (Feb. 11, 1981) at 1063-76, EL App. N at 404-17. Relying on the labor portion of this reduction in EL's operating expenses Shannon calculated the number of jobs that would be eliminated as a result of the proffered *1273 transaction as well as the types and locations of jobs that would be lost. Id. Shannon viewed this figure as the number of employees potentially adversely affected by the proffered transaction. He reduced the figure by (1) the number of positions in the EL system that would be opened by attrition; and (2) the number of new positions that EL's capital improvement projects, normalized maintenance programs, and rehabilitation projects would have required. Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 1-6, EL App. N at 517-22. Shannon concluded that because of these factors, EL's employment needs would have almost completely offset the reduction in jobs reflected in EL's Western acquisition pro formas. Shannon testified that only two isolated groups of EL operating employees would have been both adversely affected by the proffered transactions and unable to find employment elsewhere on the EL system. In all four proffered transactions 68 marine employees in the New York Harbor would have been terminated. ELRB at 203. In the Chessie acquisitions a number of employees at the Marion Yard would have been dismissed. Aside from these two groups, however, Shannon concluded that no EL operating employees would have been eligible for labor protection benefits. The GPs claim Shannon's results to be suspect as a Dr. Pangloss approach. It seems almost impossible that EL could have managed so perfectly that attrition and new job opportunities arising from its rehabilitation and capital improvement programs would counterbalance the hundreds of jobs that would be lost as a result of abandonments. Beyond this we have some more detailed misgivings regarding certain aspects of Shannon's study. Shannon performed actual calculations of the number of adversely affected employees only for the EL Western acquisition. Although these figures can provide a foundation for our discussion, which principally concerns the EL-LV Western acquisition, we must adjust these conclusions to reflect the abandonment of EL's Scranton line in the EL-LV combination. Moreover, we note that Shannon's "study", as it appears in Exhibit 16 to Shannon's deposition, was submitted long after he had prepared his statement announcing his conclusions regarding EL's LPCs. Shannon (D) (Feb. 11, 1981) at 1188-94, 30 GP Supp.App. at 273A-273G. Despite these reservations we believe Shannon's LPC study provides a sufficient basis for calculation of EL's LPCs. As discussed above, Shannon's calculations posited a substantial reduction in the number of EL operating positions. Based on Shannon's workpapers, we calculate that Shannon concluded that 1,819 EL operating jobs would have been eliminated in the five years following the EL Western acquisition. Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 1-6, EL App. N at 517-23.[156] We accept the figures contained in Exhibit 16 as reasonable estimates of the number of operating positions that would have been eliminated as a result of the proffered transactions. Although Shannon did not specifically consider the effects on employment of each of the operating changes reflected in the EL pro formas, sheets 1-16 of Exhibit 16 indicate that Shannon did consider the major dislocations that the proffered transactions caused — the Marion and Croxton Yard projects, the abandonment of 700 miles of trackage, and the acquisitions of labor-saving *1274 equipment. Although we have some reservations regarding the apparent exclusion of the smaller operating changes from Shannon's study, we are willing to discount the effect of these changes. EL's witnesses testified that small, isolated employment reductions could be managed so as to minimize LPCs through the use of attrition and by bargaining with unions. Shannon (S) (Oct. 15, 1980) at 11-12. See also Maxwell (S) (Oct. 15, 1980) at 8-11. Since there is no particularized rebuttal by the GPs of Shannon's conclusions regarding the number of jobs that would have been lost in the proffered transactions, we find these estimates sufficiently accurate to serve as a base for our calculations. (ii.) Attrition and Rehire Rates. We take a different view, however, of Shannon's application of attrition rates and rehire rates to these figures. Despite his view that more than 1,800 operating positions would be eliminated in the proffered transactions, Shannon concluded that virtually no operating employees would have been adversely affected. As noted above, Shannon's determination was based principally on his use of attrition and rehire rates. Relying on EL's historical experience, Shannon calculated and applied an individual attrition rate for each of the six "ICC Groups"[157] that he determined would be affected by the proffered transactions. These rates ranged from 5.5 percent for dispatchers to 20.5 percent for watchmen. Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 1-6, EL App. N at 517-23. Exhibit 16 reflects two different treatments of the rehire rate issue. For ICC Group II, Shannon assumed — without explanation — a rehire rate of 50 percent. Shannon (D) (Feb. 11, 1981) Ex. 16 at sheet 1, line 16, EL App. N at 517. For all other EL operating employees (ICC Groups III, IV, V, VIa, and VIb) Shannon apparently assumed a 100 percent rehire rate. Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 2-6, EL App. N at 518-23. We base our calculations on a methodology somewhat different from that used by Shannon. Rather than applying the relevant individual attrition rates to particular ICC Groups, we apply a systemwide average attrition rate to EL's entire pool of operating employees. We adopt this approach principally to facilitate our computation of EL's labor protection costs. Our methodology, however, does EL no injustice: the attrition rate we use (10.4 percent) is simply the weighted average of the attrition rates applied by Shannon. Indeed, by applying attrition and rehire rates on a systemwide basis, rather than treating ICC Groups individually, we increase the extent to which job openings in one ICC Group can be filled by dismissed employees from other ICC Groups. Similarly, the GPs have no cause for complaint. Kasher used a systemwide attrition rate (10.8 percent) that was higher than that we apply here. Kasher (S) (Nov. 1, 1979) at 23-26. Although we thus accept the attrition rate applied by Shannon, we reject his assumptions regarding rehire rates. The record contains a startling variety of rehire rates: Shannon used both 100 percent and 50 percent; Kasher used 50 percent, which he felt was extremely conservative; PC, in a 1966 study, estimated 33 percent; and Grotz used both 100 percent and 20 percent.[158] We are impressed by the arguments advanced in support of the lower rehire rates. As Kasher commented, "Rail employees are not fungible", Kasher (S) (Nov. 1, 1979) at 26. In order to obtain personnel with the requisite skills for a particular position, those managing EL's system frequently would have had to look beyond the pool of surplussed EL employees. In addition, management's ability to rehire surplussed employees would have been limited by timing difficulties; many *1275 positions would have had to be filled quickly to maintain operating efficiency. Moreover, as discussed above, labor would have demanded reasonable restrictions on an acquirer's transfer rights, and the new labor protection agreements would have contained features such as geographical restrictions on transfer rights, as well as provisions regarding seniority, craft, and class limitations that would have restricted management's ability to fill job openings with surplussed employees, see pp. 1269-1270, supra. Kasher (S) (Nov. 1, 1979) at 26-27; Grotz (S) (Oct. 15, 1980) at 24-25. Lastly, the acquirers would have been inclined to take a relatively conservative view of the rehire rate, particularly in the context of negotiations with the Ts. Although we can do no more than provide a rough estimate, these considerations suggest to us an overall rehire rate for EL operating employees in the range of 33 percent. We apply this rate both to positions opened by attrition and to positions created by rehabilitation and capital improvement programs in order to determine the extent to which these positions could have been used to reduce the number of EL employees dismissed. As noted above, we accept Shannon's raw data regarding the number of jobs that would have been lost in each of the years following the proffered transactions. These figures, according to our calculations,[159] are as follows: Table 1 Number of Positions Eliminated Following Acquisition Year 1 318 Year 2 543 Year 3 539 Year 4 418 Year 5 1 ______ ___ Total 1819 These figures, reflecting reductions in the number of operating positions, must be adjusted to reflect the number of employees who either would have been rehired shortly after dismissal or would not have been terminated because of the creation of new jobs or vacancies in existing positions due to attrition. First, we apply the 10.4 percent attrition rate to the total number of EL operating employees in each of the years following the proffered transaction. Our computations indicate that the following number of operating positions would have been vacated each year due to attrition:[160] Table 2 Attrition-Created Positions Year 1 961 Year 2 992 Year 3 970 Year 4 921 Year 5 884 Based on Shannon's data, see Shannon (D) (Feb. 11, 1981), Ex. 16 at sheets 1-6, EL App. N at 517-23, which relied on EL's pro formas, we calculate that the following number of new positions would have been available each year as a result of EL's rehabilitation and capital improvement programs:[161] Table 3 New Positions Available Following Acquisition Year 1 616 Year 2 334 Year 3 67 Year 4 68 Year 5 69 Adding the two sets of figures in Tables 2 and 3 produces the following figures for the total number of annual employment openings *1276 in the EL system in the years following acquisition: Table 4 Vacant Positions Following Acquisition Year 1 1577 Year 2 1326 Year 3 1037 Year 4 989 Year 5 953 Application of the 33 percent rehire rate arrived at above to these totals produces the following figures for the number of operating positions opened by attrition or newly created that an acquirer's management would have been willing to consider available to reduce the number of EL operating employees that would be surplussed: Table 5 Rehires and Nondismissals 1577 × 33% = Year 1 520 1326 × 33% = Year 2 438 1037 × 33% = Year 3 342 989 × 33% = Year 4 326 953 × 33% = Year 5 315 Subtraction of these figures from the numbers of EL operating positions eliminated each year (Table 1) produces the number of EL operating employees who would have been eligible for labor protection benefits each year following the proffered transactions:[162] Table 6 Employees Dismissed Following Acquisition Year 1 (202)[*] Year 2 105 Year 3 197 Year 4 92 Year 5 (314)[*] (iii.) Lump Sum Separation Payments. Under the New Orleans benefits formulas surplussed employees are offered a choice between a monthly dismissal allowance and a lump sum separation payment, see p. 1269, supra. Calculation of EL's LPCs requires an estimate of the number of employees who would have chosen each of these types of benefits. Both Shannon and Grotz assumed 50 percent of EL's operating employees would have chosen lump sum payments. Their conclusions were based on the view that EL's employees would have been confident that they could find other employment and that, therefore, a lump sum payment would be preferable to a dismissal allowance that, in theory, would be reduced by outside earnings. Shannon (S) (Oct. 15, 1980) at 17-18; Grotz (S) (Oct. 15, 1980) at 27-28 & n.17. Kasher, in contrast, assumed that only 25 percent of EL's employees would have chosen a lump sum payment, because of the importance to EL employees of fringe benefits and the possibility of reemployment on the EL system with continued seniority rights that would have been lost by the choice of the dismissal allowance option. On the whole we find the Ts' 50 percent figure closer to the mark than the GPs' estimate.[163] Seniority rights would have *1277 been of comparatively little importance to less senior EL employees who, because dismissals would have been in reverse order of seniority, would have constituted the majority of surplussed employees. We feel compelled, however, to reduce the Ts' figure to 40 percent to reflect the reluctance of EL's employees to re-enter the job market without the protection of a monthly dismissal allowance. Kasher (S) (Nov. 1, 1979) at 17-18. Application of this figure to the total number of EL operating employees who would have been eligible for labor protection benefits in each of the years following the proffered transaction (Table 6) produces the following figures: Table 7 Dismissed Employees Electing Lump Sum Payments Year 1 0 Year 2 42 Year 3 79 Year 4 37 Calculation of the costs that would have resulted from lump sum payments to these employees requires consideration of several factors. First, we must estimate the average base period salary of the employees choosing lump sum payments. We arrive at this figure by first calculating a weighted average of the average yearly compensation rates of various groups of EL employees in 1973. According to our calculations, the average 1973 salary of EL operating employees was approximately $11,800. We adjust this figure to obtain figures for 1975 — the year preceding the date when we believe the proffered transaction would have occurred — by applying "growth factors" prepared by a PC witness, Richard Olsen, and relied upon in the Grotz LPC study. Grotz (S) (Oct. 15, 1980) at 25 & Table 6 n.5, citing Olsen (S) (Cost and Revenue) (Dec. 1, 1978) at 17 & Table B-8. These factors reflect the real increase of railway wages during the relevant period. Based on these figures we calculate that the average yearly compensation (in 1973 dollars) for EL employees in these years would have been: Table 8 Real Railway Wage Projections 1973 $11,800 1974 $11,800 × 1.024 = $12,083 1975 $11,800 × 1.0242 = $12,373 1976 $11,800 × 1.0243 = $12,670 1977 $11,800 × 1.0244 = $12,974 1978 $11,800 × 1.0245 = $13,286 1979 $11,800 × 1.0246 = $13,605 Next, we adopt crude adjustments to these figures to reflect the fact that most of the employees who would have been dismissed would have had less seniority, and therefore lower average annual compensation rates, than the average EL operating employee. In addition, we adjust these figures to reflect the fact that, because the proffered transactions would have taken place on April 1, 1976, the average base period compensation would have comprised eight months of 1975 wages and four months of 1976 wages: Table 9 Dismissed Employee Average Wage 1975 $12,373 × 95% = $11,754 1976 $12,670 × 95% = $12,037 1977 $12,974 × 95% = $12,325 1978 $13,286 × 95% = $12,622 1979 $13,604 × 95% = $12,924 Base Period Salaries Employees Dismissed in Period Starting: ____________________________________ 4/76 $11,754 × 67% = $ 7,875 $12,037 × 33% = $ 3,972 _______ $11,847 4/77 $12,037 × 67% = $ 8,065 $12,325 × 33% = $ 4,067 _______ $12,132 4/78 $12,325 × 67% = $ 8,258 $12,622 × 33% = $ 4,165 _______ $12,423 *1278 Base Period Salaries Employees Dismissed in Period Starting: ___________________________________ 4/79 $12,622 × 67% = $ 8,457 $12,924 × 33% = $ 4,265 _______ $12,722 The New Orleans benefits provided for varying levels of lump sum separation payments depending upon the length of an employee's service with a carrier. The applicable figures are as follows: Table 10 Length of Service Benefit Level 1-2 years 3 months pay 2-3 years 6 months pay 3-5 years 9 months pay 5 + years 12 months pay See WJPA, Ex. GPS — Labor — 10 at 176. Accordingly, we must estimate the number of employees choosing lump sum payments who would have fallen into each of these categories. The record contains very little evidence on this issue. Grotz adopts certain assumptions regarding the LV workforce, which may be taken to be roughly applicable to EL. Grotz's figures place the vast majority of employees choosing lump sum payments in the 5 + years of service category, a result we accept principally because the acquirers would have been inclined to insist on a conservative projection. Grotz (S) (Oct. 15, 1980) at Table 4, 5.[164] Application of these figures to the New Orleans benefit formulas and to the average base period compensation rates set out above produces the following cost of lump sum payments in the EL Western acquisition: Table 11 Lump Sum Payments[*] Year 1 0 Year 2 1 × .25 × $12,132 = $ 3,033 2 × .5 × $12,132 = $ 12,132 3 × .75 × $12,132 = $ 27,297 36 × 1 × $12,132 = $436,752 ____________________________________ Total $479,214 ________ Year 3 1 × .25 × $12,423 = $ 3,106 4 × .5 × $12,423 = $ 24,846 6 × .75 × $12,423 = $ 55,904 68 × 1 × $12,423 = $844,764 ______________________________________ Total $928,620 ___________ Year 4 1 × .25 × $12,722 = $ 3,181 2 × .5 × $12,722 = $ 12,722 2 × .75 × $12,722 = $ 19,083 32 × 1 × $12,722 = $407,104 ______________________________________ Total $442,090 __________ Grand Total $1,849,924 (iv.) Dismissal Allowances. The next step in our calculation of EL's LPCs requires computation of the cost of dismissal payments made to surplussed EL operating employees. As suggested above (Tables 6 & 7), the following numbers of EL operating employees would have chosen to receive dismissal allowances in the years following the proffered disposition: Table 12 Dismissed Employees Electing Dismissal Allowances Year 2 63 Year 3 118 Year 4 55 We must adjust these figures to reflect the effect of rehiring surplussed employees into positions opened by attrition or created by EL's rehabilitation and capital improvement programs (Table 5): *1279 Table 13 Employees Receiving Monthly Dismissal Allowances -------------------------------------------------------------- Year 1 2[*] 3[*] 4[*] 5[**] 6 -------------------------------------------------------------- Employees 1 0 0 0 0 0 0 Dismissed 2 - 63 63 63 0 0 in Year: 3 - - 118 118 0 0 4 - - - 55 0 0 -------------------------------------------------------------- The New Orleans benefits provide for dismissal allowances of 100 percent of an employee's base period compensation, calculated at pp. 1277-1278, supra, minus any outside earnings or unemployment compensation received by an employee. The record contains two different conclusions with respect to this latter issue. Kasher, on the assumption that employees receiving dismissal allowances would not report their outside earnings, did not make any adjustment in the cost of dismissal allowances to reflect outside earnings. Kasher (S) (Nov. 1, 1979) at 21. Grotz adopted a figure based upon an assumption that each employee dismissed would find employment three months following dismissal at a salary equivalent to 50 percent of pre-dismissal earnings increased to reflect general wage escalations. Grotz justified this assumption on the grounds that employees would not have "flouted the express direction of the ICC" and on the fact that an acquirer would have established a "prudent procedure" for ensuring that outside earnings were reported. Grotz (S) (Oct. 15, 1980) at 28-29 & Table 9. We see no reason why an acquirer would not have considered itself able to achieve at least partial success in policing the reporting requirements of the new agreement. Based roughly on Grotz's calculations (with a crude downward adjustment to reflect enforcement costs and other difficulties), we adopt a 50 percent offset to monthly dismissal allowances to account for outside earnings. Using the figures set out above regarding the average base period compensation of the employees choosing dismissal allowances (Table 9) and the figures relating to the numbers of EL operating employees receiving dismissal allowances (Table 13), we arrive at the following conclusions regarding the cost of monthly dismissal payments: Table 14 Costs of Monthly Dismissal Allowances Year 1 0 -------------------------------------- Year 2 63 × $12,132 = $ 764,316 -------------------------------------- Year 3 63 × $12,132 = $ 764,316 118 × $12,423 = $1,465,914 ____________________________ Year 3 Total $2,230,230 -------------------------------------- Year 4 63 × $12,132 = $ 764,316 118 × $12,423 = $1,465,914 55 × $12,722 = $ 699,710 ____________________________ Year 4 Total $2,929,940 -------------------------------------- Year 5 0 -------------------------------------- Grand Total $5,924,486 × 50% __________ Total Cost of Monthly Dismissal Payments $2,962,243 (v.) Fringe Benefits. The New Orleans benefits also guaranteed employees electing *1280 to receive monthly dismissal allowances fringe benefits, such as health and welfare benefits. Both Grotz and Kasher estimated that this item would result in LPCs equal to approximately 17 percent of the cost of dismissal payments, before reduction for outside earnings. Kasher (S) (Nov. 15, 1979) at 22-23; Grotz (S) (Oct. 15, 1980) at 29. Shannon apparently adopted a similar figure in the context of operating employees. We find Grotz's and Kasher's estimates — which are based on historical data compiled by the Railroad Retirement Board — reasonable and accept them for purposes of our computations. An acquirer likely would have calculated that total costs incurred as a result of fringe benefit payments to EL operating employees would have been: Table 15 Cost of Fringe Benefits 17% × $5,924,486 = $1,007,163 (vi.) Relocation Costs. There is general agreement among the various LPC studies regarding the cost of relocating an individual employee. Both Shannon and Kasher estimated relocation expenses of $4500 per relocated employee. 2 ELRB at 217 & n.226; Kasher (S) (Nov. 1, 1979) at 28. Grotz estimated that it would require $600,000 in moving expenses to relocate 120 LV employees, an average cost of $5000 per employee. Grotz (S) (Oct. 15, 1980) at 32. USRA, in the FSP, estimated $5000 in moving costs per relocated employee. I FSP at 165. We adopt the $5000 figure, because the Ts would have been required to pay both moving and other relocation expenses, thus requiring slight upward adjustments of the parties' figures. There is very little evidence in the record concerning the number of EL employees who would have been eligible for reimbursement of relocation expenses. Shannon provides rough estimates for operating employees at the Marion shop and the marine workers at the New York Harbor, but does not address the issue as it relates to the EL system generally. Kasher, on the basis of his experience at Conrail, estimated that approximately 50 percent of all rehired employees would receive relocation benefits. Kasher (S) (Nov. 1, 1979) at 27. Although we are reluctant to apply this figure without some indication of the relevance of Kasher's Conrail experience to EL's case, the dearth of evidence on the subject in the record leaves us little choice. As reflected in Tables 1, 5, and 13 the following numbers of employees would have been rehired (or transferred prior to termination) in the years following the proffered transactions: Table 16 Rehired Employees Year 1 318 Year 2 438 Year 3 342 Year 4 326 Year 5 236 ______ ______ Total 1660 Application of the 50 percent rate (representing rehired employees eligible for relocation benefits) and the $5000 average cost per employee produces the following total cost of relocation expenses: Table 17 Relocation Costs 1660 × 50% × $5000 = $4,150,000 (vii.) Displacement Costs. The New Orleans conditions provide for payment of displacement expenses to employees who accept employment at salaries less than their base period compensation. As Kasher conceded, displacement allowances would have decreased rapidly in the years following the proffered transactions, principally because the earnings gap between base period compensation and yearly salaries would be eroded by general wage increases. Kasher (S) (Nov. 1, 1979) at 30. Kasher estimated that 12 percent of the surplussed employees rehired in the first year following acquisition would receive displacement allowances. His statement does not reveal his assumptions regarding later years, although in light of his general views that displacement allowances would have been concentrated in the period immediately following the proffered transactions, we feel safe in assuming that after year 1 this item may be disregarded as de minimis. Shannon apparently did not address this question in the context *1281 of EL's operating employees. We adopt Kasher's 12 percent figure as a reasonable estimate of the number of employees that would receive displacement allowances. We reject, however, Kasher's estimate of the average monthly displacement payment that EL operating employees would receive. Kasher used a figure ($2,717.04 per year) derived from the "average monthly displacement allowance claim" submitted to Conrail in 1978. As the Grotz study points out, Grotz (S) (Oct. 15, 1980) at 30, this figure is far larger than the average amount of displacement payments actually made by PC in 1973 and 1974: $76.28 per year. See I FSP at 165. Moreover, the Conrail figures relate to claims made under § 505(b) of the Rail Act, which — unlike the New Orleans benefits — adjusted an employee's base period compensation to reflect subsequent general wage increases. We find the figures relating to actual payments made by PC far more relevant to our inquiry than the claims submitted to Conrail, and accept the $76.28 figure. Indeed, because the PC payments were made primarily pursuant to agreements that adjusted base period salaries to reflect general wage increases, this figure is probably somewhat overstated. We accept it, however, on the theory that the acquirers would have made a conservative estimate of their potential LPC liability. Based on the total numbers of operating employees in EL's workforce following acquisition (Table 2 & note 160, supra), we calculate that LPCs attributable to displacement allowances would have totalled: Table 18 Displacement Costs 9232 × 12% = 1108 displaced employees 1108 × $76.28 per claim = $84,518 (viii.) LPCs Attributable to Abandonment of EL's Scranton Line. As noted previously, Shannon's study addressed only the EL Western acquisition. Shannon testified, however, that the EL-LV Western acquisition would involve identical LPCs. Although the EL-LV Western acquisition contemplated abandonment of the Scranton line, which would continue in rail use in the EL Western acquisition, Shannon concluded that all employees on this line would find employment on the nearby, parallel LV line. Shannon (D) (Jan. 15, 1981) at 724-26, EL App. N at 395-97. Positions on that line would have been available, according to Shannon, because of increased normalized maintenance and new rehabilitation expenses. Shannon (S) (Oct. 15, 1980) at 64; Adik (S) (Dec. 31, 1978) at 13-15. Although the GPs reject this contention, 2 GPRB at S-21, 26-27, they offer no particularized criticism of Shannon's assumption. Absent such criticism, we accept Shannon's estimate, although we think it likely that an acquirer would have determined that some relocation expenses would have been involved, particularly in the MOWE department. Based on our rough calculation of the number of employees who would have been affected and the average cost of relocating an employee, we estimate that the acquirers would have projected a cost of approximately $450,000 for this item. (ix.) Marine Employees. Shannon determined, ELRB at 203, that in all four proffered transactions 68 marine employees working in the New York Harbor would have been terminated — a conclusion we accept. Shannon assumed that all marine employees would have chosen a lump sum payment rather than a monthly dismissal allowance, principally because of the existing employment opportunities with other operators in the New York Harbor area. Shannon (D) (Jan. 14, 1981) at 372-75, EL App. N at 352-53. We see no basis for questioning this conclusion, which was based on Shannon's personal knowledge of and experience in the New York Harbor. Shannon (D) (Feb. 11, 1981) at 1136-37, EL App. N at 456-57. Shannon applied a 12 percent attrition rate to the 68 marine employees for the one year period preceding the acquisition, and concluded that only 60 employees would be *1282 employed in the harbor at the date of the acquisition. We adjust this figure somewhat to reflect our finding that the 12 percent attrition rate is too high, adopting instead a 6 percent rate. Shannon conceded that such a rate would have been appropriate, and we think that an acquirer's management would have opted for the more conservative number. Shannon (D) (Jan. 14, 1981) at 366; EL App. N at 350. Accordingly, we conclude that 64 marine employees would have received lump sum severance payments. The average yearly wage in 1973 for these employees was $11,580. Id. Adjusting this figure to reflect the increase in real railway wages, we obtain an average 1975 base period compensation of approximately $12,145. See Table 8. Assuming, as we did in the context of operating employees in general, see pp. 1277-1278, supra, that LV's average length of service distribution would apply, and using the WJPA formula (Table 10), we conclude that the total lump sum payments that EL's marine employees would have received would have been: Table 19 Marine Employees 1 × .25 × $12,145 = $ 3,036 3 × .5 × $12,145 = $ 18,218 5 × .75 × $12,145 = $ 45,544 55 × 1.0 × $12,145 = $ 667,975 ______________________________ Total $ 734,773 (x.) Total LPCs Attributable to EL Operating Employees. Addition of the LPCs attributable to lump sum payments (Table 11), dismissal allowances (Table 14), fringe benefits (Table 15), relocation expenses (Table 17), displacement allowances (Table 18), EL's marine employees (Table 19), and abandonment of EL's Scranton line provides the following estimate of LPCs attributable to EL operating employees: Table 20 Total LPCs Attributable to EL Operating Employees Lump Sum $ 1,849,924 Dismissal $ 2,962,243 Fringe Benefits $ 1,007,163 Relocation $ 4,150,000 Displacement $ 84,518 Marine Employees $ 734,773 Scranton Line $ 450,000 ____________ ___________ Total $ 11,238,621 (b) Nonoperating Employees We base our calculations of the LPCs attributable to EL's nonoperating employees on the Shannon study, principally because the GPs' study did not specifically address this precise issue. Applying the methodology discussed above in the context of EL's operating employees, Shannon concluded that approximately 213 EL union nonoperating employees would have been terminated, 125 would have been relocated, and 310 EL nonunion nonoperating employees would have been terminated. Shannon (S) (Oct. 15, 1980) at 4 & 22. Based on an assumed level of benefits for union employees roughly equivalent to the New Orleans benefits, Shannon estimated that the LPCs resulting from these terminations would total $4.1 million. As in the case of operating employees, we must adjust Shannon's conclusions to reflect our disagreement with his assumed attrition and rehire rates. Based on Shannon's workpapers, 999020-999043, EL App. N at 628-39, and testimony, Shannon (S) (Oct. 15, 1980) at 21-22, we estimate that the following numbers of EL nonoperating positions would have been eliminated in the years following acquisition:[165] *1283 Table 21 Positions Eliminated Following Acquisition MOW MOE Trans. Traff. Gen. Total Year 1 19 14 31 136 227 427 Year 2 19 14 31 136 228 428 Year 3 19 19 15 0 0 53 Year 4 6 0 0 0 0 6 Year 5 0 0 0 0 0 0 In determining the number of protected employees who would have been terminated, we must take into account the extent to which the above figures reflect terminations of unprotected employees. As discussed above, see p. 1270, supra, nonunion employees occupying executive, professional, and supervisory positions would have been unprotected under the new agreements. Our examination of the "Erie Lackawanna Railway Company, Black Book", Shannon (D) (Feb. 12, 1981) Ex. 26, EL App. N at 583, convinces us that the figure used by Shannon for nonagreement employees is a fair estimate of the number of surplussed employees who would have been unprotected under the new agreements. Most noncontract employees fall within the categories noted above — and would have been unprotected under the new agreements. We reduce Shannon's figures only slightly, from 310 to 280, to reflect our rough estimates of the number of employees — such as secretarial and clerical employees — that likely would have been entitled to protection, see id. Before applying attrition and rehire rates, we deduct the number of unprotected nonoperating employees from the total number of nonoperating positions eliminated as a result of the proffered transactions (Table 21) to obtain the number of protected nonoperating positions that would be eliminated in the proffered transactions. For each year following acquisition, we deduct unprotected employees in numbers roughly proportional to the total numbers of nonoperating positions eliminated in each year:[166] Table 22 Protected Positions Eliminated Year 1 427 - 126 = 301 Year 2 428 - 126 = 302 Year 3 53 - 28 = 25 Year 4 6 - 0 = 6 As with operating employees, attrition and rehire rates must be applied to these figures to obtain an estimate of the number of employees who would have had to be surplussed to achieve the savings reflected in EL's pro formas. We accept as a reasonable estimate of the attrition rate for EL nonoperating employees the 13 percent figure used by Shannon. This figure was based on EL's historical attrition rate among nonoperating employees, Shannon (D) (Feb. 11, 1981) Ex. 16 at sheet 36, line 6, EL App. N at 552, and the GP's offered no particularized criticism of the rate. Shannon used a rehire rate of approximately 50 percent, see Shannon workpapers 999021, 999024, 999028, EL App. N at 629, 632, 636. We reject this figure, and adopt instead a rehire rate of 33 percent, for the reasons already discussed. Our application of attrition and rehire rates to the numbers of EL nonoperating employees potentially adversely affected by the acquisition (Table 22) follows:[167] *1284 Table 23 Positions Opened By Attrition Year 1 1194 × 13% = 155 + 75 = 230[*] Year 2 767 × 13% = 100 Year 3 437 × 13% = 57 Year 4 412 × 13% = 54 Year 5 406 × 13% = 53 Year 6 406 × 13% = 53 Year 7 406 × 13% = 53 Table 24 Rehires and Nondismissals Year 1 230 × 33% = 76 Year 2 100 × 33% = 33 Year 3 57 × 33% = 19 Year 4 54 × 33% = 18 Year 5 53 × 33% = 18 Year 6 53 × 33% = 18 Year 7 53 × 33% = 18 Deducting the figures in Table 24 from the numbers of potentially adversely affected protected EL nonoperating employees (Table 22) provides the number of protected EL nonoperating employees who would have been eligible for labor protection: Table 25 Dismissed Protected Employees Year 1 301 - 76 = 225 Year 2 302 - 33 = 269 Year 3 25 - 19 = 6 Year 4 6 - 18 = (12) We must next determine which of the eligible employees would have elected to receive a lump sum severance payment rather than monthly dismissal allowances. We adopt the same assumptions here that we used in the context of operating employee LPCs, see pp. 1276-1277, supra. Table 26 Employees Electing Lump Sum Severance Payments Year 1 225 × 40% = 90 Year 2 269 × 40% = 108 Year 3 6 × 40% = 2 The figures in Table 26 reflect the number of EL nonoperating employees that would choose to receive a lump sum payment. As in the context of operating employees, we must consider the WJPA benefit formula (Table 8) and our estimate of the distribution of EL nonoperating employees by length of service to calculate the LPCs attributable to lump sum payments made to EL's nonoperating employees:[168] Table 27 Cost of Lump Sum Payments Year 1 2 × .25 × $12,419 = $ 6,210 4 × . 5 × $12,419 = $ 24,838 6 × .75 × $12,419 = $ 55,886 78 × 1.0 × $12,419 = $ 968,682 __________________________________________________ Total $1,055,616 _____ __________ Year 2 2 × .25 × $12,719 = $ 6,360 5 × .5 × $12,719 = $ 31,798 8 × .75 × $12,719 = $ 76,314 93 × 1.0 × $12,719 = $1,182,867 __________________________________________________ Total $1,297,339 _____ __________ Year 3 2 × 1.0 × $13,022 = $ 26,044 __________ Grand Total $2,379,000 ___________ __________ *1285 It remains to calculate the LPCs attributable to paying monthly dismissal allowances to EL nonoperating employees: Table 28 Employees Receiving Dismissal Allowances Per Year Year: 1 2 3 4 5 6 7 Employees 1 135 135 135 123[*] 105[*] -[**] - Dismissed 2 - 161 161 161 161 143[*] -[**] in Year: 3 - - 4 4 4 4[*] -[*] As in the case of EL's operating employees, we reduce the dismissal allowances paid to employees by 50 percent, which we think reflects the offset that may be attributed to outside earnings. We also reduce the fifth year of dismissal allowances paid to reflect the terms of the WJPA. Table 29 Cost of Dismissal Allowances Year 1 135 × $12,419 = $ 1,676,565 ----------------------------------------------- Year 2 135 × $12,419 = $ 1,676,565 161 × $12,719 = $ 2,047,759 ----------------------------------------------- Year 3 135 × $12,419 = $ 1,676,565 161 × $12,719 = $ 2,047,759 4 × $13,022 = $ 52,088 ----------------------------------------------- Year 4 123 × $12,419 = $ 1,527,537 161 × $12,719 = $ 2,047,759 4 × $13,022 = $ 52,088 ----------------------------------------------- Year 5[*] 105 × 60% × $12,419 = $ 782,397 161 × 60% × $12,719 = $ 1,228,655 4 × 60% × $13,022 = $ 31,253 ----------------------------------------------- Year 6[*] 143 × 60% × $12,719 = $ 1,091,290 4 × 60% × $13,022 = $ 31,253 ----------------------------------------------- Grand Total $ 15,969,533 50% Total Cost of Monthly ___________ Dismissal Allowances $ 7,984,767 (i.) Fringe Benefits. As in the context of operating employees we adopt a 17 percent figure as a reasonable estimate of the LPCs attributable to the payment of fringe benefits for EL nonoperating employees receiving dismissal allowances: Table 30 Cost of Fringe Benefits 17% × $15,969,533 = $2,714,820 (ii.) Relocation Costs. Application of a $5000 average relocation cost per relocated employee and a 50 percent rate of relocation among rehired employees provides the following estimates of relocation costs:[169] Table 31 Number of Rehires per Year Year 1 76 Year 2 33 Year 3 19 Year 4 18 Year 5 18 Year 6 18 Year 7 4 ______ __ Total 186 186 × 50% = 93 employees relocated 93 × $5,000 = $465,000 relocation costs (iii.) Displacement Costs. Based on the assumptions discussed above, we calculate that LPCs attributable to the payment of displacement allowances to EL's nonoperating employees would have totalled approximately: Table 32 Displacement Costs 1194 × 12% = 143 displaced employees 143 × $76.28 per displaced employee = $10,900 (iv.) Total LPCs Attributable to EL's Nonoperating Employees. Addition of the LPCs attributable to lump sum payments (Table 27), monthly dismissal allowances (Table 29), fringe benefits (Table 30), relocation expenses (Table 31), and displacement expenses (Table 32) provides the following figure for total LPCs attributable to EL's nonoperating employees: *1286 Table 33 Total LPCs Attributable to EL Nonoperating Employees Lump Sum $ 2,379,000 Dismissal $ 7,984,767 Fringe Benefits $ 2,714,820 Relocations $ 465,000 Displacements $ 10,900 _____________ ___________ Total $13,554,487 Addition of the LPCs for operating and nonoperating EL employees produces an estimate of total EL LPCs of $24,793,108. 2. Lehigh Valley We base our calculations of LV's LPCs on a study prepared by W. Arthur Grotz,[170] which we adjust to reflect our disagreement with various of the author's assumptions and choices of methodology. For the reasons discussed previously, see pp. 1271-1272, supra, we reject the GPs' study. The starting point for Grotz's analysis was a manning study estimating the number of employees that would have been required to operate the LV system following acquisition. Grotz (S) (Oct. 15, 1980) at 15-16. Grotz estimated that 2,161 employees would be needed to run the LV system immediately following acquisition, id. at Table 2, and that 2,098 of these positions would be available for protected LV employees, id. Grotz concluded that 389 protected LV employees would be terminated as of the date of the acquisition, that 91 employees aged 64 or older would be terminated, and that 120 unprotected LV employees would be terminated, id. at Table 1. Grotz estimated that an additional 116 employees would be terminated in the years following the acquisition. We accept Grotz's conclusions regarding the number of employees that would have been terminated on "Day 1", the date of the proffered transaction. Grotz's conclusions were based on estimates of LV's future employment requirements — which we find to be reasonable. Grotz had considerable experience in the field of railway economics and was particularly well-informed regarding LV's system. Moreover, his study considered the effects of acquisition on each of the various categories of LV's workforce. We note also that Grotz's estimates of the numbers of adversely affected LV employees were larger than those of Kasher.[171] Despite our agreement with Grotz's conclusions regarding Day 1 dismissals, we do not think an acquirer's management would have treated dismissals after Day 1 in the same fashion that Grotz's study did. In addition to the elimination of employment positions on Day 1, Grotz projected substantial reductions in the size of LV's workforce in the years following acquisition. Grotz concluded, however, that virtually no employees would be adversely affected as a result of these reductions in available positions on the LV system. Rather, Grotz implicitly assumed that all positions vacated through attrition in a given year could be permitted to remain open throughout that year in order to meet the reductions in employment projected for the next year, id. at Table 3. Thus, the number of positions *1287 opened by attrition was not reduced to reflect the fact that not all such positions could be filled with employees otherwise scheduled for dismissal, id. We think an acquirer would have adopted a less optimistic methodology in projecting its LPC liability under the new agreements. An acquirer's management would have recognized the serious obstacles to using attrition to reduce LPCs. In the words of the ICC, "practical limitations [such] as geography or personal qualifications would restrict the universal transfer and use of employees". Pennsylvania R.R. — Merger — New York Central R.R., 327 I.C.C. 475, 689 (1966). A leading commentator in the field has observed that there is "no certainty that the skills or geographic locations of the employees who were expected to separate from the railroad would match the skills of the employees actually affected by the consolidation". Ables, supra, at 118. Likewise, discussing an almost identical issue, the Supreme Court has written: According to the findings of the hearing examiner in this case, 863 employees will be totally deprived of employment during the five-year period following the merger. Appellants argue that there is no need for these discharges since natural attrition will open up many more than 863 jobs during the same period. However, as the railroads point out, attrition does not work in a uniform or predictable manner and there is no indication that the elimination of surplus posts can be accomplished by the method appellants suggest .... Maintenance of Way Employees v. United States, 366 U.S. 169, 178 n.11, 81 S.Ct. 913, 918 n.11, 6 L.Ed.2d 206 (1961). We think these considerations, and a desire to avoid underestimating LPCs, would have prompted an acquirer's management to adopt a methodological approach to attrition similar to that we adopted earlier in calculating EL's LPCs. Only a portion of the positions vacated each year due to attrition would be considered available to reduce the number of dismissals projected for the following year: thus, a rehire or nondismissal rate would be applied to attrition-vacated positions to determine how many of these openings realistically could have been used to reduce the number of LV employees who would be terminated.[172] We accept Grotz's use of a 12.53 percent attrition rate. This figure was derived by taking a weighted average of the historical rates of attrition for particular employee age groups that then were applied to the LV workforce, Grotz (S) (Oct. 15, 1980) at 23 & Table 4, and we see no basis for questioning Grotz's determination. Although we applied a 33 percent rehire rate in calculating EL's LPCs, we adopt a 40 percent rate in LV's case. We think this, or an equivalent, adjustment would have been adopted by an acquirer to reflect the lesser impact of geographical limitations attributable to the smaller size of LV's system. Based on the assumptions and choices of methodology set forth above, we think an acquirer's calculations of the LPCs attributable to the LV system would have proceeded roughly along the following lines. Based on Grotz's Tables 2 and 3, we estimate that the following numbers of protected positions would have been eliminated in the years following the proffered acquisition:[173] *1288 Table 34 Protected Positions Eliminated Following Acquisition 4/76 389 10/76 112 10/77 206 10/78 133 10/79 100 10/80 0 These figures must be adjusted to obtain an estimate of the number of employees who actually would have been surplussed to achieve Grotz's forecast employment reductions: Table 35 Positions Vacated Through Attrition 4/76 2161 × 12.53% = 271[*] 10/76 2015 × 12.53% = 252 10/77 1809 × 12.53% = 227 10/78 1676 × 12.53% = 210 10/79 1576 × 12.53% = 197 10/80 1576 × 12.53% = 197 10/81 1576 × 12.53% = 197 Table 36 Rehires and Nondismissals 4/76 271 × 40% = 108 10/76 252 × 40% = 101 10/77 227 × 40% = 91 10/78 210 × 40% = 84 10/79 197 × 40% = 79 10/80 197 × 40% = 79 10/81 197 × 40% = 79 Deducting the figures in Table 36 from the reductions projected in LV's available employment positions (Table 34) provides the number of protected LV employees who would be eligible for labor protection benefits under the new agreements: Table 37 Protected Employees Dismissed 4/76 389 - 108 = 281 10/76 112 - 101 = 11 10/77 206 - 91 = 115 10/78 133 - 84 = 49 10/79 100 - 79 = 21 We must next determine which of the employees eligible for labor protection (Table 37) would have elected to receive a lump sum separation payment rather than a monthly dismissal allowance. We reject Grotz's use of a 50 percent figure representing the proportion of employees choosing lump sum separation payments, and adopt instead the 40 percent figure used in calculating EL's LPCs, see pp. 1276-1277, supra. Table 38 Employees Electing Lump Sum Separation Payments 4/76 281 × 40% = 112 10/76 11 × 40% = 4 10/77 115 × 40% = 46 10/78 49 × 40% = 20 10/79 21 × 40% = 8 We next consider the WJPA benefit formula (Table 10) and the distribution of LV employees by length of service, see Grotz (S) (Oct. 15, 1980) at Table 5, 8, to arrive at a figure for the LPCs attributable to lump sum payments to LV's employees:[174] Table 39 Lump Sum Payments 4/76 .25 × 2 × $11,847 = $5,924 . 5 × 5 × $11,847 = $29,618 .75 × 8 × $11,847 = $71,082 1 × 97 × $11,847 = $1,149,159 *1289 Table 39 Lump Sum Payments 10/76 .75 × 1 × $11,988 = $8,991 1 × 3 × $11,988 = $35,964 ---------------------------------------------------- 10/77 .25 × 1 × $12,276 = $3,069 .5 × 2 × $12,276 = $12,276 .75 × 3 × $12,276 = $27,621 1 × 40 × $12,276 = $491,040 ---------------------------------------------------- 10/78 .5 × 1 × $12,571 = $6,286 .75 × 1 × $12,571 = $9,428 1.0 × 18 × $12,571 = $226,278 ---------------------------------------------------- 10/79 .75 × 1 × $12,874 = $9,656 1.0 × 7 × $12,874 = $90,118 ____________________________________________________ Total $2,176,510 The next step in our calculation of LV's LPCs requires computation of the cost of monthly dismissal allowances. We base our calculations upon the numbers of employees electing monthly dismissal allowances (derived from Tables 37 & 38) and the numbers of rehires per year (Table 36): Table 40 Employees Receiving Monthly Dismissal Allowances 4/76 10/76 10/77 10/78 10/79 4/80 10/80 4/81 10/81 ---------------------------------------------------------------------- 1 169 169 169 169 169 169 169 169 0[*] 2 - 7 7 7 7 7 0[**] 0 0 3 - - 69 69 69 69 0[**] 0 0 4 - - - 29 29 29 26[**] 26 0[**] 5 - - - - 13 13 13 13 0[**] Based on these figures, we calculate the LPCs attributable to payment of monthly dismissal allowances. Table 41 Monthly Dismissal Allowances 4/76-10/76 169 × $11,847 × 50% = $1,001,072 ----------------------------------------------------------------- 10/76-10/77 169 × $11,847 = $2,002,143 7 × $11,988 = $ 83,916 _________________________________________________________________ Total $2,086,059 ----------------------------------------------------------------- 10/77-10/78 169 × $11,847 = $2,002,143 7 × $11,988 = $ 83,916 69 × $12,276 = $ 847,044 _________________________________________________________________ Total $2,933,103 ----------------------------------------------------------------- 10/78-10/79 169 × $11,847 = $2,002,143 7 × $11,988 = $ 83,916 69 × $12,276 = $ 847,044 29 × $12,571 = $ 364,559 _________________________________________________________________ Total $3,297,662 ----------------------------------------------------------------- 10/79-4/80 169 × $11,847 × 50% = $1,001,072 7 × $11,988 × 50% = $ 41,958 69 × $12,276 × 50% = $ 423,522 29 × $12,571 × 50% = $ 182,280 13 × $12,874 × 50% = $ 83,681 _________________________________________________________________ Total $1,732,513 ----------------------------------------------------------------- 4/80-10/80 169 × $11,847 × 50% × 60% = $ 600,643 7 × $11,988 × 50% × 60% = $ 25,175 69 × $12,276 × 50% × 60% = $ 254,113 29 × $12,571 × 50% × 60% = $ 109,368 13 × $12,874 × 50% × 60% = $ 50,209 _________________________________________________________________ Total $1,039,508 ----------------------------------------------------------------- 10/80-4/81 169 × $11,847 × 50% × 60% = $ 600,643 26 × $12,571 × 50% × 60% = $ 98,054 13 × $12,874 × 50% × 60% = $ 50,209 _________________________________________________________________ Total $ 748,906 ----------------------------------------------------------------- 4/81-10/81 26 × $12,571 × 50% × 60% = $ 98,054 13 × $12,874 × 50% × 60% = $ 50,209 _________________________________________________________________ Total $ 148,263 ----------------------------------------------------------------- Grand Total $12,987,086 ----------------------------------------------------------------- Outside Earnings × 50% ___________ Total Dismissal Allowances $6,493,543 In calculating fringe benefits we apply the 17 percent figure used in computing EL's LPCs and applied by Grotz in his study for LV: Table 42 Fringe Benefits 17% × $12,987,086 = $2,207,805 Rather than assuming, as we did in EL's context, that 50 percent of all employees *1290 rehired would incur relocation expenses, we estimate that only 25 percent of the rehired LV employees would be relocated. We think it likely that an acquirer would make such an assumption to reflect the smaller size of the LV system, and therefore the lesser likelihood of a need for relocation in particular cases. Application of this figure, and an average relocation cost of $5000 per relocated employee yields the following results: Table 43[*] Number of Rehires per Year 4/76 108 10/76 101 10/77 91 10/78 84 10/79 79 10/80 79 10/81 39 _____ ___ Total 581 581 × 25% (relocation rate) = 145.25 relocations 145.25 × $5000 = $726,250 relocation cost Based on the assumptions discussed in calculating EL's LPCs, see pp. 139-40, supra, we estimate that LPCs attributable to the payment of displacement allowances to LV's employees would have totalled approximately: Table 44 Displacement Costs 2,161 × 12% = 259 259 × $76.28 per displacement = $19,757 Addition of the LPCs attributable to lump sum payments (Table 39), monthly dismissal allowances (Table 41), fringe benefits (Table 42), relocation expenses (Table 43), and displacement allowances (Table 44), provides the following estimate of LV's LPCs: Table 45 Total LV LPCs Lump Sum $2,176,510 Dismissals $6,493,543 Fringes $2,207,805 Relocations $726,250 Displacements $19,757 ________________________________ Total LV LPCs $11,623,865 3. The Effect of Bargaining and Other Factors on the Amount of the LPC Deductions The preceding sections set forth what we believe a potential acquirer realistically would have calculated as its liability for EL's and LV's LPCs in the years following the proffered transactions. The acquirers, in negotiations with EL and LV, would have sought a deduction in purchase price for the Ts' properties based on these calculations. The Ts naturally would have resisted these attempts, arguing as they have in this case that LPCs attributable to their systems would have amounted to relatively little. As noted previously, the acquirers would have made a conservative projection of their LPC liability. Our LPC calculations reflect this attitude, as three examples will illustrate. First, we did not consider the impact that attrition-created openings on an acquirer's lines would have on a T's LPCs. Some of these vacancies no doubt could have been filled by a T's surplussed employees, thus reducing LPCs. Second, our calculations did not attempt to account for measures taken by the Ts in the years preceding acquisition to reduce their workforces in an effort to minimize LPCs.[175] Third, we applied conservative 33 percent and 40 percent rehire rates to attrition-created vacancies and to positions made available by rehabilitation and capital improvement programs to determine the extent to which these could be used to reduce the number of dismissals, as well as to permit the rehiring of dismissed employees. Although we are convinced that the assumptions and methodologies we have adopted, or equivalent ones, would have been used by an acquirer, we also recognize that bargaining would likely reduce the size *1291 of the LPC reductions, in part because the T could argue that the acquirer's LPC projections were unduly conservative. Two other considerations, however, would have been even more significant in reducing the size of the LPC deduction. First, labor protection payments made by an acquirer to the Ts' former employees would have been deductible by the acquirer for tax purposes. This would have reduced both the actual cost to the acquirer of the labor protection payments and the size of the LPC deduction by 50 percent. Second, because the acquirer would have been receiving a lump sum LPC deduction from the purchase price, we must calculate the present value — as of the date of agreement on the price of the T's lines — of the acquirer's projected after-tax labor protection payments. As developed in Part VII, we use a 9.13 percent discount rate. Application of this rate to the after-tax LPCs attributable to EL and LV employees produces a deduction of $8,320,737 from the EL purchase price and a $3,929,441 deduction from the LV purchase price. These figures represent the approximate size of the LPC deduction from the purchase prices for EL's and LV's properties. Competitive bidding and the willingness of the potential acquirers to make some adjustments in the size of the LPC deduction to reflect the possibility that their LPC calculations were unduly pessimistic might have reduced the figures slightly, to $8,300,000 and $3,900,000. The acquirers would not have been likely to agree to further reductions, because of their fears that LPCs would exceed even their pessimistic projections, as had occurred in other situations.[176] We do not assert that these estimates are exact. The unpredictable nature of the bargaining process and the uncertainty surrounding attempts to predict labor protection costs would frustrate any such effort. We are satisfied, however, that our estimates reasonably approximate the outcome of negotiations among EL, LV, and the acquirers. 4. Ann Arbor AA contends that it would have incurred no LPCs in the alternative scenario because all its properties would have been acquired by the State of Michigan or private purchasers, and reductions in AA's workforce therefore would have been minimal or nonexistent. AA at 183-84. The GPs argue that, even assuming such an acquisition, the prices paid for AA's lines would reflect a LPC deduction. The GPs reason that the State of Michigan would pay only slightly more for AA's northern properties than AA's "next best marketplace alternative". This alternative, the GPs correctly point out, would have been abandonment. Accordingly, the GPs conclude, the LPCs that would result from this abandonment would be deducted from the purchase price for the northern properties. AA responds that neither existing labor protection agreements — principally the WJPA — nor ICC decisions provide a basis for concluding that it would have incurred labor protection obligations had it pursued its next best marketplace alternative. Because of the unusual context in which disputes regarding these agreements and decisions would have occurred — negotiations between the State of Michigan and AA regarding a purchase price for AA's properties — we address separately AA's contentions regarding these issues. Earlier in this Part we concluded that substantial uncertainty would have surrounded the question whether the WJPA was applicable to the proffered transactions. We think, for the reasons noted above, that similar uncertainty would have characterized negotiations between the State of Michigan and AA. We also concluded that the ICC's line of Okmulgee decisions likely would have been considered distinguishable from the circumstances of the alternative scenario. AA, however, had less trackage (185.4 miles total; 131.2 miles to be abandoned) than some of the railroads involved in the Okmulgee cases. For example, *1292 the Tennessee Central was 287 miles long, and the Rutland was 332 miles long. Moreover, the relatively isolated and small-scale dismissals that such an abandonment would entail would not have presented a particularly compelling case for ICC insistence that AA proceed under § 5(2). Nonetheless, the State of Michigan could have pointed to a number of ICC precedents which strongly suggest that some sort of labor protection conditions would have been applied had AA pursued its next best marketplace alternative. As the GPs note, in circumstances where an abandonment — even of an entire line — is closely linked to financial gain to the controlling interests of a railroad, the ICC has imposed labor protection conditions pursuant to § 1(18) of the Interstate Commerce Act. E.g., Washington & Old Dominion R.R. — Abandonment of Entire Line, 331 I.C.C. 587, 600-02 (1968); Seaboard Air Line R.R. Trackage Rights — Atlantic Coast Line R.R., 324 I.C.C. 506, 520-22 (1964); East Carolina Ry. Abandonment of Entire Line, 324 I.C.C. 506, 520-22 (1964). This line of decisions — which is at times difficult to reconcile with the Okmulgee decisions — would have provided a basis for the State of Michigan to argue that AA would have been saddled with labor protection obligations in the next best marketplace alternative scenario. Moreover, the State of Michigan could have argued that the private acquirer of AA's Toledo-Ann Arbor segment might well have conditioned its purchase on AA's proceeding under § 5(2) — in order to receive the antitrust immunity conferred by § 5(11). Lastly, we think that the State of Michigan could have argued persuasively that labor unions, through the threat of strikes and unrest, could have obtained some sort of labor protection commitments from either AA or the acquirers, even without relying on existing agreements or ICC actions. These considerations indicate that some LPCs would have been incurred in AA's next best marketplace alternative scenario. The labor protection payments likely would have been made by the acquirer of AA's Toledo-Ann Arbor segment. The acquirer would have been willing to assume the obligation for LPCs, in return for a deduction from the purchase price for AA's southern properties, in order to minimize the risk of labor unrest on the acquired southern lines. Moreover, the acquirer would have recognized that the ICC might require that it bear responsibility for labor protection payments. This outcome is suggested by ICC decisions under § 1(18) in which carriers absorbing an abandoning carrier's lines were required to make labor protection payments to the abandoning carrier's employees. See East Carolina Ry., supra, 331 I.C.C. at 507; Seaboard Air Line R.R., supra, 324 I.C.C. at 800-01. The above considerations would have provided the State of Michigan with substantial leverage in negotiations with AA regarding a purchase price for its northern properties. AA's only bargaining power would have derived from a threat to abandon its northern segment. Actual abandonment, however, would have required AA to risk the ICC's imposition of labor protection obligations or the labor union's successful enforcement of existing labor protection agreements. Moreover, it could have resulted in labor unrest that would have jeopardized the rail use sale of the Ann Arbor-Toledo line. We think it unlikely that AA would have been willing to engage in such brinksmanship. Rather, it would have stayed at the bargaining table arguing for a reduction in the size of the State of Michigan's LPC deduction. The threat of abandonment would have been a relatively weak bargaining tactic in these circumstances. The GPs contend that AA's LPCs would have totalled $2,197,000 if the ICC had imposed New Orleans conditions, Kasher (S) (Nov. 1, 1979) at 50, and $1,374,000 if the WJPA had applied, id. at 51. The estimates assume that AA's lines would have been acquired in mid-1974, and that labor protection obligations would have extended from that date until mid-1978 and mid-1979 respectively. Figures in this neighborhood would have provided the starting point for negotiations between AA and the State of Michigan. Because we have concluded that AA's bargaining position would have been inferior to that of the State, we think AA would have been unable to obtain any significant *1293 concession from the State's LPC calculations. We do think, however, that the uncertainty surrounding the likely outcome of ICC action in the next best marketplace alternative scenario would have resulted in some concessions by the State, ultimately leading to a compromise agreement basing estimates of LPCs on the WJPA, rather than the New Orleans, level of benefits. This conclusion is also supported by the fact that the GP's calculations of AA's LPCs assumed a mid-1974 disposition, while bargaining between AA and the State of Michigan would have assumed an April 1, 1976, acquisition. A later acquisition date would have permitted AA to take measures to minimize LPCs. Accordingly, the $1,190,000 figure cited above would have provided the basis for the LPC deduction for AA's northern lines. As in the case of EL and LV, this figure would have been reduced by 50 percent to reflect the deduction of labor protection payments for tax purposes. This conclusion follows from the fact that, in the next best marketplace alternative scenario, the acquirer of AA's southern segment would have assumed responsibility for AA's LPCs in return for a LPC deduction from the purchase price for the southern segment. As in the case of EL and LV, the acquirer could have deducted its future labor protection payments for tax purposes. This would have reduced the LPC deduction the acquirer could have obtained from AA, thus increasing the amount AA could realize in the next best marketplace alternative scenario — and the amount the State of Michigan would be required to pay for AA's properties. This figure also must be discounted to present value as of January 1, 1974. As we have elsewhere, p. 1291, supra, we use a 9.13 percent discount rate, which we apply to the annual after-tax LPCs reflected in the GPs study. Kasher (S) (Nov. 1, 1979) at 51. Although these figures may be somewhat overstated as regards AA's LPCs under the WJPA, because they assume a mid-1974 acquisition date, we think they are adequate as regards AA's LPCs in a 1976 acquisition under the New Orleans conditions. We calculate that bargaining between AA and the State of Michigan would have resulted in an agreement that the present value of AA's LPCs as of January 1, 1974, would have been $493,200. 5. Lehigh & New England As discussed below, we conclude that L&NE's system would have been sold for no more than X. Accordingly, as we have noted earlier, L&NE would incur no labor protection expenses. This conclusion follows from the fact that L&NE could scrap its lines completely and thus incur no labor protection obligations.[177] Thus, an acquirer would be obliged to offer L&NE at least the scrap value of its lines and to agree to hold L&NE harmless from labor protection expenses to prevent it from actually abandoning. X. Timing and ICC Conditions The element of timing is perhaps this opinion's most "prodigious and time-consuming effort in conjecture", CMV Opinion, 445 F.Supp. at 1045. It is nevertheless important for several reasons. Delay of a sale much after 1974 would decrease the chance that a prospective purchaser would look only to 1973 as the base year for calculating the earnings value of a given line. Because traffic levels, and hence revenues, declined sharply over all the Ts' lines after 1973, this would have resulted in earnings values significantly lower than ones based exclusively on 1973 data. See pp. 1352-1353, infra. Continuing physical deterioration of the Ts' rail properties during such delay would have increased necessary rehabilitation costs, which also would have been taken into account by an acquirer. Timing also bears on the amount of labor protection that would have been necessary. See p. 1290 & n.175, infra. For these issues, a *1294 rough estimate on timing might suffice. But the Ts are entitled to interim interest on any sums they could have realized earlier than April 1, 1976, the date their rail properties were conveyed to Conrail; conversely, the proceeds from a disposition after that date must be discounted back to that date. It is therefore necessary for us to find a precise date by which each T's hypothetical sales probably would have been completed. We have no illusions about the rigor with which such an inquiry can be conducted. Although the GPs attempt to analyze the stages of the disposition process in greater detail than do the Ts, the parties agree that it would be futile to attempt to write a detailed "history" of the disposition process. See, e.g., 6 GP at Z 9-10; ELRB at 6-7. At best, it is possible only to render a cumulative judgment on how long each broad phase of the disposition process would have been likely to take. As with several other points in this opinion, resolution of the timing issue requires not a "yes-no" judgment of the sort called for in cases requiring determination of historical fact, but rather a choice of one among a number of plausible outcomes. The conventional concept of burden of proof has little meaning in such circumstances. Any particular choice can, in truth, be little more than an impartial guess based upon the evidence and the argument. We tentatively considered the timing issue in the CUE Opinion, 439 F.Supp. 1351, when evaluating the Ts' entitlement to compensation for erosion of their estates due to the compelled continuation of operations at a loss before the conveyance date, April 1, 1976. We held there that such erosion cannot be laid at the government's door unless a T can show that, but for the Rail Act, it would have sought and should have obtained ICC approval to sell or abandon its rail properties before the conveyance date. The issue of timing therefore became decisive on that of compensable erosion. We stated that the Ts' "burden of demonstrating the probability of obtaining a final decision on and consummating major abandonments or sales before April 1, 1976 ... will be extremely heavy", 439 F.Supp. at 1375, but left it open for each T to try to do so. 439 F.Supp. at 1374-75, 1379-80. The evidence presented in this phase of the case generally confirms that preliminary judgment. We find that all the Ts' proposed sales and abandonments[178] probably would have been completed on, but not before, April 1, 1976. A. The Parties' Positions Except for CNJ, whose unique timing arguments we consider and reject at pp. 1380-1383, infra, the various Ts' contentions about the disposition process are consistent with each other. Indeed, the GPs and the Ts do not seriously disagree about the basic steps of the disposition process, at least when those steps are stated in sufficiently general terms. See 6 GP at T 28-29. Each T, including PC and R before their dismissal from the case, claims that it would have sought to dispose of its rail properties as the first of two steps[179] in a plan of reorganization under §§ 77(d) and (e) of the Bankruptcy Act.[180] Those provisions require such a plan to have the approval *1295 of both the ICC and the reorganization court.[181] The Ts and the GPs envision each T's disposition process as taking place in three broad phases. Commencing with each estate's determination to liquidate its rail properties, the first phase would have encompassed the reaching of sale agreements. Each T would have announced a program to dispose of its rail properties by either sale or abandonment and would have solicited expressions of interest from solvent railroads and public bodies. Each potential purchaser would have had to study the costs and benefits of the various possible acquisitions before serious negotiations could take place. Some negotiations would have been complicated by the need for several interested parties to negotiate with each other simultaneously. For example, no solvent railroad would have purchased any T's lines until it had reached an accommodation with that T's unions over labor protection. Some negotiations would have been wholly contingent on others, as in the case of AA's contention that the potential acquirers of PC's Chicago-Detroit or Detroit-Cincinnati IFLs would have bid for its Toledo-Ann Arbor segment. See AA at 31. Even after the main contours of the reconfigured rail system in the Northeast began to emerge, traffic and revenue projections for many freight lines would have depended in part on the terms on which freight traffic would have had access to connecting lines slated for acquisition by public bodies; similarly, public bodies acquiring parts of the NEC would have had to agree among themselves on how to allocate the costs of intercity passenger service. While these studies and negotiations were taking place, each T would have instituted formal proceedings before the ICC to lay the groundwork for approval of the transactions; they would have filed applications to abandon or sell their entire rail systems at the beginning of the process, submitted to the ICC the data that were being made available to prospective purchasers, and filed applications for approval of specific sales as each was negotiated. At the same time, the ICC would have begun an investigation on its own initiative to develop detailed information about the transportation needs of the Northeast for use in those proceedings. The parties are in apparent agreement that this first phase of the process, however long it would have taken, would have resolved the question of price. Conceivably, the parties to a sale might have reached only an agreement to agree later on a price, but this is unlikely. None of the parties makes this argument, and the briefs apparently assume that the Ts and their acquirers would have agreed on a price before seeking ICC approval. See, e.g., 6 GP at T 28. In addition, some of the witnesses thought the negotiation phase would resolve the question of price. Whitehurst (S) (Nov. 1, 1979) at 39; Hesse (S) (R Probable Actions) (Dec. 31, 1978) at 18-20. Finally, the ICC, at least to the extent it was proceeding under § 5 of the Interstate Commerce Act, would not have considered an application that did not include a price term. 49 C.F.R. § 1111.1(c)(2) (1974). The second phase, partly overlapping the first, would have consisted of the ICC's deliberations over the sales and abandonments. As we discussed in Part IX, supra, sales would have been evaluated under § 5(2) of the Interstate Commerce Act; abandonments would have been evaluated under § 1(18).[182] In parallel with these deliberations, the ICC would have conducted any further inquiries into the Ts' financial condition required by the Bankruptcy Act. After giving its approval to the sale and *1296 abandonment applications, along with any conditions it saw fit to attach, it would have certified the Ts' plans of reorganization, as modified, to their respective reorganization courts. The final phase of the disposition process would have encompassed any steps necessary to complete the dispositions after the ICC approved them. At a minimum, § 77(e) of the Bankruptcy Act would have required each estate's reorganization court to hold a hearing on and give its approval to the estate's proposed dispositions. Further time would have been necessary for the Ts and the acquirers to negotiate any changes made necessary by ICC modifications of the proposals as initially submitted and to formalize the sale agreements, and for the acquirers to finalize credit arrangements. In addition to agreement on the broad steps of the disposition process, there is now no dispute as to when each T would have begun to dispose of its properties. This issue was settled by paragraph 10 of the Second Pre-Trial Order, which provided that "the transferors (including the Trustees of the Erie Lackawanna) are entitled to assume that, in the absence of the Rail Act, serious consideration and discussion of dispositions for continued rail use would have begun not later than February, 1973". Thus, even though EL did not lose hope of achieving an income-based reorganization until January 1975 in the real world, it now believes, with the other parties, that it would have filed for authority to sell or abandon its freight lines in mid-February 1973 in the alternative scenario. EL assumes that PC would have filed for abandonment in February 1973, and that EL's management then would have realized that Western carriers would have been prompted to purchase PC's IFLs, thereby threatening to divert traffic from EL and dashing its hopes for a successful reorganization. 1 ELF at 19-20. LV, which actually filed a petition for abandonment with its reorganization court in early March 1973, contends that it would have filed at least as soon in the alternative scenario, LV at 20; R would have followed suit sometime in the spring of 1973, 2 RF at 214-15; and AA, which had had an application to abandon its car ferry operations across Lake Michigan pending since 1972, would have applied to the ICC for permission to sell or abandon its rail properties not later than May 1974. AA at 6, 28 n.1. Although the general framework and the starting point of the disposition process are thus not in dispute, the parties' disagreements over how long each phase of the process would have taken lead to sharp differences at the bottom line. The GPs offer relatively detailed testimony on each phase of the disposition process. Several experts analyzed the time required for potential purchasers to study the proffered lines and for the basic terms of purchase to be negotiated.[183] Others addressed the difficulties involved in the multi-sided negotiations that would have had to take place among the Ts, solvent buyers, and labor;[184]*1297 among various Ts, solvent buyers, and public bodies over who would acquire the commuter lines and other lines in the JT areas, and over the switching and trackage fees to be levied by public bodies that acquired such lines; and among the various public bodies that would have acquired segments of the NEC spine.[185] Still others addressed the various tasks the ICC would have had to perform, and the steps necessary to consummate the transactions after ICC approval.[186] Relying partly on analogies to past sale and abandonment transactions and partly on ad hoc judgments about the transactions involved here, those experts concluded that at least three-and-one-half to four years would have been necessary for potential buyers and labor to complete necessary studies and reach agreement with the transferors, see 6 GP at T 50-52, 58-64, 94-96; that at least four-and-one-half to five years would have been necessary to obtain ICC approval for the transactions, see id. at T 104; and that at least eight months would have been necessary to complete the transactions after final ICC action, see 6 GP at T 147-52. Assuming that PC began its disposition process in February 1973, these figures would point to a completion date no sooner than October 1981. The GPs choose to err on the side of generosity and argue for a completion date no sooner than late 1978 or early 1979. Underlying the GPs' timing case are three vital premises. First, the GPs contend that the federal government would have been willing to provide to the Ts for an indefinitely long period whatever aid was necessary to prevent them from shutting down for want of cash. Second, they contend that Congress would neither have imposed time limits on the ICC's deliberations nor have altered the ICC's substantive legal mandate for the occasion. Finally, the GPs contend that the ICC would not have considered each T's sale and abandonment applications in isolation, but rather would have consolidated all the proceedings in order to evaluate the effect of each proposed transaction on the others and on the region as a whole. The effect of this last contention is to slow each disposition process at least to the pace of the slowest T, which, because of its size and complexity, inevitably would have been PC. The Ts' timing cases (including that of PC, whose timing evidence the Ts remaining in the case generally adopt) rest on the negation of these three premises.[187] Their *1298 primary argument is that Congress would not have provided them with any financial aid, and that their dispositions therefore would have to have been completed before they ran out of cash and had to cease operations. Content to argue that because potential buyers, labor, and the ICC could not tolerate a shutdown, they would have found some way to complete the transactions before the Ts became cashless, the Ts eschew discussion of supposedly analogous historical transactions, and they offer little testimony about the details of each phase of the disposition process. Under this theory, PC, LV, and R equate their disposition dates roughly with the date by which each would have run out of cash: July 1974, March 1974, and mid-1974, respectively.[188] AA, rather differently from the others, claims that it would have become cashless in May 1974, but does not deny that it would have received interim aid from either the state or federal government; rather, it contends that it could have completed its dispositions by the spring of 1975 because all of its potential acquirers were familiar with its operations and hence would have needed little time to formulate a bid, and because none of its dispositions would have been opposed before the ICC. EL, true to its theory that its disposition process would have paralleled that of PC, links its disposition date to PC's. PC and EL thus project completion of their disposition processes roughly eighteen months after their determination to liquidate; the others project even faster dispositions. Partly in support of their cashlessness argument and partly as a fallback, the Ts argue that Congress would have imposed a deadline on the ICC's deliberations or drastically changed its legal mandate. Failing that, they argue, the reorganization courts would have imposed deadlines on the ICC. Finally, several Ts deny that the ICC would have consolidated their applications with those of PC or other Ts. B. Analysis 1. Likely Congressional Initiatives in the Alternative Scenario The most serious source of uncertainty in any inquiry into the disposition process lies in estimating the likely behavior of Congress in the alternative scenario. We specifically opened the door to argument on this issue in the CMV Opinion, 445 F.Supp. at 1008-09, and suggested that Congress might have provided "interim financial aid" to the Ts or enacted "regulatory reforms". The GPs and the Ts both contend that Congress would have undertaken special initiatives shortly after the Ts began their liquidation programs, but they focus on different halves of this suggestion. The GPs posit an open-ended program of operating subsidies designed to keep the Ts afloat during an indefinitely prolonged disposition process. At the same time, they reject any argument that Congress would have attempted to shorten the process by imposing a deadline for ICC action or by changing its mandate. The Ts' position is the opposite of the GPs': they contend that Congress would have granted them no interim aid, but would have placed time limits and other constraints on the ICC's deliberations. Their argument against interim aid implicitly depends on a single premise: that Congress would have realized that the Ts' dispositions could be completed speedily if the potential acquirers, interested third parties, and the ICC were all forced to meet the deadline imposed by the Ts' imminent shutdown for want of cash. The Ts make a plausible case that extremely fast dispositions, on the order of eighteen months, might have been feasible if Congress had been absolutely determined to force such an outcome. Nevertheless, we think it almost certain that Congress would have chosen to lengthen the disposition process to some extent, and therefore would have provided some form of emergency aid to the Ts. The Ts do not deny that Congress' primary motive in the alternative *1299 scenario would have been to maximize the likelihood that the Ts' lines would remain in rail use. See 2 RF at 255-57. According to the premise laid down in the CMV Opinion, 445 F.Supp. at 1005, we must assume that Congress would not have ensured that outcome by acquiring the lines for the federal government; hence that motive would have manifested itself by actions designed to maximize the chance that sales such as those the Ts propose would be consummated. But Congress would have seen that the speedy disposition schedule the Ts propose could backfire, for, as the GPs' evidence shows, potential bidders, particularly solvent railroads, might simply have refused to bid at all if they were not given enough time to perform the studies necessary to form a rational judgment about the value of the Ts' lines. See 6 GP at T 177. The Ts' evidence does not rebut this; it purports to show only that potential bidders could have met the Ts' schedule if forced to do so. That evidence does not bear on whether Congress would have risked that experiment. In deciding whether more time than the eighteen months allowed by the Ts would have improved the likelihood of substantial rail use sales, Congress would have been entitled to look to prior sale and abandonment transactions. The Ts, however, cite no transaction that was anywhere near the size and complexity of the ones PC alone would have undertaken and was completed in anywhere near the eighteen months they postulate, and the GPs' survey of prior transactions demonstrates that much simpler transactions commonly took far longer. See 6 GP at T 152. Congress could hardly have escaped the conclusion that more time would have substantially helped the chances for a successful disposition. Apart from the objective reasonableness of providing for a longer disposition process, we are persuaded that it would have been inevitable as a matter of political reality. Representative members of most affected interest groups — shippers, labor, and state and local governments — testified persuasively and without effective contradiction that they would have demanded that Congress provide enough interim aid to allow for a disposition process long enough for them to sort out their options and organize an effective defense of their interests before the ICC. See 6 GP at T 167-172.[189] Such efforts could not have been effective until the Ts' sale and abandonment plans were completely negotiated and set out before the ICC. If, as we find below, that would take a minimum of a year, then at most six months of the Ts' eighteen-month schedule would remain for ICC proceedings. Even if the ICC could have met its statutory obligation to afford a "reasonable opportunity for interested parties to be heard", Interstate Commerce Act § 5(2)(b), 49 U.S.C. § 11344, in that period, which we doubt, these constituent groups would have argued to Congress, with much justice, that the ICC could be little more than a rubber stamp for the Ts' proposals in such a condensed proceeding. With virtually every constituent group but the Ts themselves clamoring for more time to prepare their cases, Congress would inevitably have responded. For the same reason, while some of the Ts suggest that Congress would have curtailed the ICC's role in the disposition process or eliminated the right of third parties to be heard, we agree with the GPs' experts that such a move would have been politically unlikely. This is the price the creditors of the Ts' estates would have paid for living in a democracy. See CUE Opinion, 439 F.Supp. at 1373. We therefore find that Congress would have allowed more time for the disposition process than the eighteen months for which the Ts argue, and hence would have provided the Ts with at least some interim financing. On the other hand, we also reject the GPs' contention that Congress would have provided such aid without limit as to time or amount. Congress would have had to announce a deadline on interim aid if it wanted sales to be negotiated at all. To *1300 show this, it is necessary to note only that some of the sales would be of admittedly losing operations to public bodies, and that none of the sales could take place without the acquiescence of labor unions. Common sense and the GPs' own evidence both indicate that neither labor nor public bodies would have agreed to do those things unless and until faced with the even worse prospect of imminent shutdown. The GPs suggest no reasons why these actors would not have been content simply to put off the evil day as long as the federal government was willing to subsidize continued operations.[190] Under the principles stated in the CUE Opinion, however, the government could constitutionally have forced the Ts to continue their hopelessly losing rail operations, while their estates continued to erode, only if the government could show a certainty or a high likelihood that the erosion would cease within a reasonable time.[191] We have grave doubts whether it could have made that showing before the Ts' reorganization courts if it had followed the course the GPs suggest. We need not go so far as to hold that a reorganization court would have been justified in granting a T immediate permission to shut down. It is enough to say that Congress would have been aware that the indefinite continuation of interim aid would have created a danger of triggering a court-ordered shutdown. See CUE Opinion, 439 F.Supp. at 1377-78.[192] But Congress was acutely sensitive to that danger in the real world, see id. at 1361; Northeast and Midwest Railroad Transportation Crisis: Hearings Before the Subcommittee on Surface Transportation on S. 2188, 93d Cong., 1st Sess. 886-87 (1973) (comment of Mr. Adams), and, as the GPs' own evidence shows, preservation of essential rail service would have been its overriding priority. Clausen (S) (Nov. 1, 1979) at 3-4. Hence, we are certain that it would have avoided that danger by announcing, in advance, a date in the reasonably near future after which it would cease to subsidize the Ts, thereby forcing all necessary actors to agree by that date. In effect, Congress would have been forced to write the sort of "essay in brinksmanship" we condemned on the part of the Ts in the CUE Opinion, 439 F.Supp. at 1379.[193] Apart from fears of a court-ordered shutdown, Congress would have had more mundane reasons to place a deadline on the negotiation process. The GPs' thesis requires us to believe that Congress would have knowingly opened the federal treasury to a drain of unknown but very substantial size for an indefinite period, ceding all control over the duration of the subsidy to any necessary actor that chose to be intransigent, and without any assurance of a desirable result. The record supports our conclusions. Aage Clausen, the GPs' main witness on *1301 congressional actions in the alternative scenario,[194] correctly stated that "[e]vidence of prior Congressional action is the best predictor of how Congress would have acted in the period under consideration". Clausen (S) (Nov. 1, 1979) at 9. Of the rail legislation analyzed by Clausen, however, none committed the federal government to provide substantial aid to privately owned railroads without retaining strict control over its amount, duration, and use. None is remotely like the open-ended commitment to underwrite operating deficits that the GPs' thesis requires.[195] To be sure, Congress would have needed to give potential acquirers enough time for study so that they could submit rational bids, since they might otherwise have balked at bidding at all. It would also have needed to allow some time for the voluminous details of the transactions to be negotiated. But there is no evidence, and no reason to believe without evidence, that any time beyond this bare minimum would have made sales for rail use any more likely. Hence there would have been no reason for Congress to spend substantial amounts of taxpayers' money on subsidies for any longer period. For similar reasons, we are persuaded that Congress would have placed a time limit on the ICC's deliberations. The fact that Congress placed a deadline on the USRA's creation of the Final System Plan in the real world is a strong indication of that, for while the USRA formulated its plan, Congress was providing much the sort of subsidy that would have been necessary in the alternative scenario. Congress would probably have seen it as far less expensive to hasten the ICC's deliberations by expanding its staff for the occasion, concentrating the same resources that in the real world went into the creation of the USRA and the ICC's Rail Services Planning Office, and expansion of the Department of Transportation, than to pay operating subsidies to the Ts for an extended period. Even aside from its wish to protect the public purse, we doubt that Congress would have relied on the ICC's normal procedures without giving it some kind of prod. Congress showed its mistrust of those procedures at the time of the rail crisis and on *1302 many occasions thereafter. Its rejection of the ICC's proposed solution to the rail crisis and creation of an entirely new agency to draw up the Final System Plan were partly due to the mistrust engendered by the Commission's traditional delays.[196] At the same time that it provided interim aid, Congress would have set time limits on the ICC's deliberations. Congress repeatedly passed legislation imposing deadlines on the ICC's handling of sale and abandonment applications relatively soon after the Rail Act, as discussed in detail below. The first such deadline, enacted in the Railroad Revitalization and Regulatory Reform Act of 1976, was proximately inspired by the ICC's unusual delays in handling the Penn Central and Rock Island merger applications. See Corber (D) (July 7, 1980) at 128-30, EL App. K at 198-200; S.Rep.No. 499, 94th Cong., 2d Sess. 19, reprinted in [1976] U.S. Code Cong. & Ad.News 14, 33. But the Penn Central merger had been completed and the Rock Island had already dragged on for ten years by 1973. Thus there is no reason to expect that Congress' actions in the alternative scenario would not have been guided by the same inspiration. GP witness Robert Corber, formerly a Commissioner of the ICC, admitted that Congress "would have urged the Commission to promptly dispose of the applications in the proceedings that were before it". Corber (D) (July 7, 1980) at 137, 8 RF App. at Ex. 77. Given its record of placing deadlines when it considered speed important, we cannot believe that it would have relied on moral suasion alone. The GPs advance little evidence to the contrary. Virtually all of their experts assume that we ruled out the possibility of such legislation. See, e.g., Corber (D) (July 7, 1980) at 80-82, EL App. K at 195-196; Brooks (D) (July 14, 1980) at 74-75, EL App. K at 214-215. In fact, in the CMV Opinion, 445 F.Supp. at 1008-09, we explicitly ruled it in.[197] Although some witnesses testified, and we have held, that pressure from shippers, labor, localities, and other interested parties would have prevented Congress from cutting off their right to participate in ICC proceedings or from making drastic changes in its substantive mandate, see 6 GP at T 168-72, this factor does not proscribe time limits. Indeed, those same pressures did not prevent Congress from passing the ICC reform legislation we noted above. As the ICC's experience under those reform statutes shows, its procedures could have been expedited considerably without drastic changes in its mandate. For example, although interested parties have a statutory right to submit evidence and argument in sale and abandonment proceedings, the GPs concede that they have no statutory right to submit it orally. See 6 GP at T 122. The ICC could have limited or dispensed with hearings, as it already did for certain classes of cases, see Brooks (D) (July 14, 1980) at 164, 7 RF App. Ex. 71, and it could have placed a time *1303 limit on submission of written evidence and argument, see Milwaukee Road Abandonment Proc., 44 Fed.Reg. 50975. Similarly, it could have dispensed with initial hearings before an administrative law judge. See Brooks (S) (Nov. 1, 1979) at 16. Preparation of an environmental impact statement as required by the National Environmental Policy Act (NEPA) could have been a lengthy process. Congress, however, granted exemptions from NEPA in the Rail Act and in two ICC reform statutes aimed specifically at expediting deliberations over sales and abandonments proposed by railroads in reorganization. Milwaukee Railroad Restructuring Act, § 19, 45 U.S.C. § 917; Rock Island Railroad Transition and Employee Assistance Act, § 113, 45 U.S.C. § 1010. These statutes, combined with Congress' frequent practice of granting NEPA exemptions when time is of the essence and to particularly complex transactions, see, e.g., Trans-Alaska Pipeline Authorization Act, 43 U.S.C. § 1652(d); Alaska Natural Gas Transportation Act, 15 U.S.C. § 719h(c)(3), convince us that a similar exemption would have been provided here, especially since continuation of rail service rather than its abandonment and replacement by trucks and buses was obviously in the interest of the environment. The principles that would have guided Congress are thus relatively clear. Estimating the numbers it would have chosen is more difficult. With respect to the time limit it would have placed on the ICC, the deadline imposed on USRA in formulating the Final System Plan is highly pertinent, for USRA dealt with the same properties that would have been before the ICC. That deadline was originally fifteen months, later extended to nineteen months because of delay by President Nixon in appointing the Board of USRA, see CUE Opinion, 439 F.Supp. at 1363. See Rail Act § 207(c). That extension would not have been necessary in the alternative scenario. We stressed in the CUE Opinion, 439 F.Supp. at 1373-74, 1375, differences between the task that actually faced USRA and the one that would have faced the ICC. See also 180-Day Appeals, 384 F.Supp. at 922. While they are undoubtedly substantial, on further reflection they are not so great as we there portrayed them. Indeed, in one important respect the ICC's task would have been the simpler one: a basic outline of what rail lines would stay in rail use and who would own them would have been presented to it as a product of negotiations. By contrast, USRA had to design a rail system for the Northeast from scratch. It had not only to decide what would stay in rail use and what would not, but also who would own what lines; it had to negotiate with solvent railroads to persuade them to acquire lines, see 3 GP at I 28-34; and it pondered alternative structures for Conrail itself, 180-Day Appeals, 384 F.Supp. at 958 n.73. It managed to meet its deadline even though a major carrier, EL, was thrown into its lap only three months before the deadline, see CUE Opinion, 439 F.Supp. at 1386-87. Furthermore, we did not envision the possibility of regulatory reform at the time of the CUE Opinion. NEPA, which we viewed as a major stumbling block for the ICC, 439 F.Supp. at 1374, probably would have been rendered inapplicable to the ICC as it was to USRA. The ICC's "time consuming hearing requirements", id., are less onerous than we then imagined, given that it had power to limit the submission of evidence to writing. Even the scrutiny it would have had to give to "considerations of competitive effect", id. at 1373, would have been lessened by analogy to the failing company defense of anti-trust law, see CMV Opinion, 445 F.Supp. at 1009-10. Still, assuming, as we do, that Congress would not have made major changes in the ICC's legal mandate, Congress would have had to allow at least somewhat more time for the ICC in the alternative scenario than it thought necessary to give the USRA in the real world. The ICC would have had to develop an adequate record to support its decisions on judicial review. Swift & Co. v. United States, 343 U.S. 373, 381-82, 72 S.Ct. 716, 720-21, 96 L.Ed. 1008 (1952); I.C.C. v. Union Pacific R.R., 222 U.S. 541, 547-48, 32 S.Ct. 108, 110-11, 56 L.Ed. 308 (1912). By comparison, the Final System Plan was not subject to judicial review, but only to review by Congress, see Rail Act §§ 208(a), *1304 209(a) — and Congress was doubtless more interested in the USRA's results than in its reasons. Although its hearings could have been conducted mostly or entirely on paper, the ICC would have been bound to consider each argument raised by the scores of interested parties who certainly would have intervened. Solvent railroads that had bid unsuccessfully on a given line would have tried to convince the ICC that they could provide more or better service, or were in better financial condition, than the high bidder; other solvent railroads would have sought traffic protection conditions, which would have forced the ICC to study possible traffic diversions that might result from the sales; public bodies that were cast in the role of buyers of last resort for hopelessly losing passenger operations would have pressed the ICC to fob them off on the acquirers of potentially profitable freight lines; shippers could have been counted on to contest abandonments; and the ICC would have had to approve the labor protection agreements previously negotiated. We therefore find that the fifteen months Congress initially allowed USRA constitutes a lower bound on the amount of time it would have allowed the ICC. For the reasons stated in the preceding paragraph, we find it is not a very distant bound. Further insight can be gained by examining the large number of statutes enacted since the Rail Act that have placed time limits on the ICC's handling of sale and abandonment applications. Most closely analogous are three statutes specifically applicable to railroads in reorganization: the Milwaukee Railroad Restructuring Act, P.L. 96-101, 93 Stat. 736 (1979), the Rock Island Railroad Transition and Employee Assistance Act, P.L. 96-254, 94 Stat. 399 (1980), and the provisions of the new Bankruptcy Code applicable to railroads in reorganization, 11 U.S.C. §§ 1170, 1172. Section 5 of the Milwaukee Act, 45 U.S.C. § 904, transfers all jurisdiction over abandonments by the Milwaukee Railroad from the ICC to the railroad's bankruptcy court, and authorizes the court to set a time limit of six months, or even less if it chooses, on the ICC's deliberations over sale applications by the Milwaukee; in the latter case, the ICC may act without a hearing if it chooses, id. § 5(b)(2), 45 U.S.C. § 904(b)(2). Sections 1170 and 1172 of the Bankruptcy Code closely parallel the Milwaukee Act. Section 111 of the Rock Island Act, 45 U.S.C. § 1009, fixes a maximum deadline of 100 days for ICC evaluation of sale applications filed by the Rock Island. These very short deadlines show great concern for the rights of creditors of an eroding estate, a concern that could only have been heightened in the non-Rail Act world, for each month's delay by the ICC would have cost the government a substantial amount in interim subsidies. Two enactments place time limits on ICC handling of sale and abandonment applications filed by any railroad, including ones not in reorganization. Section 402 of the Railroad Revitalization Reform Act of 1976 placed a basic thirty-one month time limit on ICC consideration of sale and merger applications, although the ICC was empowered to violate that limit, if necessary, upon report to Congress. The Staggers Act of 1980 made that thirty-one month time limit absolute, and shortened it to ten months for transactions other than mergers and acquisitions of control, which would be the case here. The thirty-one month deadline set by the RRRR Act and confirmed by the Staggers Act is a plausible upper bound on the amount of time Congress would have allowed the ICC in the alternative scenario. It is contained in statutes of general applicability, and so it was designed to apply even to very large transactions. But for the reasons stated above, we think Congress would have imposed an even tighter time limit than usual in this case.[198] Between the lower bound of fifteen months and this upper bound of thirty-one months, we think that a range of twenty to twenty-four months is a reasonable estimate of the time limit Congress would have set. *1305 Knowledge that Congress' provision of interim aid would be sharply limited in time would also have shrunk the period devoted to studies and negotiations. Shannon testified on behalf of EL that a study sufficient to support a "reasonable business decision" to buy portions of PC or EL could be made in six months. Shannon (D) (May 23, 1979) at 343, EL App. K at 163. That estimate is supported by two objective sources. First, an employee of Norfolk & Western, studying the possibility of a purchase of PC and other Ts by Norfolk & Western, Chessie, other solvent railroads, and public bodies, estimated in June 1973 in an internal memorandum that potential acquirers would be ready to bid within six months. Zeis (D) (May 8, 1980) Ex. 5, EL App. K at 245-246. Second, Chessie began serious negotiations with USRA over purchase of EL, R, and PC lines designated for sale to it by the Preliminary System Plan after only five months of study.[199] That transaction was similar in size and complexity to those in the alternative scenario, since it encompassed essentially all the freight lines in R's rail use proffer, including those originally proposed for the JT, most EL freight lines east of Akron, plus a branch of PC. See 3 GP at I 29. We have elsewhere noted that Chessie's study was a reliable one, see pp. 1248-1249, supra, and the GPs give no substantial reason why its real-world experience would not have been comparable to what would have obtained in the alternative scenario. Nor do they even attempt to discredit Norfolk & Western's estimate. That neutral and contemporaneous judgment and Chessie's real world experience are far more persuasive to us than the GPs' evidence, which consists entirely of expert testimony prepared for this litigation. We therefore find that five to six months would have been sufficient time for potential acquirers to study the Ts, including PC, before serious negotiations could begin. Further time would have been necessary for negotiations between the Ts, acquirers, and labor, and for any legislation necessary to enable public bodies to buy the Ts' properties. The evidence on how long would have been necessary to do that is surprisingly scant. The Ts offer virtually none.[200] The GPs consider that the limiting factors would have been labor negotiations and negotiations with public bodies. Several experts testified that labor negotiations would have taken two further years. See, e.g., Myles (S) (Nov. 1, 1979) at 15-22; Luna (S) (Nov. 1, 1979) at 17-18. The one expert who gave a number for public body purchasers stated that they would have needed at least three-and-one-half years for study and negotiations. Whitehurst (S) (Nov. 1, 1979) at 45. We give little credence to this testimony, however, for all of these witnesses followed the GPs' erroneous assumption that unlimited financial assistance would have been available to the Ts. See Myles (S) (Nov. 1, 1979) at 2; Luna (S) (Nov. 1, 1979) at 2; Whitehurst (S) (Nov. 1, 1979) at 9. Furthermore, the GPs' two most important labor witnesses on this issue refused to accept the premise, required by the CMV Opinion, that Congress would not have stepped in to save jobs if the Ts had shut down. See, e.g., Luna (D) (June 4, 1980) at 91, 8 RF App. Ex. 79. As we have discussed elsewhere, labor would have been more pliant in such a situation than its history would suggest. An analogy is suggested by the Miami Accords, discussed at p. 1268, supra, negotiated in only a few weeks, which resolved the labor protection issues arising out of the reorganization of the Rock Island and Milwaukee Railroads. The Miami Accords thus show that labor negotiations can move quickly if the alternative is a complete shutdown. All things considered, we estimate that the necessary studies could have been made *1306 and the Ts could have concluded sale agreements, including agreements on price, within ten to twelve months after PC announced its intention to liquidate. We further estimate that Congress would have set a twenty to twenty-four month deadline on the ICC's consideration of the Ts' sale and abandonment applications. These conclusions imply that Congress would have had to provide interim assistance to the Ts in some form. We find that aid probably would have been provided in the form of grants, rather than priority loans. That the interim assistance provided for by the Rail Act was in the form of grants is persuasive of this.[201] That choice was not accidental; the legislative history shows that Congress believed that forced priority loans would have been unconstitutional. See Clausen (S) (Nov. 1, 1979) at 37-38. Whether a court should have held such forced loans unconstitutional is a difficult question that we need not decide. It is enough that Congress thought it was; Congress is entitled to give the creditors the benefit of its own understanding of the just compensation clause. 2. Probable Actions of the ICC Our finding that Congress would have been likely to impose time limits on the ICC in the alternative scenario generally spares us the task of reconstructing the likely actions of the ICC in any detail. Nevertheless, two broad clusters of issues require discussion. First, the GPs argue that the ICC would have modified or placed conditions on the sale agreements and abandonment proposals presented to it by the Ts. Each of the modifications and conditions they propose would have reduced the price the Ts would have realized. Second, the Ts challenge the GPs' assumption, which we have accepted arguendo thus far, that the ICC would have consolidated all the Ts' sale and abandonment applications and considered them together in a regional planning process. (a) Conditions Reducing Value The GPs argue that the ICC would have placed several kinds of conditions on the Ts' applications to sell and abandon their properties. At various points, they argue that the ICC would have required labor protection benefits for displaced employees, imposed traffic protection conditions to prevent undue diversion of traffic from solvent carriers' lines, and approved sales of certain lines only if the acquirer agreed also to acquire either other lines slated for abandonment by the Ts or unprofitable lines slated for acquisition by public bodies. They so argue partly to support their timing argument by illustrating the complexity of the ICC's task, and partly because some conditions would in themselves have diminished the amounts the Ts would realize. Insofar as the GPs offer these arguments on the issue of timing, they are subsumed within our conclusion in the preceding section that Congress would have placed a time limit on the ICC's deliberations. Labor protection is discussed in Part IX.[202] Hence, only the contentions that the ICC could have conditioned the sale of some lines on acquisition of less profitable lines or of lines that would otherwise be abandoned need be addressed here. These contentions lose much of their force because of our finding that the private acquirers of EL's and LV's main freight lines would have acquired those carriers' predominantly freight lines in the joint terminal area as well. See pp. 1313-1316, infra. Except for the contention that the ICC would have required the acquisition of lines slated for abandonment, the only lines to which the GPs' argument now applies are EL's passenger operations and, assuming that a solvent railroad would have been the high bidder for the AA's Toledo-Ann Arbor segment, the AA line north of Ann Arbor. The GPs offer considerable evidence that shippers on lines *1307 slated for abandonment and the public bodies slated to acquire these lines would have petitioned the ICC to foist them onto the solvent acquirers of those carriers' other lines. We are persuaded that they would have tried, but we hold that they would have failed. In the CMV Opinion we "record[ed] our disagreement with the suggestion of the Government parties that the Interstate Commerce Commission would or could have used its power to withhold approval so as to insist on the inclusion of unprofitable mileage to a degree that would have driven down the price that purchasers would pay down to the level of scrap value ...." 445 F.Supp. at 1010 & n.19. Although we also stated that "[w]e do not intimate that the Commission would have had no alternatives other than complete acceptance or rejection of the proposals as made", we had in mind such common conditions as traffic protection, trackage rights, or inclusion of other carriers. With one exception, the cases the GPs cite either involve conventional conditions or simply allude to the ICC's power to impose conditions in general terms; we distinguished those and similar cases in the CMV Opinion, id. The sole exception, and the GPs' only specific authority for conditions involving acquisition of unwanted trackage, Chicago, Kalamazoo & Saginaw Ry. Acquisition, 166 I.C.C. 265 (1930), is of little precedential value here. The condition imposed in that case required the addition of only two extra miles of track to a purchase of 10.3 miles. Moreover, that case was decided by a division of the ICC and was not reviewed by the full Commission or the courts, and we have neither been referred to nor found in our own research any instance where it has been cited or followed by the ICC in the fifty years since it was decided. Assuming the ICC had power to condition the Ts' sales to solvent railroads on acquisition of substantial trackage slated for abandonment or for acquisition by public bodies because of inability to operate at a profit — a question we need not decide — we reject the GPs' arguments that it would have done so. In the first place, if the GPs wished to show that the ICC would not have permitted certain branch lines slated for abandonment to go out of rail use, they bore the burden of coming forward with reasonably specific evidence that the ICC would have insisted on adding particular lines or sets of lines to the Ts' proffers, see Second Pre-Trial Order ¶ 9(e), Sp. Ct. Rep. at N 11800. The GPs, however, have made no attempt to show that any particular nonproffered line would have been added. Nor did they attempt to show that any particular nonproffered line did not qualify for abandonment under the ICC's usual standards, even though their own witness stated that the ICC would have relied substantially on established abandonment criteria. Corber (S) (Jan. 31, 1979) at 42. Hence, as to lines slated for abandonment, the GPs' argument fails for want of proof. Furthermore, we seriously doubt that the ICC would have conditioned acquisition of EL's main freight lines on the acquirer's conducting EL's passenger operations, regardless of pleas by New Jersey for it to do so. The relative bargaining strengths of solvent acquirers and the State would have been the same as they were in the negotiation phase. For the reasons discussed at pp. 1317-1319, infra, solvent acquirers almost certainly would have refused to undertake those carriers' passenger services on any terms, even to the point of preferring to let the Ts' freight lines go out of rail use. The solvent acquirers could credibly threaten to walk away from the entire transaction if the ICC imposed such a condition and the GPs recognize that the ICC could do nothing to stop them. 6 GP at W 41-42; see also ELRB at 270-71. The GPs suggest that the ICC in that circumstance could have required EL to operate a viable pared-down core system; but, passing the question whether the ICC could constitutionally have done so, the record contains no evidence that EL had any such viable core standing alone. Unable to sell any part of its system with such a condition imposed and unable to operate it profitably, EL would have had a constitutional right to cease all its operations. See CUE Opinion, 439 F.Supp. at 1356-57, 1364. The GPs' own witnesses, however, testified that the *1308 ICC's "foremost objective" would have been to preserve needed rail service. Corber (S) (Jan. 31, 1979) at 19. Because there would have been no real doubt that EL's passenger operations would have been kept in service by public bodies in any event, and because a stubborn attempt to foist them onto solvent railroads could well have resulted in loss of the entire EL system for rail use, the ICC almost certainly would have refused to gamble.[203] This reasoning does not apply to AA, since Michigan stood ready to buy the Toledo-Ann Arbor segment if a solvent bidder for that branch balked at a condition that it also acquire the lines north of Ann Arbor. But the only reason the GPs advance for supposing that the ICC would have conditioned the acquisition of one line on another is that the ICC "would have prevented the transferors from skimming the cream from their systems, while in allocating to public bodies the cost of operating other transferor lines essential to both the public interest and the very earnings claimed by the transferors". 6 GP at X 11. Passing the "public interest" clause, which the GPs do not expand upon, that reason does not apply to AA, for the earnings value of its Toledo-Ann Arbor segment does not depend on traffic from the northern line. See p. 1365, infra. Such a condition would tend to drive the price AA could realize down to X. But the GPs' own evidence is that the ICC would not deliberately have sought to reduce prices paid to the Ts without a good reason. Corber (S) (Jan. 1980) at 24. Because the GPs do not suggest a reason, we find that the ICC would not have imposed such a condition. (b) Early ICC Approval of Particular Transferors' Sale and Abandonment Applications In our CUE Opinion we indicated our preliminary view that "[t]he solution which the ICC would have had to devise to the regional rail crisis would have involved a complex piece of regional planning, and we think it unlikely that the ICC would have tied its hands early in the proceeding by permitting abandonment or sale of major lines even before it knew what shape the surviving system might take." 439 F.Supp. at 1375. As we envisioned this regional planning process, the ICC would have recognized in advance the possibility or likelihood of interrelationships between the various Ts' sale and abandonment applications. It therefore would not have given individual applications detailed attention until all sale applications, or at least all of the major sale applications, were before it. All of the applications would have been consolidated and studied as a whole, and the ICC would not have permitted any single transaction to be consummated until it was ready to give its approval to all of them. Consolidation of the Ts' applications thus has an important effect on the time at which each could have completed its sales and abandonments. No T except CNJ has gone so far as to argue that it would have reached agreement with a purchaser and received all necessary administrative approvals before it became clear that other Ts were also going to propose major sales and abandonments, and we reject that argument at pp. 1380-1383, infra. Nevertheless, several Ts take issue with our preliminary view in another way: they argue that they could have persuaded the ICC to approve part or all of their transactions before those of the other transferors — principally PC — even after the need for regional planning became evident. The evidence is clear that the ICC would have entertained a strong presumption in favor of consolidation of all the Ts' applications. In its report rejecting the PC trustees' *1309 proposal to liquidate, the ICC emphasized the need for a "solution which deals with Penn Central in the context of a territory-wide approach encompassing the other carriers serving the Northeast". Penn Central Transportation Co. Reorganization, 347 I.C.C. 45, 112 (1973). GP witnesses Robert Brooks, who was director of the ICC's Office of Proceedings at the time the Ts' applications would have been filed in the alternative scenario, and Robert Corber, who served as a Commissioner of the ICC from 1975 to 1976, both testified persuasively that they would have followed that course.[204] None of this evidence focuses on the situations of particular Ts, but it places a very heavy burden on the Ts to demonstrate that it would have been an abuse of discretion for the ICC not to consider and approve a particular transaction before the others. That burden can be met only if, at a minimum, it is shown that, first, advance approval of the transaction would not have tied the ICC's hands with respect to other transactions in any important way, and second, the transaction would have been essentially unopposed. The first requirement, noted in the passage from the CUE Opinion quoted above, needs no explanation. The second is a consequence of the relatively short time limit we have assumed that Congress would have set on the ICC's deliberations. Before 1973, contested sale and abandonment transactions of any size rarely if ever received administrative approval in less than the twenty to twenty-four months we assume that the ICC would have had; the creditors of the Ts' estates could not reasonably have expected them to conclude sooner. Given its past record, any procedures used by the ICC that resolved large, contested sales and abandonments within this short a time would have been per se reasonable in our view. Only large transactions are in issue, however, for in accordance with the warning in our CUE Opinion, 439 F.Supp. at 1375 n.40, the Ts direct their early-approval arguments only to their entire systems or major portions thereof. We conclude that none of the T's has met this burden. EL does not attempt to show that it could have escaped consolidation of its case with PC's. None of the other T's has made the necessary showing that its disposition would have gone unopposed before the ICC. Abandonments would doubtless have been opposed by shippers on the relevant lines. See Hansen (S) (Nov. 1, 1979) at 7-8. Public bodies faced with having to acquire losing lines would likely have tried to persuade the ICC to include those lines in another system. See Stangl (S) (Nov. 1, 1979) at 17. And sales to solvents would have raised opposition from other solvents who had submitted competing bids, from solvents seeking traffic protection conditions, and from shippers seeking to condition the acquisition of profitable lines on acquiring lines slated for abandonment.[205] Apart from the general question of opposition to their respective dispositions, each of the Ts remaining in the case (except EL) presents its own specific arguments for early disposition. We discuss and reject the arguments of CNJ in Part XIII, infra. Here we find that the arguments of LV and AA are also without merit. LV argues that by June 1973 it was well on its way to cessation of its rail operations, and that the ICC would not have been likely to "stop LV in its tracks so that the other Transferors could catch up". LVRB at 17. LV also argues that because of its small size, its disposition would have posed few problems for the ICC. Id. Both arguments are apparently intended to apply only to LV's stand-alone case. See LV at 8. *1310 Although we have not discussed this case in detail, we think it founders for lack of identification of even one, let alone two, competitive bidders. If these contentions are meant to apply to an EL-LV acquisition, then they clearly lack merit. Neither LV's progress in its reorganization nor its size would have been relevant if its disposition were bound to EL's.[206] AA argues that its small size and relative geographic isolation would have rendered it unnecessary for the ICC to include it in any regional planning process. AA at 36-39. As the GPs correctly point out, however, 2 GPRB at U 22-23, the argument that AA's disposition would have preceded PC's is inconsistent with AA's own premise that the earnings value of the Toledo-Ann Arbor segment would have been established by the bidding of an acquirer of PC's IFLs. Moreover, AA has not demonstrated that its sale would have been unopposed. As AA strenuously argues, there would have been numerous bidders for its profitable Toledo-Ann Arbor segment. AA at 28. Any sale would thus likely have been opposed before the ICC by the unsuccessful bidders. The State of Michigan, faced with the prospect of acquiring AA's losing northern segment, would doubtless have tried to persuade the ICC to require the private purchaser of the Toledo-Ann Arbor segment to take the northern segment as well. As in the cases of the other Ts, the ICC would have taken the full reasonable time available to it to consider these objections. (c) Necessary Transactions Following Approval Conveyance of the Ts' properties in the alternative scenario would have been neither automatic nor immediate after ICC approval. Undoubtedly a few "tag ends" would remain to be tied up. Shannon (D) (May 24, 1979) at 575, EL App. K at 187.[207] The GPs contend that the transactions could not have been completed until at least eight months after they were approved by the ICC. 6 GP at T 147. EL's witness, Shannon, on the other hand, testified that two months would have been sufficient.[208] Our review of the evidence and arguments indicates that Shannon's estimate is more nearly accurate. The GPs' estimate of eight months rests on four assumptions, all of which we find flawed. First, the GPs contend that renegotiation resulting from conditions that the ICC would have imposed on the sales would have taken time. 6 GP at T 148. But we have already held that the ICC would have imposed no conditions of any significance. The GPs also argue that conveyance would have been delayed by petitions for reconsideration that opponents of the transactions would have filed with the ICC. The Commission, however, would have placed a strict time limit on the filing of petitions for reconsideration and could have been expected, in line with the congressional policy, to decide them very promptly. The GPs' third argument, and perhaps its best, is that opponents of the approved transactions would have sought and obtained judicial review of the ICC's actions, and that this would have resulted in significant delays. 6 GP at T 50. There is evidence, however, that approval by reorganization courts would have been almost immediate.[209] It is likely that any reviewing court, faced with both a clear congressional policy reflected in the deadline imposed on the ICC and the possibility of massive shutdowns, would *1311 have moved most expeditiously. Indeed, as in the case of the Rail Act's provision that this court must decide appeals from the initial decisions of the reorganization courts within 80 days, § 207(b)(1), Congress could well have placed severe time limits on the taking and decision of appeals. The GPs' final argument, that arrangement of credit terms would have delayed the process, 6 GP at T 151, is also without merit. The needed financing could and would have been assured before the sales agreements were concluded. C. Conclusion Arriving at a precise date on which the Ts' properties would finally have been conveyed in the alternative scenario is now a matter of summarizing the conclusions we have reached and performing arithmetic. The process of disposing of rail properties would have begun in February 1973. Studies by acquirers of the properties and negotiations would have taken ten to twelve months, and ICC deliberations would have taken twenty to twenty-four months. Taking into account the possibility of judicial review, the properties would have been conveyed about April 1, 1976. We recognize that in arriving at this date we have relied on crude estimates for each phase of the process. Our result necessarily reflects the roughness of its components; some of the periods may be too short, others too long. The conveyance of Ts' properties might have taken place somewhat earlier or somewhat later than April 1, 1976. But we are satisfied from the evidence that this is the most likely date, and that neither an earlier nor a later disposition is any more likely. In our CUE Opinion we held, tentatively, that conveyance would not have occurred before April 1, 1976. Today we reaffirm that holding, and hold further that it would have taken place no later. XI. Acquisitions by Public Bodies The JTTs allege that public bodies in New Jersey and Pennsylvania would have paid net salvage value to acquire their predominantly freight lines in the terminals areas and would have paid prices greater than net salvage value to purchase their predominantly passenger lines.[210] The GPs' assault on these contentions is multi-faceted. The GPs categorically reject the notion that the JTTs could have erected a firewall between their allegedly profitable main freight lines in the west and the losing public body lines in the east. In the GPs' view, public bodies would not have paid some $265.5 million for the predominantly freight lines within the JTs,[211] and another $265.1 million for the predominantly passenger lines,[212] when the continued operation of these lines was essential to the JTTs' realization of earnings values from the sale of the main freight lines. Moreover, the GPs allege that the acquisitions envisioned by the JTTs would have imposed an additional operating burden on the public bodies of $250 million annually.[213] In light of the difficulties faced by states in the Northeast in the mid-1970's, the assumption of such *1312 staggering costs would have been unlikely. Rather, public bodies would have used every means at their disposal to keep ownership of the "public body" lines in private hands. This would have involved attempts, including offers of financial subsidies if necessary, to induce the potential purchasers of the JTTs' main freight lines to acquire the eastern lines as well, and failing that, petitions to the ICC to condition the sales of the main freight lines on inclusion of the public body lines, a move to which the ICC would have responded favorably. Collapsing the firewall would have had the effect of substantially reducing the proceeds claimed by the JTTs from separate sales of their western and eastern lines. Assuming sales of the main freight lines based on earnings values, inclusion of break-even public body lines would not have affected the sales price, but would have deprived the JTTs of additional consideration from separate sales to public bodies. To the extent the public body lines were losing propositions, acquirers of the main freight lines would have paid less for the total package than for the main freight lines alone. In the alternative, the GPs argue that had the public bodies made these purchases they would have paid far less than the JTTs allege, and would have insisted on an allocation of capital and operating costs much less favorable to the purchases of the main freight lines than the JTTs envision. With respect to acquisition price, public bodies would have paid no more than an amount that would leave the T only slightly better off than it would have been had it pursued its best private marketplace alternative. This pricing principle entails rejecting premium prices for passenger lines, and, because of the dependence of main freight line revenues on the continued operation of the public body lines, requires deductions from the X of the latter in the event the main freight lines were to bring prices in excess of their X. Concerning the allocation of costs incurred in connection with the terminals area lines, public bodies would not have been so supine as the JTTs imagine. The switching charges set by the acquirers of the JTs would, for a number of reasons, have been much higher than those calculated by the JTTs and would have been determined according to a more equitable allocated cost approach. They also would have been calculated so as to recover capital costs associated with the acquisition and rehabilitation of the predominantly passenger and Northeast Corridor lines. In addition to these factual contentions, the GPs present a number of legal arguments against awarding values to the Ts based on hypothetical sales to public bodies.[214] First, the GPs maintain that the Ts' claims impermissibly assume changes in state law that were not reasonably foreseeable in February 1973. Second, they allege that the Ts mistakenly based certain of their claims on actions by the federal government that would have amounted to alternative, long-term federal responses to the rail crisis. Third, the GPs argue that federal condemnation law forbids compensating the Ts for the value to the public or social value of their properties. Fourth, the GPs contend that the Ts' cases offend the Supremacy Clause by relying in several respects on limitations on state power that do not bind the federal government. As a threshold question, the GPs hotly contested the JTTs' classification of terminal area lines into "predominantly freight" and "predominantly passenger" segments. In classifying their lines, the JTTs followed a rule of thumb dubbed somewhat derisively the "six-train rule" by the GPs. As described by the JTTs' witness: A line was deemed to be predominantly passenger if it was used for (i) six or more round trips per day of passenger trains regardless of freight density; or (ii) some passenger trains but less than six round trips per day of freight operation. All other lines were considered to be predominantly freight lines. Storm (S) (JT) (Dec. 15, 1978) at 7 n.1. In their briefs the GPs attack the six-train rule as designed to result in the smallest possible amount of trackage being designated *1313 predominantly freight. The JTTs' classification yielded a system in which, with few exceptions, every segment that carried any passenger traffic was designated a passenger line even though virtually all of these also carried some freight traffic, and some carried substantial amounts of freight traffic. Lines designated as predominantly freight were almost without exception those dedicated exclusively to freight service. See Tomlinson (S) (JT) (Dec. 1, 1978) at 12-13. Moreover, the GPs argue, the predominantly freight lines do not form a contiguous system. According to the GPs, the minimization of the predominantly freight trackage benefits the JTTs' case in three ways. First, because the JTTs allege that predominantly freight lines would have been acquired at net salvage value, while passenger lines would have brought a premium in excess of net salvage value, maximizing the passenger trackage would yield greater sales proceeds to the JTTs. Second, regardless of any passenger premiums, if the predominantly freight lines were included in the sales of the main freight lines to solvent railroads at no extra cost, while the predominantly passenger lines were sold to public bodies for net salvage value, maximizing the number of passenger lines would similarly increase the total proceeds from these sales. Third, reducing the number of predominantly freight lines minimizes the switching costs that the JTs would have charged back to the connecting line-haul carriers, thus inflating the earnings values of the main freight lines. We resolve the dispute over designation of lines by accepting most of the contentions of both sides. We agree with the JTTs that the existence of important commuter service distinguishes the predominantly passenger lines because only the public would have provided the necessary funding to keep the passenger operations running. See pp. 1317-1320, infra. The GPs' objections center upon the enhancement of value of the JTTs' lines as a whole that results from the JTTs' designations. As the GPs' counsel put it at oral argument, "it doesn't matter whether you call something a passenger line or a freight line, or you call it a rail line, or you call it a tangerine. The important point is the way that you allocate costs." Tr. at 303. We accept the GPs' objections in determining the prices that public bodies would have paid for the predominantly passenger lines, see pp. 1333-1342, infra, and the trackage rights fees public bodies would have charged for use of the passenger lines for freight movements, see pp. 1342-1346, infra. The GPs' arguments we accept as to the passenger lines are largely the same as those we accept as to the predominantly freight lines in the terminal areas. The analysis of these freight lines is more straightforward in that there is no need to consider complications introduced by the presence of passenger traffic on the same lines. Thus we begin our examination of the prospects for public acquisition by considering the JTTs' case for creation of public joint freight terminals. A. The Joint Freight Terminals Case Despite the great expense and the extensive studies the JTTs undertook in formulating the JTs case, their argument that public bodies would have acquired the terminal freight lines and set up JT companies is surprisingly weak. The JTTs do not claim that it would have been necessary for public bodies to have purchased the predominantly freight lines as a last resort, stepping in reluctantly to avert the disaster that cessation of freight service in the terminals areas would bring. On the contrary, the JTTs maintain that private purchasers of the main freight lines would have been willing to acquire the predominantly freight lines in the terminal areas as well had the public bodies declined to venture into the freight business. See 1 ELF at 57-59 (ELLV); 1 RF at 192-213. The JTTs' position is that, having solved their most vexing problem by acquiring the predominantly passenger lines, public bodies would have forged ahead and purchased the predominantly freight lines in order to create the JTs. The primary purpose of the public bodies in undertaking these projects would have been to promote industrial development by eliminating duplicative facilities and providing more efficient service to connecting carriers and local users. Combined *1314 ownership of the freight and passenger lines in the terminal areas would also have allowed better coordination of the two types of traffic, including reduced interference between freight and passenger trains and insured equality of access to freight service by industries in the area. Public purchase of all the Ts' terminal properties would have been more likely than piecemeal acquisitions by individual solvent railroads, since the acquiring roads would not have been able to achieve the economies made possible by a unified public acquisition and also would have been unwilling to risk being saddled with responsibility for passenger operations. See Hesse (S) (JT) (Dec. 15, 1979) at 29-30. There is a fundamental problem with the JTTs' argument. While the central purpose of the JTs was to eliminate duplicative facilities and services and to cut switching costs, the JTTs presented no evidence to show that the JTs' switching charges were lower than the switching costs that would have been encountered by prospective purchasers of their main freight lines in the absence of the JTs. In fact, evidently concerned that the GPs' assault on the JTs case might prove successful, the individual JTTs alleged alternative outcomes in which switching costs on their lines would have been lower and earnings values correspondingly higher without the "benefit" of the alleged economies achieved by establishment of the JT. See Statement of EL Trustees in Response to the Court's May 21, 1981 Memorandum (June 1, 1981) at 8-9, Sp. Ct. Rep. at N 35818-19; ELRB at 124-28; 1 RF at 192-213. EL suggested at oral argument that the NJ/NYTerm's switching charge might have been advantageous to purchasers of the lines of some of the Ts, such as PC, but not to purchasers of others, such as EL or EL-LV. See Tr. at 54-55. And the fact that the JTTs themselves never even attempted to form a JT in the real world suggests that the prospects for effecting major savings while preserving the status quo of traffic distribution had never been thought very good. Beyond this, we do not think that a solvent railroad interested in buying a T's main freight line would have been reluctant to purchase the T's predominantly freight lines as well. These were not segments that the solvents could afford to ignore.[215] Without the continued operation of the terminal lines, the sales of the main freight lines would not have been possible. See Hesse (D) (May 2, 1979) at 132, 11 GP Supp. App. at 180. The inability of individual purchasers to achieve the economies alleged to be possible in a JT would at best have affected only the price a solvent railroad would have been willing to pay for the package and not its willingness to buy. Indeed, prospective purchasers of EL and LV would have had reason to regard the creation of the JTs with some apprehension. Unlike publicly owned terminal railroads in other cities, which are rather local in extent, the terminals proposed by the JTTs cover vast areas.[216] In New Jersey, for example, the terminal includes LV's main line, spanning 76 miles from Jersey City in the east to Phillipsburg in the west, R's line running 27 miles from West Trenton north to Bound Brook, and 42 miles of the old Lackawanna line from Dover to Phillipsburg, much of which lies in areas of the State far from the New York port and which no one, prior to this litigation, ever would have conceived of as forming part of a terminal railroad. We are also incredulous that a Western purchaser bent on creating a transcontinental system would have been content to stop at the western border of New Jersey, leaving control of its entry to New York in the hands of a public agency that lacked any experience in freight service. A purchaser's apprehension at this state of affairs would only have been increased *1315 by the ambitious design of the JTs experiment. Even if we were to assume that the public bodies would have followed the JTTs' detailed operating plans and costing decisions to the last letter, purchasers of the main freight lines would have had to consider the significant risk that savings achieved on paper would not have been realized in the yards. A prospective purchaser of the main freight lines would thus have to reckon with a host of unknowns, including the possibility that a public terminal's switching charge might have exceeded the costs the purchaser would incur by acquiring and operating the predominantly freight lines itself, or that existing traffic patterns in the terminal area would be disturbed to its detriment. Moreover, it is somewhat incongruous for the JTTs to argue that Western Railroads would have engaged in competitive bidding to acquire their main freight lines to the east in order to protect or promote their position but would have declined to compete in the terminals area where much of the traffic they desired originated. Even if a purchaser were agreeable to the creation of a JT, it would have had cause to fear that an aggressive competitor acquiring another main line would have insisted on purchasing some of the predominantly freight lines for itself in order to gain a competitive edge. For all these reasons, solvent railroads looking to acquire the main freight lines probably would have taken a cautious attitude towards the establishment of publicly owned JTs. None of this is to suggest that if the public bodies had insisted on acquiring the JTs properties, purchasers of the main freight lines would have absolutely refused to cooperate. We doubt, however, that public bodies would have been particularly interested. The chief reason cited by the JTTs why public bodies would have sought to acquire the JT lines is their desire to take advantage of the unique opportunity provided by the rail crisis to rationalize and upgrade terminal facilities and services, thereby improving the climate for industrial development. As evidence of this, the JTTs cite several studies and reports recognizing the need for consolidation and modernization of rail facilities in the New York and Philadelphia port areas, see JT at 29-30 & n.19, a need which the GPs do not dispute. The JTTs, however, have not provided any credible evidence that New Jersey and Pennsylvania would have volunteered to undertake these grand projects in the circumstances of the alternative scenario. By contrast, the GPs rely on a bevy of public transportation officials who uniformly testified that public bodies had virtually no history of involvement in freight service,[217] that freight service had long been regarded as the responsibility of the private sector, and that acquisition would have been considered only if the alternative was abandonment *1316 of essential freight lines. See Altshuler (S) (Nov. 1, 1979) at 23; Simon (D) (Apr. 20, 1979) at 47, 20 GP Supp. App. at 50; Stangl (S) (Nov. 1, 1979) at 20-21; Stangl (D) (Mar. 19, 1980) at 198-99, 256, 19 GP Supp. App. at 728-30; Tennyson (D) (Mar. 19, 1980) at 38-39, 43-45, 20 GP Supp.App. at 714-18; Whitehurst (S) (Jan. 30, 1980) at 46-49. In the words of one New Jersey official, in 1974 the state was concerned about the possibility of being required to take a more direct role in the passenger business and the last thing we wanted to do or, in fact, were prepared to do, was get into the rail freight business. Stangl (S) (Nov. 1, 1979) at 20. Moreover, as developed below in our discussion of the financial feasibility of public acquisitions, public bodies' efforts to come up with the funds necessary to acquire and operate the passenger lines would hardly have been painless. While we do not say that they would have been unable in addition to finance a purchase of the predominantly freight lines if this were necessary to preserve freight traffic, it is rather plain that public bodies would not have been willing to spend the amounts sought by the JTTs solely to attain new heights of freight service. The only immediate beneficiaries of such largesse would have been the bankrupts, which would have raised their asking prices for the main freight lines to the extent that the costs of terminal services were reduced and also would have realized additional amounts from the separate sales of the terminal lines to public bodies. The public benefits of more efficient terminal services, if realized at all, would doubtless have taken years to emerge. This leaves only the argument that combined public ownership of the freight and passenger lines would have had the advantage of increased possibilities of coordination of rail service. Contrary to the JTTs' view, public bodies would have had no need to acquire the terminal lines to reduce interference between freight and passenger trains, since virtually all of this would have occurred on the passenger lines they are already presumed to control. Insuring equal access to freight service is hardly a sufficient incentive to justify an investment of hundreds of millions of dollars. We conclude that public bodies would have resisted buying the predominantly freight lines and that these would have been acquired instead by the purchasers of the main freight lines.[218] We thus do not reach the complex issue of the reasonableness of the switching charge computed by the JTTs. B. Predominantly Passenger Lines in Northern New Jersey We find that EL's passenger lines in the North Jersey area would have been publicly acquired in the absence of the Rail Act.[219] In reaching that conclusion, we consider the strength of public bodies' motivation in the alternative scenario and their legal and financial capacity to consummate the acquisitions. After resolving the likely disposition, we face the more complex questions of what price would have been paid for EL's passenger lines and what trackage rights fees would have been charged for freight movements over the publicly acquired passenger lines. 1. Legal Capacity As a threshold question the GPs contend that, since none of the states involved possessed the legal authority to acquire railroads and the requisite changes in the law were unforeseeable, acquisition by public bodies may not be considered in determining the value of the Ts' property.[220]*1317 The legal rule cited by the GPs is applied when a condemnee's valuation is premised upon a hypothesized private purchaser that would have put the property to a new use forbidden by law. But, for example, [w]here ... there is a possibility or probability that the zoning restriction may in the near future be repealed or amended so as to permit the use in question, such likelihood may be considered if [and to the extent that] the prospect of such repeal or amendment is sufficiently likely as to have an appreciable influence upon present market value. 4 Nichols, Eminent Domain § 12.322[1] (3d ed. 1979) (footnote with citations omitted). In this case, however, the potential buyer is the very body that makes the law. The absence of preexisting authority cannot, therefore, be said to diminish the prospect of public acquisition since a state's decision to purchase portions of the Ts' railroads would be embodied in legislative or constitutional enactments providing the necessary authority.[221] Since we reject out of hand the applicability to this case of the legal principle cited by the GPs, we will forego examination of various Ts' responses to the factual predicate of this argument. Thus, the only two relevant questions are the ones we turn to next: whether New Jersey would have been sufficiently motivated to acquire the Ts' passenger operations in the absence of the Rail Act[222] and whether New Jersey would have had the financial capacity to complete the transaction. 2. Public Acquisition The GPs' further response to the Ts' passenger cases is a variant of their general argument that, so long as substantial profits were projected on the main freight lines, public bodies would have refused to buy any lines on which those profits were dependent. Just as public bodies would have opposed the erection of an artificial firewall between the main freight lines and the predominantly freight lines in the JTs, so would they have resisted the establishment of an "inner" firewall between privately owned, profitable freight lines and publicly owned, losing passenger lines. Rather than acquiesce in such "cream-skimming", public bodies would have sought to persuade the private purchasers of the Ts' main freight lines to buy and operate the passenger properties as well. If necessary, public bodies would have offered to raise their passenger service subsidies above the avoidable cost level. If that effort failed, states would have proceeded to petition the ICC to condition approval of a sale of the main freight lines on inclusion of Ts' passenger properties. See pp. 1306-1316, supra. If *1318 all these efforts proved unsuccessful, public bodies would have been forced by lack of funds and the political infeasibility of increased spending for public transportation to acquiesce in the abandonment of the passenger lines. Given the alternatives of increased bus and automotive transportation, the consequences of a shutdown of rail passenger service would not have been so catastrophic as to foreclose it as an option. Finally, the GPs contend that even if public bodies did agree to buy some of the Ts' passenger lines, they would not have agreed to buy them all. We shall address each of these contentions in some detail. In this subsection we consider how strong New Jersey's interest in acquiring EL's passenger lines would have been in the absence of the Rail Act. EL stands on rather firm ground with respect to public acquisition of its passenger lines. EL emphasizes that no solvent railroad contemplating a purchase of its main freight lines or of the EL-LV package would have been willing, or could have been forced, to go through with the deal if responsibility for the passenger service was included, see Shannon (S) (EL) (Dec. 31, 1978) at 91, because passenger service had been a losing proposition for many years and a substantial drain on the bankrupts providing it. Although public bodies were heavily involved in subsidizing passenger service, they had insisted on paying no more, and in some cases less, than the avoidable costs of passenger operations.[223] Under this approach, common costs such as track maintenance were allocated to the "base" or freight service. See Whitehurst (S) (Jan. 30, 1980) at 7. EL's witness Shannon bluntly stated: [R]unning passenger service is a pain. It's politically oriented. I've been in the passenger business for almost 25 years and ran suburban service and nobody in their right mind will buy it, except the public. Shannon (D) (May 23, 1979) at 389-90, ELP App. at 79-80. When pressed, Shannon did state that a private purchaser theoretically would have been willing to assume responsibility for conducting passenger operations if guaranteed 100 percent subsidization of operating deficits plus a fair return on investment. He thought that under those conditions it would have been cheaper for the State to buy and operate the passenger service itself. Shannon (D) (May 23, 1979) at 381-86, ELP App. 87-93. That seems most unlikely since, even if operating costs would have been lower under public ownership, the public would have saved the full acquisition price if it could have succeeded in having the lines included in the sales of the main freight lines without affecting the price in the private sale. But there are other reasons to believe that the passenger lines would have been disposed of differently from the freight lines. Potential acquirers of the freight lines would have been reluctant to become involved in passenger service even indirectly, for there was no likelihood that any public body would make, or could persuasively claim to make, a guarantee of 100 percent subsidization for the indefinite future. Further, since the demands of scheduling passenger service are greater than those for freight, a public body would have seen some advantage in owning the passenger lines so as to minimize scheduling conflicts. Finally, if New Jersey had not been willing to purchase EL's passenger lines, at least some of them would have been abandoned since they would net more to EL as scrap than if they were included in a freight package. This is not to deny that New Jersey would have been reluctant to purchase and operate the passenger lines or that the State would have attempted instead to persuade the solvent railroads to accommodate it. It is simply to say that the State's arguments in this regard would have lacked force. The weakness of the State's position is reflected in the evidence on this point. While the GPs produced several witnesses *1319 who testified that public bodies would have tried to negotiate a private purchase of the passenger lines, none was willing to say that such efforts would have been likely to achieve their goal. By contrast, EL's witnesses emphatically denied that any solvent railroad could have been persuaded to enter the passenger business. This testimony went without significant contradiction, and is corroborated to some extent by the fact that in connection with the FSP proposal that Chessie buy EL and R, "it became clear that as a condition of purchasing these properties, Chessie did not want to assume the obligation of negotiations with the commuter authorities." I FSP at 46 n.5. As a result, USRA recommended that Chessie not acquire any passenger lines but rather obtain trackage rights over them for freight service. Id. at 232-33. We think that New Jersey, unlike a private acquirer, would have been willing to purchase EL's passenger lines. These lines, including about 165 miles running through some of Northern New Jersey's most heavily populated bedroom communities, see EL Map 1, carried more than 35,000 passengers in each direction on an average weekday. The lines ran to Hoboken and Newark, and connected to the PATH rapid transit system to Manhattan at Hoboken. New Jersey had demonstrated its financial commitment to EL's passenger service over the years in the form of operating assistance payments, capital improvements, and various kinds of tax relief. The size of these operating assistance payments had been rising dramatically in the 1970's — $5.2 million in 1970, $9.1 million in 1973, and $17.9 million in 1975. See Ex. ELR-NJ-1 at 3, ELP App. at 16. While some of the publicly financed capital improvements had been paid for by federal as well as State funds, others were shouldered exclusively by the State, as evidenced by the purchase between 1967 and 1972 of 155 passenger cars and 32 locomotives at an approximate cost of $41 million.[224] See Ex. PCT-NJ-48 at A 17, ELP App. at 23. We find it hard to believe that New Jersey could have seriously contemplated abandonment of the State's passenger lines. See Ex. PCT-NJ-13 at 1-2 (Transportation Commissioner Kohl) ("40 lanes of expressways would be necessary to provide equivalent travel during rush hours," which would entail "huge cost [and] would be socially and environmentally unacceptable"); Ex. PCT-NJ-114 at 5-7 (Commissioner Kohl) (infeasibility of transferring rail commuters to buses and automobiles); Ex. ELR-NJ-6 at vii (New Jersey Commission on State Tax Policy) ("great public cost, if not impossibility, of providing equivalent highway transportation facilities"). Apart from the State's need to purchase its lines to keep the passenger trains running, New Jersey would have some affirmative motivation to make the acquisition in order to achieve the financial and operational control that comes with ownership. This advantage was becoming increasingly clear to State transportation officials during the 1970's as the cost of subsidizing private carriers rose while service declined, and was in fact cited by the State in 1978 when it acquired the major portion of EL's passenger lines. See ELP at 19-27. Unless the State took this next logical step to ownership, it would have had to face the prospect of continuing to sink huge sums into capital improvements and to pay out ever larger subsidies to an entity which had no incentive to improve service or achieve economies and over which the State had no legal control. In all, EL's commuter service was too important to the State for abandonment to be considered a realistic option. New Jersey could not have convinced solvent railroads or EL to include losing passenger lines in the sale along with EL's freight lines. Under these circumstances, New Jersey would have had no choice but to acquire EL's passenger lines. The Ts' cases for *1320 public acquisition of their predominantly passenger lines are not immune, however, from our criticisms of the JT case since these lines did carry freight traffic. But arguments that were decisive against public acquisition in the case of the terminal freight lines have an impact in the case of the passenger lines limited to the prices public bodies would have paid and the trackage rights fees they would have charged. 3. Financial Capacity The GPs allege a second barrier to New Jersey's purchase of EL's passenger lines for continued rail use: the inability of the State to finance the acquisition.[225] Most of their analysis, however, purports to be offered in support of the more limited proposition that "the financial constraints on public bodies would have reinforced their determination to pay no more than the minimum for any lines".[226] We find this latter proposition persuasive; in fact, it was conceded by EL's only witness on the issue of New Jersey's financial capacity. Cahill (S) (EL) (Oct. 15, 1980) at 10. But the GPs' broader contention fails. Their reply brief properly identifies the question. After conceding that virtually no single undertaking would be impossible financially, it states "the financial ability issue ... cannot realistically be divorced from the practical ability and political willingness of public bodies to cut existing programs and impose new taxes", 2 GPRB at K 17. Yet the GPs' argument fails to meet this issue, since it analyzes the alleged historic unwillingness of New Jersey to commit massive funding to public transportation projects as though the rail crisis would have had no impact upon its financial decisions. Examination of New Jersey's financial capacity must begin with reference to the cost of taking over the passenger lines. Based upon the Ts' allegations, the GPs indicate that the total acquisition cost for all of the proffered public body lines in New Jersey would have been approximately $518 million.[227] 4 GP at P 132 (Chart P-1: "New Jersey as Fairy Godmother"). From this must be subtracted the costs of the JT freight properties, leaving a total of some $363 million — a figure including $109 million for the Northeast Corridor Spine and $100 million claimed by CNJ for its freight and passenger lines. This figure is an upper limit since it assumes that the Ts will prevail on all their contentions relating to prices that public bodies would have paid. Since we reject EL's claim for a premium above X, see pp. 1324-1333, infra, and accept the GPs' argument for deductions from X, see pp. 1333-1341, infra, it is evident that the $363 million figure substantially overstates the actual financial requirement. And to the extent that any of the GPs' claims in the nonrail use phase of the case prevail, the burden would have been even less. An additional financial burden would have resulted from the State's assumption of all operating costs of the proffered lines.[228] The GPs' estimates place the annual *1321 burden, above preexisting commuter subsidies, which amounted to $30 million in 1974, at nearly $110 million. 4 GP at P 135. For purposes of determining the State's ability to finance the takeover of passenger lines, the figures are overstated in that nearly half of the alleged increase is to cover the cost of freight operations in the NJ/NYTerm. See Aaron (S) (Jan. 30, 1980) App. J at Table 2. Thus the $30 million subsidy already being expended in 1974 was a substantial fraction of the alleged total, and, because the subsidy had been increasing, see p. 1319, supra, the amount by which the total burden would have exceeded subsidy levels absent public acquisition would have been even less in future years.[229] The GPs also present evidence from First Boston indicating the hypothesized increases in New Jersey's sales, corporate income, and motor fuels taxes that would have been necessary to finance $114 million in annual operating burdens in an attempt to put their operating cost estimates in perspective. First Boston's figures, present a somewhat distorted figure because they are based on the assumption that only one revenue source would shoulder the full burden, First Boston (S) (Jan. 30, 1980) at 42, giving an impression that in effect triple-counts the total cost.[230] Their budget figures, id. at 44-45, repeat this illusion by implicitly assuming that the entire burden would be met by budget cuts concentrated in a single area.[231] If, more realistically, this burden were spread among tax increases and budget cuts in several areas, it seems clear that the burden would be bearable in light of the public necessity. The GPs correctly note that the mid-1970's were a time of tight money, when numerous demands competed for limited State funds. 4 GP at P 69-87; 2 GPRB at K 31 & n.48. The GPs argue, for example, that New Jersey's historically low government expenditures reinforce the conclusion that the State would have been inclined to spend as little as possible. 4 GP at P 145 (indicating, for example, that New Jersey ranked 49th among the states in per capita government expenditures per dollar of disposable income). But this also cuts the other way, because, if necessary, New Jersey was more capable of assuming an additional financial burden than were most other states. See Kaden (D) (July 22, 1980) at 241-42, ELP App. at 234-35 ("By comparison [to] New York, ... New Jersey has today almost an infinite amount of fiscal capacity."). We also discount as self-serving some rather extreme statements indicating lack of financial ability made by state officials before Congress while it was considering rail legislation, see 4 GP at P 77-79. We agree with the GPs that "public bodies would have had the strongest possible economic motivation to adopt strategies to minimize any new financial commitment to rail transportation, and to pay the absolute minimum for any properties they did buy", id. at 79, so attempts by state officials to convince Congress to shoulder the burden were to be expected. But we are not convinced that permitting abandonment of the *1322 passenger lines would have been part of such a strategy. EL identified several sources of funds available to New Jersey. Debt finance would appear to be one of the more likely. The GPs note that all three major transportation bond issues proposed in New Jersey in the 1970's were defeated at the polls, 4 GP at P 149, including a 1974 proposal for $100 million for the preservation of rail lines that might have been excluded from Conrail or, if the Rail Act was declared unconstitutional, for other lines that might have been abandoned by the bankrupts. Here the GPs are on weak footing. That proposal failed by less than a three percent margin.[232] EL argues that defeat may well have been due to a failure to explain the proposal, the credibility gap created by the continuing failure of the State transportation department to spend public transportation proceeds remaining from a 1968 transportation bond issue, and guilt by association with the accompanying $200 million highway construction bond program that was defeated by a substantial margin. ELP at 47 n.41; Cahill (S) (Oct. 15, 1980) at 8. In addition, in the absence of the Rail Act the prospect of abandonment of these lines would have appeared a far more realistic possibility than it did at the time of the actual 1974 bond proposal. The GPs' primary response is to note that the legislative findings made at the time of this bond issue were similar to those EL hypothesizes would have been made in the absence of the Rail Act. 4 GP at P 149 & n.60; 2 GPRB at K 34. This point is nonresponsive to some of EL's allegations and is simply unpersuasive on the facts. The existence of a "whereas" clause in the legislative act referring to loss of essential services, especially when coupled with references to federal support under the Rail Act, is simply no substitute for the facts as they would have been perceived by the public in the absence of the Rail Act, when abandonment of the spine, EL commuter lines, and additional properties otherwise would have been imminent.[233] Much of the GPs' attack on other suggested revenue sources is based upon the conclusions of First Boston relating to the impact of financing acquisition with general obligation bonds. EL claims that First Boston considered a narrow range of alternatives, whereas in another study First Boston had noted the existence of many more. ELP at 32 & n.27. The GPs reply that general obligation bonds were the focus of the inquiry because they have the minimum impact of any funding source — which they claim was one of the conclusions of First Boston's earlier study that had considered other options, 2 GPRB at K 21-22. Although persuasive, this cannot substitute for a more detailed investigation of alternatives. One might well conclude that general obligation bonds would have been the State's first choice. But that does not rule out all other possibilities in the event that New Jersey deemed it wise to spread the cost of the acquisition among a variety of funding mechanisms.[234] EL also alleges that some funds could have been diverted from approximately $100 million that remained unspent from the previously noted 1968 transportation bond issue. The GPs' first reply is that "it would have made no sense" to have used the funds in this manner because nothing would have remained to make improvements that had been promised to voters for years. 2 GPRB at K 35; see also 6 GP at U *1323 64-65. This is a "patently ridiculous" reply, since as EL points out, if the lines were not purchased there would have been no railroads left to improve. ELP at 45-46. Of course, the funds were needed for buses as well, 2 GPRB at K 37, and some were eventually spent there, Cahill (S) (Oct. 15, 1980) at 7, but in the absence of the Rail Act it seems likely that funds originally intended for rail improvement would have been available if necessary to finance rail passenger acquisition rather than being fully diverted to finance the acquisition of new buses. It might appear that no income tax revenue would have been available to finance the purchase of the passenger lines, given that it took a number of years for New Jersey to enact its state income tax in response to the added financial requirements for the state school system after Robinson v. Cahill, 62 N.J. 473, 303 A.2d 273, reargued, 63 N.J. 196, 306 A.2d 65, cert. denied, 414 U.S. 976, 94 S.Ct. 292, 38 L.Ed.2d 219 (1973); see 4 GP at P 145-46. Yet former Governor Cahill's remarks, Cahill (S) (Oct. 15, 1980) at 6, that there was no real crisis in the interim years and that the income tax did pass shortly after the court ordered the schools closed in the summer of 1976 seem quite persuasive. The question here is only whether the state income tax would have been available to help avert a crisis in the absence of the Rail Act, and again there is reason to believe that it would. Id. at 6-7. EL also makes reference to legislation introduced in 1973 to consolidate the State's toll roads, under which revenues made available from increased tolls could supposedly have financed between $550 and $800 million in new bonds. ELP at 39-43. EL asserts that this specific proposal is one way that acquisition could be financed with less credit impact than general obligation debt. Id. at 42 n.39. The GPs respond that this was one of many pieces of transportation legislation not enacted at the time and that the legislation eventually enacted in 1979 did not link this revenue source to the acquisition of transportation facilities. 2 GPRB at K 30 & n.47. Of course, in 1979 the State did not face the rail crisis it might have faced years earlier in the absence of the Rail Act. The GPs are more persuasive in noting that the toll consolidation proposal "is just another way to raise taxes", 2 GPRB at K 31, although it is not necessarily the case that opposition to increased tolls would have been as strong as that to imposing the State's new income tax at a higher level to cover the costs of acquiring rail lines.[235] A final major funding source relied upon by EL is the use of federal funds, including the application of funds from the Urban Mass Transportation Act (UMTA), Federal-Aid Highway Act (FAHA) and, in particular, the Federal Aid-Urban System program (FAUS) under the FAHA, and revenue sharing. As a threshold argument, the GPs contend that consideration of this option amounts to allowing the Ts to benefit impermissibly from an alternative long-term federal response to the rail crisis. We reject the GPs' objection because the federal funds at issue are not part of a federal response directed at the rail crisis. These programs were all in place prior to the Rail Act. Furthermore, the scope of each program extends far beyond rail services formerly conducted by private railroads. If EL depended upon increased funding of these programs resulting from the rail crisis we would reject their claim, but we do not understand it to rely upon any such increases. The GPs also would have us reject this option because of the need for project approval by federal officials. 1 GPRB at B 23 n.28. Yet such approval, to the extent it is consistent with preexisting legislative mandates, cannot be cast aside as an alternative federal response any more than various actions of the ICC and Congress posited by the GPs both in this[236] and earlier[237] phases of the case can be. *1324 EL offered evidence indicating that New Jersey would have access to $400 million in UMTA funds and that there were $80 million of unobligated FAUS monies, the latter amount increasing at a rate of $50 million annually. ELP at 55-56. While acquisition of existing properties was accorded a low priority under UMTA policy, in the context of possible abandonments those priorities probably would have changed dramatically. This is not to deny that the limited federal money was substantially less than New Jersey's transportation funding needs at the time. See, e.g., Kaden (D) (July 23, 1980) Ex. 39 at 6. Moreover, as the GPs note, because these would not be a new source of funds, their use would be viewed by the legislature as similar to using other State revenues. 2 GPRB at K 27-28; see Tennyson (D) (March 19, 1980) at 116, 31 GP Supp. App. at 677. The difficulty in using funds from the highway acts to finance acquisition of rail lines, see 2 GPRB at K 25-26 & nn.37-38, supports essentially the same proposition. But once again the GPs have failed on the ultimate issue. All the Ts need to establish as to federal funding sources is that in the absence of the Rail Act the State could have marshalled sufficient resources to meet what would have been one of its highest priorities. The availability of federal funds undoubtedly would have assisted New Jersey in purchasing the Ts' properties. The only significant uncertainty is which sources in fact would have financed New Jersey's takeover of the passenger lines in the absence of the Rail Act. That is an issue we need not decide. There can be little question that New Jersey would have taken the concededly difficult and painful course of sacrificing some expenditures on other programs and increasing the overall financial burden on its citizens to prevent the rail crisis from becoming even more unpleasant. 4. Pricing We find that the prices public bodies would have been willing to pay in the alternative scenario would have been negotiated with reference to the value of EL's next best alternative. The GPs devote substantial efforts to establishing this pricing principle, see, e.g., 4 GP at P 7-10, 66-242, and note that it was endorsed as the likely course by testifying public officials from all the affected states, 4 GP at P 10-11; 2 GPRB at K 10-13; Cahill (S) (Oct. 15, 1980) at 10, evidence that is most probative since the relevance of pricing theories is dependent upon the extent to which "particular prospective buyers would in fact have given weight to them", CMV Opinion, 445 F.Supp. at 1047. But the dispute for the most part is not so much over the principle itself as over its application to the alternative scenario, see, e.g., ELRB at 287-88. For the passenger lines, the Ts concede that the next best alternative would have been liquidating the property for nonrail use. Thus, most pricing arguments begin with a base price of X, and argue for premiums and deductions from this as yet undetermined figure on the basis of the value of the hypothesized sales to and the relative bargaining positions of the Ts and the public bodies. (a) Erie Lackawanna's Claim for a Premium in Excess of X EL suggests that New Jersey would have paid in excess of X in order to avoid having to condemn EL's lines. One proffered motivation is that New Jersey would have sought to avoid the litigation costs incident to condemnation proceedings. This argument ignores what EL's own witness admits: similar costs would have been entailed for the condemnee. Shannon (D) (May 23, 1979) at 472-73, 18 GP Supp.App. at 270-71; Netzer (S) (Nov. 1, 1979) at 33-34. As a result, this consideration has a neutral effect in that it gives both parties the same incentive to avoid condemnation. Altogether different is EL's conjecture that New Jersey would fear a condemnation award in excess of X. Specifically, EL argues that a New Jersey court would award not merely salvage value for its passenger lines — approximately $104 million according to EL — but an additional "premium" *1325 consisting of $4.9 million for going concern value[238] and $5.7 million for the difference between gross and net liquidation value as well, Shannon (S) (EL) (Dec. 31, 1978) at 92 — or, at least, that New Jersey would have had sufficient fear of this that it would have paid the relatively modest premium claimed by EL. We see no basis for thinking that New Jersey would have feared an award based on gross liquidation value. Shannon conceded he knew of no case awarding both "gross liquidation value plus a going concern value premium." Shannon (D) (May 24, 1979) at 606-07, 18 GP Supp.App. at 284-85. In fact, we have been unable to discern the reasoning upon which EL and Shannon rely for the portion of EL's claim for the excess of gross over net liquidation value. If the argument is, as we suppose it to be — that since the condemnor would not dismantle, the cost of dismantling should not be taken into account — it runs afoul of the principle that the condemnor need pay only the condemnee's next best alternative, to wit, dismantling with its attendant costs. EL's claim for a premium based upon the going concern value of its passenger lines is based upon two New York precedents, In re Port Authority Trans-Hudson Corp. (PATH), 20 N.Y.2d 457, 285 N.Y.S.2d 24, 231 N.E.2d 734 (1967), cert. denied, 390 U.S. 1002, 88 S.Ct. 1244, 20 L.Ed.2d 103 (1968), and In re Fifth Avenue Coach Lines, Inc., 18 N.Y.2d 212, 273 N.Y.S.2d 52, 219 N.E.2d 410 (1966), appeal dismissed, 386 U.S. 778, 87 S.Ct. 1480, 18 L.Ed.2d 524 (1967). Fifth Avenue Coach is readily distinguishable since, there, going concern value was awarded because the appellate court found that only the condemnor's stringent rate regulation had prevented the condemnee from operating at a profit. No such circumstances are present here. By contrast, PATH permitted augmentation of the award based upon going concern value in the case of an unambiguously unprofitable enterprise, although the basis for the court's generosity remains somewhat of a mystery. The only apparent explanation is that the motivation for the court's approval of valuation based upon original cost less depreciation — to avoid an award reflecting an excessively low scrap value (zero in the case of the tunnel properties) — was extended to permit a going concern value premium. Here X will presumably be substantial.[239] The GPs' arguments indicating that earnings values from the EL's main freight lines are in part dependent upon continuing its passenger lines in rail use, see pp. 1333-1334, infra, further distinguish PATH. Quite apart from these distinctions we would be hesitant to conjecture a condemnation premium based on these New York cases. We have previously expressed our disagreement with Fifth Avenue Coach, CMV Opinion, 445 F.Supp. at 1022, as well as with the going concern value holding of PATH, CMV Opinion, 445 F.Supp. at 1037 n.54, and indicated how Justice Cardozo's opinion in Roberts v. New York, 295 U.S. 264, 55 S.Ct. 689, 79 L.Ed. 1429 (1935), relied upon by both the majority and dissent in PATH, was unhelpful in assessing its merits, CMV Opinion, 445 F.Supp. at 1028-29. Rejecting any going concern value premium above a base award of X for a losing operation seems most in accord with previous pronouncements on the question. This is true even when the claim for going concern value is not a percentage of the physical values, a position rejected in PATH itself, 20 N.Y.2d at 471-72, 285 N.Y.S.2d 24, 231 N.E.2d 734, see also Galveston Electric Co. v. City of Galveston, 258 U.S. 388, 393-96, 42 S.Ct. 351, 354-55, 66 L.Ed. 678 (1922) (claim of confiscatory rate), but costs of the sort claimed by Shannon that would in fact *1326 be incurred in inaugurating operations. The reason is the same as that for rejecting gross liquidation value. The condemnee is entitled only to the value of its next best alternative, to wit, sale for scrap; trained personnel, operating records, etc., would have had no value in such a sale. As said in the New Haven Inclusion Cases, 399 U.S. 392, 482, 90 S.Ct. 2054, 2104, 26 L.Ed.2d 691 (1970), a hopelessly losing railroad cannot be treated "as a liquidating enterprise with respect to certain items and as an operating railroad with respect to others, depending on which happens to yield the higher value." See also Kimball Laundry Co. v. United States, 338 U.S. 1, 9, 69 S.Ct. 1434, 1439, 93 L.Ed. 1765 (1949); New York, New Haven & Hartford R.R. v. United States, 289 F.Supp. 418, 435 (1968). This line of decisions is particularly important because a New Jersey court would not be bound by the New York decisions. While technically the New York Court of Appeals in PATH applied New Jersey law to the condemned property in New Jersey, all the court had to say on that score was that New Jersey law was similar to the general rule followed in New York and in the federal jurisdiction and that the case before it was sui generis, rendering it impossible to make with certainty any statements predicting the reaction of the New Jersey Supreme Court. No New Jersey cases following PATH have been cited to us and we are not aware of any. Thus, there is no good reason to take the going concern value portion of the PATH opinion as indicative of what New Jersey courts would decide New Jersey's own law to be in a condemnation of the Ts' passenger lines. And there can be little doubt that public officials would not have paid such premiums by way of settlement for fear that a New Jersey court would have followed the going concern value holding of PATH. See, e.g., Stangl (S) (Nov. 1, 1979) at 28-29; Stangl (D) (Mar. 25, 1980) at 419-20 (New Jersey), 19 GP Supp.App. at 743-44; Tennyson (S) (Nov. 1, 1979) at 5, 9-10 (Pennsylvania); Schuler (D) (Apr. 29, 1980) at 253-54 (New York), 19 GP Supp.App. at 786-87.[240] We thus need not rule on the GPs' argument that for this court to recognize a premium based upon a state court's decisional law that is in conflict with federal condemnation law would violate principles of federal supremacy. Finally, EL refers to some past public acquisitions of railroad property as a basis for its claim that New Jersey would have paid some premium over X. Rather than present any substantial testimony concerning these transactions on its own, EL relied almost exclusively upon evidence offered by PC, and advanced by R after PC settled, in support of its claim for premiums 50 percent in excess of X for publicly acquired passenger lines. See ELP at 67 n.60 (LIRR) (citing Ronan and noting "other transactions referred to by him"); ELRB at 28 & n.24 (Rutland Ry.), citing 3 CRA (S) (Dec. 1, 1978) XIV at 1-45. This evidence includes a massive study performed for PC by the Charles River Associates (CRA) of historical transactions involving the acquisition of rail properties by public authorities. This was supplemented by the testimony of William Ronan, who, beginning in 1959, held a variety of high transportation positions, mostly in the State of New York. Although we agree with the Ts that information concerning past public acquisitions, properly applied, can provide useful guidance in predicting actions of public bodies in the alternative scenario, we find the Ts' evidence *1327 gives no support to their claims for premiums. The CRA Study examined fifteen negotiated acquisitions by nonfederal public bodies of rail properties for continued rail use. The study calculated the consideration received by the railroad in each transaction and compared it with independently determined measures of "common basis".[241] Implicitly, the Ts ask that we equate the latter measure with X for the purposes of this litigation, with the result that the ratio of consideration to common basis would represent premiums actually paid.[242] "Consideration" was defined to include all benefits to the railroad that were intended as part of the agreed compensation for the conveyed property; thus it was not restricted to the cash actually paid, but included relief from operating deficits, tax relief, trackage rights, and any other concessions made by the public body. In 9 of the 15 transactions, consideration thus defined exceeded common basis.[243] The GPs criticized this study at great length, see 9 GP at App. X 1-138; 2 GPRB at N 20-29. We find it unnecessary to consider all the GPs' criticisms because we find a number of them sufficient to reject the study outright. The GPs' analysis of the CRA Study falls into two categories: general methodology and reliability of execution. Defects identified in the latter category as a result of the GPs' devastating cross-examination of CRA witnesses are alone sufficient to destroy all credibility of CRA's evidence. These comprised substantial inadequacies in document collection[244] and failure to interview most of those involved in the transactions they studied.[245] The latter is particularly important given that CRA's methodology was to determine the value of benefits intended as compensation as perceived by the negotiating parties at the time of the transactions. It also appears that their investigative technique was at times most haphazard.[246] Even more troubling is that when conflicting information was received, that unfavorable *1328 to the Ts' desired conclusions was sometimes ignored.[247] CRA also relied upon doubtful legal conclusions wholly beyond their expertise.[248] A substantial portion of the consideration received in the transactions as measured by CRA was the cash flow gain to the railroad due to relief from future operating losses.[249] This was calculated by taking the gains — measured by operating loss, or amount received by the railroad in the transaction under service contracts — typically for the year of the transaction and multiplying by ten to reflect the assumption that absent the transaction the operating burden would have continued well into the future. Even if we were to accept this approach, CRA's computations under it are seriously flawed. In one case, the benefit was determined by calculating the avoided costs of passenger service without subtracting the revenues derived from that service. See Lamond (D) (June 7, 1979) at 87-89 (PRSL), 13 GP Supp.App. at 356-58; 2 CRA (S) (Dec. 1, 1978) at VII 68. When service contracts were part of the transaction, CRA did not deduct the cost to the railroad of providing the service. On at least one occasion, the base figure, which in turn was inflated by the multiplier of ten, was grossly in excess of that used contemporaneously by the transferring railroad.[250] In other transactions, CRA deemed past operating subsidies[251] and subsidies totally unrelated to the transaction[252] as part of the consideration. Support for CRA's use of a multiplier of ten is also highly dubious. For the multiplier to be that high, assuming a 6 percent interest rate, operating losses would have had to be expected to continue for at least sixteen years. CRA (S) (Dec. 1, 1978) at 54-56. Yet there is no apparent basis to assume that losses would have continued for so long. The public bodies acquired the properties only after the prospect of such deficits made discontinuation of passenger services a serious possibility. See Nathan (D) (Apr. 15, 1979) at 379-80, 15 GP Supp. App. at 215-16. Available evidence in a number of transactions indicates that losses would have continued for only a few years,[253] suggesting that a multiplier closer to two might have been in order. More disturbing is evidence suggesting that the multiplier of ten was not merely an arbitrary choice but rather a figure derived after the fact to ensure that CRA's study reached conclusions favorable to the Ts. It appears that CRA had originally used a multiplier of 4.31, based upon the assumption that operating losses in the absence of the transaction would have had a five year duration. It was admitted that the decision to move to a factor of ten might have been made after CRA began receiving the common basis numbers against which their consideration figures were to be compared. J. *1329 Boyd (D) (June 7, 1979) at 376 (Methodology), 9 GP Supp.App. at 165. CRA did not indicate any new information upon which this change was based, see J. Boyd (D) (June 19, 20, 1979) at 283-84 (West End), 446-47 (Harlem-Hudson), 9 GP Supp.App. at 220-21, 238-39; LaMond (D) (June 7, 1979) at 15, 71-72 (PRSL), 13 GP Supp.App. at 346, 354-55; Wise (D) (June 8, 1979) at 91-103 (Cleveland), 21 GP Supp.App. at 701-13, and the GPs are persuasive in refuting Boyd's contention that the original 4.31 factor was merely a placeholder, later to be replaced by some other figure that was being independently determined.[254] We also find the theory offered to justify the inclusion of avoided future passenger deficits unpersuasive. Where consideration was based upon payments for service contracts, it is consideration for those services, not for the properties conveyed. See Nathan (S) (Nov. 1, 1979) at 41; Nathan (D) (Apr. 18, 1980) at 872-73, 15 GP Supp.App. at 232-33; Netzer (S) (Nov. 1, 1979) at 16-17; Prince (S) (Nov. 1, 1979) at 9 (West End), 14 (Harlem-Hudson); Salvucci (S) (Nov. 1, 1979) at 3 (B&M). This much was virtually conceded upon cross-examination. J. Boyd (D) (June 7, 1979) at 454-56 (Methodology), 9 GP Supp.App. at 192-94; Ronan (D) (July 9, 10, 1979) at 248, 337, 17 GP Supp.App. at 501, 541. In the alternative scenario, public bodies would have had to make additional payments for these services, whether provided by themselves or contracted to other operators. See Ronan (S) (Dec. 1, 1978) at 92 (MTA and the Connecticut Transportation Authority "[s]urely" would not have paid these "staggering" amounts of future operating losses in cash for the Harlem-Hudson and West End properties); Ronan (D) (July 10, 11, 1979) at 337, 511, 586, 17 GP Supp.App. at 541, 586, 611. Finally, even if such consideration were significant, it would be of no help to the Ts' claims for premiums because the Rail Act has already provided equivalent compensation by relieving them of future operating deficits. To the extent the actual cash paid in historical transactions must be augmented to reflect this benefit, so the cash due from the federal government need not cover the future operating losses that the government has already taken over. In fact, the Ts' theory suggests that the GPs would have been entitled to claim the benefit of relief from anticipated future passenger operating losses as an offset to the awards claimed by the Ts. Cf. p. 1337, infra (special benefits doctrine). CRA's use of common basis data is also questionable. CRA makes no claim that these valuations were based upon the beliefs of the parties at the times of the transactions. See CRA (S) (Dec. 1, 1978) at 30; J. Boyd (D) (June 6, 1979) at 180-81, 223-24 (Methodology), 9 GP Supp.App. at 96-97, 122-23. These beliefs are important in predicting public body actions in the alternative scenario because prices can only be based upon X as perceived by the parties to the negotiations, and not upon some other figure determined by hindsight. Throughout its study, CRA explicitly used calculations differing from those of the parties and included elements not known to the parties, particularly in connection with appraisals of common basis. See, e.g., J. Boyd (D) (June 6, 7, 1979) at 185, 192, 329 (Methodology), 9 GP Supp.App. at 98, 103, 146; Gillan (D) (Apr. 25, 1979) at 343, 9 GP Supp.App. at 777; 2 CRA (S) (Dec. 1, 1978) at I 35-36; Caust-Ellenbogen (D) (July 3, 1979) at 71-74 (SIRT), 9 GP Supp.App. at 481-84; Nathan (S) (Nov. 1, 1979) at 9; note 251, supra, p. 1330, infra, (appraisals in B&M and LIRR transactions). In addition, some of the common basis values used were contradicted by available appraisals.[255] *1330 CRA's analyses of the LIRR and B&M transactions, which were cited by EL, exemplify some of CRA's errors. First, the actual cash paid in both transactions exceeded only slightly the so-called common basis for these properties. The figures for cash paid and common basis were $39.5 million and $37.5 million for the B&M, CRA (S) (Dec. 1, 1978) at 154, and $65 million and $63 million for the LIRR, id. at 119. Much of the "excess consideration" in the B&M transaction was attributable to the value of operating subsidies which a Massachusetts official denied were intended as compensation for the property conveyed; rather the money was paid for services B&M had contracted to provide. Salvucci (S) (Nov. 1, 1979) at 3. Moreover, in determining the common basis figure for B&M, CRA ignored an appraisal of two-thirds of B&M's conveyed real estate at a value over fifty percent higher than the figure CRA used for the whole. See Seymour (S) (Dec. 1, 1978) at 78; 9 GP at App. X 130. For the LIRR, Ronan's prior statements in 1967 and 1974 indicate that the cash paid constituted the whole cost to the MTA. Ronan (D) (July 9, 1979) Ex. 4 at 52; id. Ex. 6 at 964-65 (also indicating that "the real estate alone was worth more than the purchase price we paid"). In addition, the common basis figure relied upon for comparison in the LIRR transaction is quite suspect.[256] Finally, as to the Rutland acquisition, contrary to EL's indication that a premium was paid, see ELRB at 28 n.24, no premium could be calculated, because PC consultants never came up with a common basis figure. In short, we find CRA's method inapposite for determining whether any premium would have been paid by public bodies in the alternative scenario and conclude that in any event their data collection and computations are wholly unreliable. In fact, if one were to recompute the measures of consideration and common basis based upon the unrebutted criticisms of the GPs, the results would indicate that public bodies in these historical transactions paid less than common basis. However, especially since the underlying data is so unreliable, we accept the GPs' invitation to ignore CRA's study altogether. Ronan's testimony purportedly provided independent support for the 50 percent premium. He asserts that the amount public bodies would have paid for the Ts' lines would have been dependent upon their social value: the least important would be valued at scrap, but those for which abandonment was completely beyond the realm of possibility would have brought a higher price. Ronan (S) (Dec. 1, 1978) at 79-80. He claims to find empirical support for his position in some of the historical transactions studied by CRA, and reasons that the public would have had no choice but to pay significant premiums because of the fear that a condemnation award would have been even higher. We do not, however, believe that Ronan's testimony adds anything to CRA's study or to EL's case for a premium based upon the desire of public bodies to avoid condemnation. It is unclear how Ronan came to choose his 50 percent figure for the premium in excess of X. Neither he nor R contends that CRA is the source of this figure, and he did not perform his own study of appropriate values for consideration as compared to X for these historical transactions. He asserts that "[e]xperience, ... reinforced by CRA's analysis of the earlier transactions, demonstrates that the difference between the consideration paid and any measure of `liquidation value' grows with the importance of the particular line to the public", Ronan (S) (Dec. 1, 1978) at 79, but does not provide any estimate of the value *1331 to the public in any of the transactions for which he noted CRA's ratios of consideration to common basis or even a hint of a reason suggesting that those transactions for which CRA found larger ratios involved more important properties. Ronan (D) (July 10, 11, 1979) at 459-64, 626-27, 632-33, 17 GP Supp.App. at 573-78, 624-25, 627-28. It is impossible for us to accept so large a premium based only upon very general theorizing followed by what seems to be little more than a guess.[257] Ronan's theory seems to offer little more than repetition of EL's claim for a going concern value premium. His theory that premiums would vary with the importance of the lines is based upon his assumption that "importance is what establishes the bargaining position of the parties". Ronan (S) (Oct. 15, 1980) at 100. He relies heavily upon the the notion that the public would have been under duress, and that the railroads thus would have had the ability to hold out for large premiums. See Ronan (D) (July 9, 10, 11, 12, 1979) at 63-64, 152, 301, 319-20, 325-26, 376, 641-42, 1048, 17 GP Supp.App. at 451-52, 485, 518, 526-27, 532-33, 559, 636-37, 772. Accepting as true the proposition that states would have been forced to condemn the Ts' passenger properties, a contention we reject at p. 1334, infra, it is still unclear how the fear of having to pay a condemnation award somewhat higher than X would result in a 50 percent premium being paid in negotiations since the anticipated condemnation award places a ceiling upon what a public body would willingly have paid. Unlike EL's argument that condemnation might produce an award for going concern value based upon PATH and Fifth Avenue Coach, see pp. 1325-1326, supra, Ronan's theory of an award based upon value to the public lacks even a colorable claim to support in condemnation law. See, e.g., Boston Chamber of Commerce v. Boston, 217 U.S. 189, 195, 30 S.Ct. 459, 460, 54 L.Ed. 725 (1910); United States v. Chandler-Dunbar Water Power Co., 229 U.S. 53, 80-81, 33 S.Ct. 667, 678-79, 57 L.Ed. 1063 (1913); CMV Opinion, 445 F.Supp. at 1013-16. All that remains is Ronan's claim based upon historical transactions. R emphasized in its opening brief, RP at 39, 41-42, that Ronan's theory was amply supported by CRA's study. However, at other points and in reply, they emphasize Ronan's independence from CRA, RP at 64, RRB at 49-50. But, with one exception, we are not told on what other historical evidence he did rely. In the one transaction in which he was involved, his contemporaneous statements contradict CRA's conclusions. See pp. 1329-1330 & n.257, infra (LIRR). Finally, he conceded that if CRA's ratios had come out differently, he would have had to reconsider his conclusions. Ronan (D) (July 10, 11, 12, 1979) at 276-80, 296-97, 603-06, 1028-29, 1068, 17 GP Supp.App. at 504-05, 516-17, 616-19, 769-70, 775. In rebuttal to Ronan, the GPs offered the testimony of Edson Tennyson, Deputy Transportation Secretary for the Commonwealth of Pennsylvania from 1972 until 1979. Tennyson originally testified for North Penn, supporting its claim that public bodies would have acquired its passenger properties. At deposition, North Penn did not permit Tennyson to answer questions pertaining to the prices Pennsylvania would have paid. Tennyson (D) (Apr. 3, 1979) at 17, 26-27, 45, 48, 54, 20 GP Supp.App. at 711-13, 720-21, 724. When testifying for the GPs, however, he stated that he had thus far presented only half of his story and volunteered to appear as a GPs witness, without remuneration, in order to make the record complete, Tennyson (D) (Mar. 19, 1980) at 37, 20 GP Supp.App. at 731. Given his chance, Tennyson flatly stated that Pennsylvania would not have paid more than X for the passenger properties. Tennyson (S) (Nov. 1, 1979) at 4-5, 8-12. The final attempt to rescue the claimed premium is R's evidence concerning Philadelphia's offer of $472,441 for an easement over 3.5 miles of R's Chester Branch. But *1332 the circumstances surrounding this acquisition are not comparable to those that would have prevailed in the alternative scenario where R and North Penn's passenger lines would have been acquired. Most notable is that the Chester Branch was not a passenger line. Moreover, there is no evidence indicating an estimate of X for this segment, so we cannot determine whether any premium was involved. Using the per mile figure derived by R from this example, the value of all their proffered passenger lines comes to $13.7 million, which is less than 25 percent of the $61 million claimed for their passenger properties and less than 35 percent of their $41 million estimate of X (without premium) for those lines. Finally, there is no evidence indicating the basis upon which the acquisition figure was derived by the parties to the transaction. The GPs also played the game of basing conjectures of what would have happened in the alternative scenario on evidence of past transactions. They refer to the 1972 Southeastern Pennsylvania Transportation Authority (SEPTA) Memorandum of Understanding whereby all the Philadelphia area passenger properties were to be leased for a period of fifty years, subject to three fifty-year renewals, for which R and PC were not to receive any cash at all. Ex. GPS-Public Bodies-175, 176. This Memorandum was approved by R's reorganization court, but on appeal to the Third Circuit, the lower court's order was vacated and remanded for further proceedings in light of the Rail Act, which had just been enacted. In the Matter of Reading Co., No. 73-1631 (3 Cir. Mar. 22, 1974). The GPs offered this as support for their argument that public bodies would demand deductions from X in setting prices, see pp. 1333-1334, infra, but, since the issues parallel those discussed in connection with the CRA Study and the other transactions, we consider the SEPTA Memorandum here. In response to the GPs' argument that only one year before the Rail Act it had been willing to give away its passenger properties, R claimed that, in that case, it had in fact received substantial consideration. The most significant alleged consideration was the public's assumption of R's passenger operations and their attendant deficits. This claim parallels CRA's inclusion of such relief as part of consideration in their study and is subject to the same criticism. R conceded at oral argument that including this benefit as an element of value would amount to "double-counting" since the benefit is already received as a result of the proffered acquisition by the public in the alternative scenario. Tr. at 506, see p. 1329, supra. Moreover, R's claim that it would have been able to abandon its passenger operations, 2 RF at 214-305, undermines the value of this alleged element of consideration. A second element of consideration is the free trackage rights granted to R and PC for freight use of the passenger lines covered by the Memorandum. To demonstrate the magnitude of this element, R refers to an agreement by Chessie to pay Conrail $2.2 million annually for operating freight service over what had been R's passenger properties. The GPs note, however, that this was a reciprocal agreement under which Conrail was to pay Chessie a similar sum, Tr. at 352, casting doubt on the true value of trackage rights over these properties. By contrast, at the time of the Memorandum, PC valued the trackage rights over its passenger lines at only $1.6 million annually, Ex. GPS-Public Bodies-188 at 6, an amount seriously questioned as too high by PC's creditors, Ex. GPS-Public Bodies-178 at 28. And it is difficult to square R's valuation of trackage rights with JTTs' witness Jennison's estimate that the annual charge for Philaterm's use of the terminal passenger lines, excluding the Northeast Corridor, amounted to merely $68,774. Jennison (S) (JT) (Dec. 15, 1978) at 35, 52. Finally, even if free trackage rights were in fact extremely valuable, this simply supports the GPs' claim for deductions by indicating the importance of the Ts' passenger lines for freight traffic, see pp. 1333-1334, infra. R also asserted that retained air rights were of significant value. However, since no quantification is available and since the Ts were only to receive half the proceeds from sales of air rights after application to *1333 commuter deficits, see 4 GP at P 200, we cannot attribute any significant value to them. We thus conclude that the SEPTA Memorandum seriously undermines the claim that Ts would have received premiums for their passenger lines, and further suggests that prices less than X would have been likely. (b) Government Parties' Claim for Deductions from X Having dismissed EL's plea for prices for its passenger lines significantly in excess of X, we now consider the GPs' claim that public body prices would instead have reflected deductions from X. The argument begins with the observation that New Jersey's purchase of EL's passenger lines would have conferred a benefit upon EL to the extent that in the sale of its main freight lines it would have been able to derive additional earnings value attributable to traffic dependent upon the passenger lines' continued existence in rail use — an earnings value hereafter referred to as "D". From this premise the GPs argue that public bodies, seeking to pay no more than necessary for the Ts' passenger lines, would have deducted D from the X of those lines to arrive at the price they would have been willing to pay. The theory can be explored from the perspective of the Ts' contention that profitable acquirers of their freight lines would not have purchased predominantly passenger lines under any circumstances, which leaves abandonment as the only private marketplace alternative. If any T would have pursued that alternative, it would have realized the full value of X from its scrap sale, but would have lost D on the sale of its main freight lines since their earnings value would have been diminished by that amount as a result of the abandonment.[258] Thus, EL's claim that X was "the minimum price which EL would have been required to accept on a sale made in order to keep those properties in rail use", ELRB at 290-91, simply misses the point. In essence, the GPs' argument is that a state's payment of X, without deduction, for the passenger lines amounts to a double recovery since the affected T would receive the full benefits of two mutually exclusive dispositions of its passenger lines. It is credible that a state would have followed the pricing rule urged by the GPs, since it derives from the more general principle of paying no more than the best marketplace alternative for the Ts' lines, see p. 1324, supra. The GPs' presentation of testimony by public officials indicating that they would have insisted upon a wide array of other deductions, 4 GP at P 13-21, and that they would have taken into account the degree to which the profitability of the Ts' main freight lines would have been contingent upon the continued operation of public body lines[259] strongly suggests that *1334 the GPs' theory is a sound indicator of what would have happened in the alternative scenario. In fact, the GPs' position is little different from that advanced by some Ts in claiming that off-segment earnings should be accounted for in determining earnings value. See pp. 1369-1370, infra. One caveat to this analysis is necessary. The deduction from the price to be paid for a passenger line cannot exceed its X.[260] D in excess of X is, after all, simply another way of stating that earnings value, broadly defined, see pp. 1219-1220, supra, exceeds X. When that occurs, it would be profitable to include the segment in the package with EL's freight lines rather than permit it to be abandoned. To a private acquirer, a predominantly freight line and a predominantly passenger line would be viewed equivalently to the extent that each made an equivalent contribution to earnings value, so long as responsibility for passenger operations was not linked to purchase of the track. The flaw in the Ts' analysis, see, e.g., ELP at 64 n.58, lies in the inarticulated assumption that passenger operations and the rail lines themselves are inseparable. See Tr. at 311, 318 (GPs). Compare ELRB at 270-72, 290 with id. at 1336 n.265 (essentially conceding the point). It is not necessary in determining the value of a T's property to resolve whether a passenger line for which D exceeds X would be privately or publicly acquired. If it would have been the former, the sale price of the freight package would not be enhanced since the T's estimates of earnings value already account for the earnings attributable to these lines; if it would have been the latter, the private acquisition price is unchanged since the lines will continue in rail use and the public simply pays zero. Of course, in considering these deductions there remains the further caveat that the proceeds from the sale of the freight lines at earnings value combined with those from sales to public bodies cannot be less than X for the entire system. Whether this point will require consideration will not be known until completion of the nonrail use phase of the case. Determining the amount of such deductions is more complicated. The Ts at no point offered any specific evidence on the significance of D. Although the CERL Study found that 44 percent of the freight revenues on the three T's lines derived from traffic exclusively served by public body lines, Waldner (S) (Jan. 30, 1980) at 11,[261] it did not distinguish between the predominantly passenger and predominantly freight lines. Yet there can be little doubt that the values of D for some passenger lines would have been quite significant. Shannon indicated that freight traffic on about half of the New Jersey suburban lines was "substantial". Shannon (D) (May 21, 1979) at 46-47. Figures submitted by both the GPs and EL in response to a letter from the court dated August 28, 1981, Sp. Ct. Rep. N 36850, discussed at pp. 1337-1342, infra, confirm this. We reserve consideration of the magnitude of the deduction applicable to particular passenger lines until after our discussion of bargaining between public bodies and the Ts. (c) Bargaining with Public Bodies Public bodies would have entered negotiations with the Ts for their passenger lines with two objectives in mind: avoiding the loss of important rail service and keeping the purchase price to a minimum, see Ronan (D) (July 10, 11, 1979) at 305, 362-66, 631, 17 GP Supp.App. at 520, 551-55, 626. The latter desire would have been particularly acute given the large amounts at stake, the financial difficulties facing state governments at that time, and the publicity that *1335 would have accompanied any such transaction. See pp. 1320-1324, supra; 4 GP at P 96-102. The Ts' objective would have been to maximize the aggregate return from the sales of their property: the combination of the prices to be received from sales of their freight lines to profitable railroads and of their passenger lines to public bodies. We find that under these circumstances the public bodies would have had a strong upper hand, with the result that successful negotiations, avoiding the need for condemnation, would have yielded prices of X minus D for the Ts' passenger lines. Cf. pp. 1220-1221, supra (bargaining with only one bidder). The source of the Ts' weakness in the bargaining is the dependence of the sales of their freight lines to profitable railroads upon the disposition of their passenger properties. See, e.g., Altshuler (S) (Nov. 1, 1979) at 21, 26-27 (prospective solvent acquirers would have pressured Ts to dispose of public body lines rapidly to assure earnings on main freight line acquisitions); Netzer (S) (Nov. 1, 1979) at 8-9, 41-42. To the extent that earnings on the main freight lines did not depend upon the continued operation of any particular passenger line, the deduction issue is moot since the D for such passenger lines would be zero. Conversely, the greater the applicable deduction, the greater by definition is the dependence of the freight sales upon public acquisition of the passenger lines and the weaker would have been the Ts' position in bargaining with the states.[262] The only way a T could have extracted more from a public body would have been if it could have effectively threatened to abandon its lines and thus disrupt the state's commuter operations. But the Ts would have been unable to mount a persuasive threat for a number of reasons. First, such a threat with respect to a line contributing significant earnings value to the main freight lines would not be credible on its face. Witnesses for both the Ts and the GPs agreed that cessation of operations by a T for any substantial period of time would have been irreversible.[263] It might be argued that similar damage would be inflicted on the public, but the public is in a much different bargaining position. Although harmful to the public, such action would amount to suicide for the T, and public bodies were well aware of this.[264] EL's witness Shannon had previously counseled EL against using the threat of shutting down in order to obtain bargaining leverage. Shannon (D) (May 23, 1979) Ex. 6 at 7. *1336 In addition, the state would have realized that any particular T would have been a one-time player at the bargaining table. Thus no T could have hoped to gain from any enhanced credibility in future negotiations that might have resulted from actually carrying out the threat.[265] By contrast, the state is a repeat player that must bargain with other Ts as well as with other entities in related future situations, such as, for example, possible acquisitions of bus companies, Cahill (S) (Oct. 15, 1980) at 11, railroads not yet bankrupt, and a wide variety of other enterprises. Thus it has much to lose from giving in to any unreasonable demands of a T. See Stangl (D) (Mar. 25, 1980) at 420, 19 GP Supp.App. at 744.[266] Finally, even if a T could mount a persuasive threat and overcome any ICC barrier to acting upon it, a state would have been able to resort to quick-take condemnation powers. See note 221, supra. Of course, condemnation was at the state's option. The Ts claim that a negotiated solution would have been most likely for a number of reasons. See, e.g., ELRB at 291 n.294 (arrangement of financing easier for state with negotiated sale).[267] But the GPs are correct in noting that condemnation would be available, and, if the Ts held out for excessive prices, likely. Schuler (S) (Nov. 1, 1979) at 14 (referring to PC's Remsen line: "... I was convinced that, as a public official, it would have been irresponsible for me to offer any more to Penn Central than they could have realized absent a state acquisition"). All that remains to be determined is what condemnation award would have been most likely if that course had been taken. We have already dismissed the Ts' claim that premiums above X would have been awarded by a condemnation court, see pp. 1324-1326, supra. Here we note that it would have been appropriate, and thus reasonable to expect, that a condemnation court would recognize the GPs' theory for deducting D from X in determining an award. The reasoning is a simple extension of the principle of permitting offsets for special benefits.[268] The justification for the special benefits doctrine transcends mere metaphysics. It recognizes that the purposes behind the constitutional requirement of just compensation are both consistent with and served by the rule. See Bauman v. Ross, 167 U.S. 548, 584, 17 S.Ct. 966, 980, 42 L.Ed. 270 (1897). It is consistent with providing justice to the condemnee since the purpose of compensation is to ensure that the one affected is not left worse off by the government's taking. Id. at 574, 17 S.Ct. at 976. By hypothesis, D is a measure of the gain in price from the private sale of the Ts' freight lines afforded by continued operation over the passenger lines, and is *1337 thus a correct measure of a deduction from an award for the passenger lines.[269] It would be anomalous to consider "mere use of a railroad [to] be a special benefit if the owner of the [condemned] land may connect his establishment with the main line by spur tracks or the like", see 3 Nichols, supra, § 8.6203[3] at 8-180 to -181 & n.42, but not to consider the use of a railroad a special benefit if the owner of the taken property may connect its entire freight system directly with the now public railroad. Just as in more traditional applications of the special benefits doctrine, this benefit is not one generally received by a group of property owners. In this case the special benefit is unique to the Ts as they benefit from the taking in a different and far more significant way than does any member of the general public. And since the requirement of just compensation includes justice to the public, see Garrison v. New York, 88 U.S. (21 Wall.) 196, 204, 22 L.Ed. 612 (1875); Searl v. School District No. 2, 133 U.S. 553, 562, 10 S.Ct. 374, 377, 33 L.Ed. 740 (1890), a condemnation court would feel compelled to permit the deduction, which has the effect of dispersing widely the social benefits from the taking. In order to avoid misunderstanding, we repeat two previously highlighted limits to the implications of our conclusions. First, the deduction urged by the GPs is limited by the actual contribution the publicly acquired passenger lines make to the Ts' sales. Second, a corollary to this proposition is that deductions would not be permitted to bring the total for all the Ts' properties below X for their entire system. (d) Magnitude of the Deductions The theory upon which the deduction argument is predicated, viewed in the context of the bargaining process just described, provides the basis for estimating the magnitude of deduction from X that would have been demanded by a public body in the alternative scenario. As we noted previously, see p. 1334, supra, however, the briefs indicated only the qualitative significance of D, without providing the data necessary for computations. As a result, the court sent a letter on August 28, 1981 to counsel for EL and the GPs, Sp. Ct. Rep. N 36850, requesting more focused information based upon materials in the record. The responses to this letter provide data indicating the total freight traffic moving over each of EL's predominantly passenger segments, broken down into the number of carloads originating or terminating on the segment, the number interchanged on the segment, and the number simply passing over it. The data also indicate how much system revenue derives from cars moving over each of the segments for each of the hypothesized EL-LV acquisition packages. Finally, the GPs performed some simple computations to indicate the earnings value that would be lost on the EL-LV system in the absence of traffic that moves over EL's predominantly passenger lines. The parties' replies exhibited substantial agreement as to the answers to the court's questions; disagreement focused upon the implications of those answers. See, e.g., Reply of the Erie Lackawanna Trustees to the Government Parties' Response to the Court's Letter Dated August 28, 1981, Concerning *1338 "Predominantly Passenger Lines" in the North Jersey Terminal Area (Sept. 21, 1981) at 1-3, Sp. Ct. Rep. at N 37082-4 (hereinafter EL's Sept. 21, 1981 Response); GPs' Reply to EL's Response to Court's Letter of August 28, 1981, (Sept. 21, 1981) at 1-2, Sp. Ct. Rep. at N 37071-72 (hereinafter GPs' Sept. 21, 1981 Response). In particular, EL conceded that its line from Port Morris through Denville and Boonton into Croxton Yard and Hoboken (Boonton line) "was essential to the profitability of the EL-LV acquisition". EL's Sept. 14, 1981, Response at 6, Sp. Ct. Rep. at N 36894. As to the remainder of EL's predominantly passenger lines, the focus of EL's argument continues to be that they are not essential for the sale of the freight lines to a solvent railroad to go through. The GPs respond correctly that this does not address their claim that these lines contributed to the earnings value of the freight lines, with consequent effect upon the price New Jersey would have paid and the trackage rights fees it would have charged. GPs' Sept. 21, 1981, Response at 6, Sp. Ct. Rep. at N 37076. The GPs' calculations indicate that in excess of $100 million of earnings value for the EL-LV system may be due to traffic moving on EL's passenger lines,[270] figures that rival EL's claim for $103.8 million in salvage value for these lines, see Shannon (S) (EL) (Dec. 31, 1978) at 92. Given EL's concession as to the essentiality of the Boonton line, it seems quite likely that a properly computed D will be in excess of X for the Boonton line, suggesting that New Jersey would have acquired this for a price only marginally in excess of zero. The outcome for the other predominantly passenger lines is far less certain because the corresponding D's are much smaller than for the Boonton line and because we will not have conclusive estimates of the X for any of the passenger lines until completion of the next phase of the case. Thus we must proceed on the assumption that D will not exceed X for at least some of the lines. This possibility raises a basic methodological question which we have thus far postponed: Must D be computed segment-by-segment, or is it satisfactory to estimate a single D for the passenger lines as a whole? The latter "package" approach essentially permits excesses of D over X on some passenger lines, such as the Boonton line, to overflow onto other lines, thereby diminishing or fully negating any positive prices that would be derived under an "itemized" approach for segments having an X greater than D. Concededly, negotiations between EL and New Jersey in the alternative scenario probably would have produced a package deal rather than several line-by-line agreements. But the most reasonable assumption is that the price for the package would reflect estimates based upon figures for each segment. This follows from the fact that if New Jersey offered a lower price based upon, for example, the package methodology, EL would have followed the option of including any lines, such as the Boonton line, for which D exceeded X, shorn of responsibility for the deficit producing passenger operations, in the sale of its freight lines to a solvent acquirer, and abandoned the rest. Simply, EL would find it profitable to abandon any segment for which EL did not implicitly receive a price of at least X minus D from the State. While we conclude that we must compute D for each segment, application of this principle is complicated by deficiencies in available data. To the extent traffic could be rerouted over freight lines, or concentrated on a few of the passenger lines, EL would have bargaining leverage as to those passenger lines upon which its freight movements were not truly dependent. But there is no testimony demonstrating the feasibility *1339 of rerouting freight traffic that historically moved over EL's passenger lines, or indicating over which lines such rerouted traffic might travel. To the extent traffic cannot be rerouted, there is no evidence indicating whether it could be loaded elsewhere, or whether the most likely alternative loading points for such traffic would be onto the EL-LV system or some other. Finally, we cannot be certain of the amount of double-counting that would be involved in examining revenue figures for different passenger segments since some of the freight traffic moved over more than one segment. To be certain that no evidence was overlooked, we explicitly inquired about the "feasibility and cost of rerouting freight traffic from EL `predominantly passenger' lines," Court's Sept. 24, 1981 Letter, Sp. Ct. Rep. N 37187. The parties' replies demonstrate that the record is essentially silent on the question. Government Parties' Response to Court's Letter of September 24, 1981, (Oct. 7, 1981) at 3-4, Sp. Ct. Rep. N 37235-36; Response of the Erie Lackawanna and Lehigh Valley Trustees to the Court's Letter Dated September 24, 1981, Concerning "Predominantly Passenger" Lines in the North Jersey Terminal Area (Oct. 7, 1981) at 5-6, Sp. Ct. Rep. at N 37258-59.[271] We have made every attempt to elicit all record materials relevant to determining the magnitude of D and now must make the most reasonable estimate possible from available information. The first step is to estimate the traffic that would be lost if any particular segment were abandoned. We choose a base figure that includes all originating and terminating traffic and excludes the rest. Traffic that merely passed over a segment or was interchanged on a segment is that most likely to be reroutable. In addition, considering such traffic would lead to some double-counting since in many cases it originated or terminated on one of the other segments. By contrast, limiting consideration to originating and terminating traffic makes double-counting most unlikely since movements that both originate and terminate within the small area covered by the passenger lines are negligible. See ELRB at 275 & n.273. Moreover, originating and terminating traffic cannot diTraffic Originating or Terminating on Passenger Segments: EL-LV Western Acquisition[*] Line Cars O/T[**] Total Cars Ratio Hoboken to Port Morris 148,022 213,060 .695 Croxton to Gladstone 1,058 3,615 .293 Croxton to Suffern 47,576 67,266 .707 Suffern to Sparkill 1,231 1,273 .967 NJNYE to Nanuet 7,839 7,848 .999 Hoboken to Rdgwd. Jct. 11,252 11,401 .987 Summit to Denville 5,861 6,376 .919 Newark to Montclair 341 341 1.000 Hoboken to Ptrsn Jct. 7,355 8,682 .847 *1340 rectly be rerouted over another EL line, so it is the traffic most likely to move by some other mode of transportation, or move over some other carrier's system, in the event that the passenger segment was abandoned. The exclusion of overhead traffic in estimating D results in some understating of the importance of the passenger lines for main line freight earnings because not all such traffic is counted twice and because rerouting is not costless. Including all originating and terminating traffic no doubt overstates the importance of some passenger lines. But in the absence of any evidence by the Ts concerning the feasibility of rerouting no other option is available. This base figure includes virtually all traffic on some of the smaller branch lines but excludes a significant portion (although not a majority) of traffic on the major lines into Croxton Yard. From these base traffic figures, we derive a revenue estimate. We simply multiply the revenue figures attributed to all traffic on each of the segments by the ratio of originating and terminating traffic to total traffic. Although it is possible that the average revenue on carloads that originate and terminate on the passenger lines is somewhat higher or lower than the average for all carloads moving over the passenger lines, it is doubtful that any such error is significant and no evidence has been produced suggesting that there is any error at all, or providing any way to correct for any imprecision that might result. Revenue Loss: EL-LV Western Acquisition[*] Line Total Revenue Revenue Loss Hoboken to Port Morris 72,852,084 50,613,494 Croxton to Gladstone 454,480 133,012 Croxton to Suffern 20,969,680 14,831,467 Suffern to Sparkill 477,037[**] 461,298 NJNYE to Nanuet 2,701,073 2,697,975 Hoboken to Rdgwd. Jct. 5,015,226 4,949,682 Summit to Denville 1,117,467 1,027,207 Newark to Montclair 118,351 118,351 Hoboken to Ptrsn Jct. 2,381,465[**] 2,017,470 Finally, we must estimate the earnings value attributable to the revenue that would be lost from the abandonment of each segment. We adopt the approach used, for example, by Shannon in determining the earnings value contribution resulting from projected increases in revenue on the EL system, Shannon (S) (EL) (Dec. 31, 1978) at 49 (Western acquisition), and by the GPs in their Sept. 14, 1981 Response, at 18-21, Sp. Ct. Rep. at N 36873-76. This method multiplies the figure for the change in total revenue by a factor indicating the excess of revenue over variable cost. This is reasonable because for relatively small changes in traffic levels over a railroad's entire system, only variable costs are affected. For example, an increase in traffic of 2 percent would increase labor, fuel, and maintenance costs somewhat, but typically would not require laying new track. Similarly, if traffic were to decrease 2 percent, the number of trains, level of maintenance costs, and overhead would not decrease by a full 2 percent. By contrast, if traffic levels were to change by a large magnitude, the *1341 concomitant change in costs would be more nearly proportionate, although still not one-to-one. See generally Jennison (S) (EL) (Dec. 31, 1978) at 19; Shannon (D) (May 21, 1979) at 99, EL App. A at 221. The 60 percent variable cost-to-revenue ratio used by Shannon for changes in traffic with the Western acquisition in the range of 6 percent, Shannon (S) (EL) (Dec. 31, 1978) at 49, is persuasively defended by EL in the context of arguing that it should be applied to its projected traffic increases. See 1 ELF at 79 n.64 (indicating use of similar ratios in prior ICC decisions); see pp. 1354-1355 & n.293, infra. The revenue loss that would result from abandonment of the Boonton line, however, is significantly in excess of that range. For this segment, Jennison's cost-to-revenue ratio of 71.4 percent, developed for application to an almost 20 percent decline in LV traffic volume in a region, Jennison (S) (EL) (Dec. 31, 1978) at 18, is most reasonable.[272] We need not determine comparable figures for the Chessie Acquisition for reasons given at note 151, supra. Annual After-Tax Earnings Loss Line Earnings Loss Hoboken to Port Morris $7,237,730 Croxton to Gladstone 26,602 Croxton to Suffern 2,966,294 Suffern to Sparkill 92,260 NJNYE to Nanuet 539,595 Hoboken to Rdgwd. Jct. 989,936 Summit to Denville 205,441 Newark to Montclair 23,670 Hoboken to Ptrsn. Jct. 403,494 The final step is to capitalize the annual earnings value estimates. The appropriate discount rate, 9.13 percent, was derived in Part VII. The resulting values for D are as follows: Magnitude of Deductions[*] Line Deduction Hoboken to Port Morris $79,274,148 Croxton to Gladstone 291,374 Croxton to Suffern 32,489,524 Suffern to Sparkill 1,010,510 NJNYE to Nanuet 5,910,132 Hoboken to Rdgwd. Jct. 10,842,677 Summit to Denville 2,250,180 Newark to Montclair 259,257 Hoboken to Ptrsn. Jct. 4,419,430 We recognize that these estimates are imprecise. In addition to the reasons previously noted, a number of problems arise from using revenue estimates and variable cost ratios that are derived from historic information and do not reflect various adjustments to the EL-LV pro formas noted elsewhere in the opinion. Our conclusion that the EL-LV figures overstate earnings values of the main freight lines because trackage rights fees for use of publicly owned passenger lines are understated, see pp. 1342-1346, infra, provides one example. Higher trackage rights fees imply a lower earnings value for traffic moving over the passenger lines, and thus lower values for D. Allowance for the labor protection costs on the main freight lines, see Part IX, supra, would have a similar effect. Beyond this, if EL had decided to abandon a passenger line because D exceeded X, this would entail further labor protection costs. Since, although the GPs devoted significant attention to the theory behind the deduction argument, neither party considered the issues arising in attempts to quantify its magnitude, we rest with the above figures *1342 as a fair approximation rather than attempting further refinement. 5. Trackage Rights Fees EL's argument that "the State, once having acquired the commuter lines, would have granted the necessary trackage rights to the solvent acquirers", ELP at 17 n.14, neither addresses the question of what fee would have been charged nor indicates what assumption concerning acquisition price is implicit in the proffered testimony. The EL-LV pro formas are based on the assumption that the public body would have charged a solvent acquirer the avoidable maintenance cost for freight movements over passenger lines.[273] In sharp contrast, the GPs contend that public bodies would have included common costs, rehabilitation costs, and acquisition costs in the trackage rights fee. The GPs further argue that trackage rights fees would have been set to recoup from acquirers of the Ts' main freight lines as much of the new financial burden on public bodies as possible, attempting to share in the earnings values contributed to the main freight lines by the passenger lines. We find substantial merit in some but not all of the GPs' criticisms, and conclude that public bodies would have computed trackage rights charges based upon an allocated cost formula that covered charges for both maintenance of way and rehabilitation, but not the acquisition cost itself. First, it is necessary to determine the overall formula that a public body would have applied. The GPs present a number of cases where public bodies successfully resisted increasing passenger subsidies to cover anything more than the avoidable cost of the passenger operations. See, e.g., Whitehurst (S) (Jan. 30, 1980) at 14 (New Jersey and PC), 17-18 (New Jersey and CNJ); Stangl (S) (Nov. 1, 1979) at 17 (New Jersey's historical unwillingness to pay any "freight costs" on the passenger lines). And the testimony of public body officials confirmed that these principles would have continued to govern decisions in the alternative scenario. See Altshuler (S) (Nov. 1, 1979) at 28-29; Stangl (S) (Nov. 1, 1979) at 21-22; Whitehurst (S) (Jan. 30, 1980) at 52-53. The Ts do not offer proof to the contrary. However, the GPs, despite their evidence, based all of their projections upon an allocated cost formula. See, e.g., Government Parties' Response to Court's Letter of August 28, 1981 (Sept. 14, 1981) at 24 & n.18, Sp. Ct. Rep. at N 36879. Since this formula is both easier to compute and accords better with a sense of fairness in light of the heavy passenger traffic on these lines, it seems more likely than any other to have governed negotiations in the alternative scenario. Deciding that the trackage rights fee would be based upon an allocated cost formula does not end our inquiry since there *1343 are countless theories upon which to base the allocation. Passenger traffic requires track to be maintained at a higher level than is necessary for freight both because frequent, moderate delays that may be of little consequence to freight movements can be a major disruption for commuter service and because passenger trains operate at higher speeds. On the other hand, in considering the level of necessary maintenance for heavy density freight lines (over 5 million gross tons per mile), Jackman, on behalf of the JTTs, assumed that track conditions should permit movements at 50 mph. Jackman (S) (JT) (Dec. 15, 1978) at 3. Since freight traffic is significantly in excess of 5 million gross tons per mile on the Boonton line, and very near that level on the Suffern line, GP's Sept. 14, 1981 Response, Table A, Sp. Ct. Rep. at N 36883, it cannot be said that most maintenance on EL's passenger lines is required only because of the existence of commuter traffic. Moreover, since freight cars are heavier, they cause more wear regardless of the level to which track must be maintained. For example, the presence of significant freight traffic would make it necessary to lay heavier rail and replace it more frequently. This is presumably the basis for the ICC's formula that allocates costs based upon gross ton miles. In fact, a gross ton mile formula was used to determine B&M's share of maintenance costs over its passenger-carrying lines that were acquired by the MBTA. See 2 CRA (S) (Dec. 1, 1978) at VI 63. Still, given the added requirements of passenger traffic, EL's witness' contention that allocation on a gross ton mile basis is unfair, Shannon (D) (May 22, 1979) at 250, EL App. N, at 78, has some force. We believe that the most reasonable allocation and the one that would have been agreed upon in negotiations between public bodies, the Ts, and acquiring railroads in the alternative scenario is a formula based upon car-miles. It is more generous to freight than the gross-ton-mile formula used by the ICC or the avoidable cost passenger formula that is supported by the GPs' witnesses although not advanced by the GPs themselves, but reflects more reasonably than either of the other two formulas the additional maintenance required for passenger traffic. That the trackage rights fee would cover regular maintenance expenses is not disputed. The Ts assume, however, that rehabilitation would not be included. Yet no reasoning or evidence is offered in support of their position. As revealed in our discussion of rehabilitation expenses in Part VIII, the distinction between maintenance and rehabilitation is somewhat artificial. See pp. 1249-1250, supra. Rehabilitation is necessary when regular maintenance has lapsed. Both expenses are necessary to make the line suitable for carrying traffic. It would be ironic for the Ts to avoid completely the burden of rehabilitating passenger lines that are necessary for their freight movements when their failure to keep them in adequate condition in past years caused their current condition. At some points, the GPs argue that acquisition costs should be covered as well. Over lines for which D exceeds X, we have already concluded that the acquisition price would be near zero, so there is no acquisition cost to recoup. As to the others, where X exceeds D, inclusion of acquisition costs in the trackage rights fee would make little difference. We have already demonstrated that the earnings values attributable to publicly acquired lines, D, would be deducted from X in setting the purchase price while the Ts would have sold the main freight lines to solvent acquirers for the full amount of earnings, values, see pp. 1333-1334, supra. Were the public to charge a trackage rights fee set to recover any portion of the acquisition cost, the value of the freight lines and thus the price any solvent railroad would have been willing to pay in acquiring them would be reduced by the same amount.[274] As a result, the acquisition *1344 price for the passenger lines must be reconsidered. Since that price equalled X minus D, every dollar by which D decreases leads to a one dollar increase in the price the public must pay. D represents the maximum that public bodies could extract through any combination of trackage rights fees and deductions from the purchase price for the passenger lines. On lines for which X exceeds D, the public has already extracted the maximum amount possible in setting the purchase price.[275] It is more plausible to assume that D would have been deducted from X in setting the purchase price for the passenger lines than to assume that the public would have paid more for the lines and charged the difference in trackage rights fees. First, on lines for which D is less than X, a trackage rights fee set to recoup the full purchase price would, for the reasons noted above, induce the acquirer to cease operations over such lines since the trackage rights charge would be in excess of the earnings value contribution of the traffic. As a result, to recover D through trackage rights fees reflecting acquisition costs, the fees would have to vary line-by-line to reflect differences in the ratio of D to X on each line, an unduly complex and unlikely arrangement. Second, higher trackage rights fees induce freight users to eliminate marginally profitable movements and divert others. It is more efficient, therefore, both in terms of resource allocation and collecting the maximum amount of revenue, for the public body to recoup D in a lump sum through a deduction from X in setting the acquisition price than through a usage charge reflecting acquisition costs.[276] Thus we have chosen to apply deductions to the acquisition price for the passenger lines and to exclude acquisition costs from trackage rights fees. A final possibility that must be addressed in the GPs' suggestion that public bodies would use trackage fees to extract any additional earnings value. As noted previously, since acquisition price already reflects a deduction from X, the only lines over which any additional extraction would be possible are those having D in excess of X. But those lines would already be selling at a price near zero, and freight is already hypothesized to be bearing its share of maintenance and rehabilitation costs. Thus, if public bodies proposed trackage rights fees *1345 that were any higher, the Ts would simply have included the lines for which D exceeds X in the freight sale package, leaving the public bodies with the burden of conducting passenger service. We indicated previously that, for scheduling reasons and because of the ability to control maintenance more directly, public bodies would presumably have had a preference for ownership, especially since they inevitably would have owned the other passenger lines in the alternative scenario. But if the threat of excessive trackage rights fees were serious, there is no reason to doubt that those passenger lines that contributed in excess of their X to earnings value would have been privately acquired, with the public left to conduct passenger operations. Since none of the GPs' evidence suggests that the public would attempt to go so far, apparently against its own interest, we continue to assume that New Jersey would have acquired all of EL's passenger lines. All that remains is to compute the trackage rights fee. Our letter of Sept. 24, 1981, Sp. Ct. Rep. N 37178, requested the parties to provide information concerning trackage rights charges and maintenance of way and rehabilitation costs on the predominantly passenger lines because the briefs devoted virtually no attention to this issue. The parties' replies continued a heated controversy over the level of maintenance of way costs on the passenger lines and the proper allocation of these costs between freight and passenger users. In Part VIII we indicated why we find EL's maintenance of way figures more reliable than those presented by the GPs, see pp. 1247-1251, supra. Since there is sufficient evidence in the record to estimate maintenance of way costs on the passenger lines based upon EL's studies, we use them here as well. We have no reason to question EL's figure of $1,987,000 for maintenance of way costs attributable to passenger service, which is taken from EL's Form R-1 for 1973. EL attributes $2,788,000 of its freight maintenance of way costs to freight traffic on the passenger lines. This figure is based upon the average system freight maintenance of way costs per equated track mile, which was multiplied by the number of equated track miles of the passenger lines. See EL's Oct. 7, 1981, Response at 9 & nn.8-10, Sp. Ct. Rep. at N 37262. The GPs question the accuracy of such an approach, GPs' Oct. 7, 1981, Response at 9, Sp. Ct. Rep. at N 37241, but do not prove that the result is too low. EL's method provides a reasonable, although very rough, estimate, which we accept as the best available. We have rejected Shannon's conclusion that freight's share of this $4,775,000 total is merely $733,000. See note 273, supra. We also are not persuaded by EL's repeated assertion that Shannon's figure is reasonable on the ground that the approximately $4,000,000 share it assigns to the State of New Jersey is a lesser burden than the $4,300,000 assumed by the State in 1973 under its avoidable cost subsidy arrangement with EL. First, the evidence does not indicate the total maintenance of way cost or allocation formula upon which the $4,300,000 figure was based. The GPs are correct that it defies belief that passenger avoidable costs were approximately 90 percent of the total maintenance of way costs on these lines. EL's position also is hard to reconcile with its $2,788,000 figure for freight maintenance of way costs. In fact, EL refers to this figure as "the cost of maintaining the line for freight service," EL's Oct. 7, 1981, Response at 12, Sp. Ct. Rep. at N 37265; hence passenger avoidable costs for maintenance by definition must be at least $2,788,000 less than the total maintenance of way costs for these lines. Second, none of the evidence presented proves that the State actually paid $4,300,000 in 1973. The figure is taken from a study prepared for the State that arrived at a figure of $12,742,000 as EL's passenger deficit for 1973. EL App. O at 23-66. Yet other evidence relied upon by EL demonstrates that in 1973 the State's subsidy payment was only $9,106,000. Ex. ELR-NJ-1 at 225, ELP at 16. Third, the GPs are correct in noting that "EL's position that no acquirer of its main freight lines would have acquired the `predominantly passenger' lines (EL October 7 Response at 25) cannot be reconciled with EL's simultaneous contention that under its scheme passenger *1346 service would cost less than New Jersey paid under the 1973 contract. See id. at 3." GPs' Oct. 14, 1981, Response at 8, Sp. Ct. Rep. at N 37360. We therefore conclude that the most reasonable estimate of the maintenance of way component of the trackage rights fee New Jersey would have charged can be computed by directly considering our best estimate of the total maintenance of way costs on the predominantly passenger lines and allocating that cost based upon car miles of traffic. The $4,775,000 figure must be adjusted to normalize the level of maintenance, which adds $133,650 to the total, see workpapers ELR 300 206165 (revised), Workpapers to Memorandum of the Erie Lackawanna Trustees in Response to the Court's Letter of June 19, 1981 (July 10, 1981) at 30, yielding a total of $4,908,650. There seems to be little dispute that freight traffic constituted 56.2 percent of the total movements over EL's predominantly passenger lines on a car mile basis, see GPs' Sept. 14, 1981, Response at 25, Sp. Ct. Rep. at N 36880, which yields an annual charge of $2,758,661 to cover maintenance of way expenses.[277] A further adjustment must be made to account for freight's share of other costs related to operation of the passenger lines. See GPs' Oct. 7, 1981, Response at 12 n.10, Sp. Ct. Rep. at N 37244.[278] The GPs indicate that they did not separately compute an appropriate figure, id., so we must rely upon the only available estimate, which is that computed by Jennison in the context of the JT studies. The relevant figure is $.02199 per car mile, Jennison (S) (JT) (Dec. 15, 1978) at 33-35, which we multiply by the number of freight car miles, 15,426,404, Workpaper to EL's Sept. 21, 1981 Response, yielding an addition to the trackage rights charge of $339,227. The total annual charge is thus $3,097,888, not the $733,000 included by EL. The solvent acquirer of EL's freight lines must also pay for its share of rehabilitation costs. We accept Shannon's overall estimate of these costs subject to the few minor adjustments noted in Part VIII and believe that an acquirer of EL's and LV's main freight lines would have paid 56.2 percent of the total cost of rehabilitation. The acquirer might have paid this amount to the public body as a one-time trackage rights assessment or might have performed the work itself and been reimbursed by the public body for its share of the bill. XII. Proffers of Freight Lines at Capitalized Earnings — Issues Not Heretofore Considered In this Part we consider a number of questions relating to proffers of freight lines at capitalized earnings with which we have not previously dealt. The largest portion of this Part will relate to the EL-LV combination. We shall then pass on to AA and L&NE. A. The EL-LV Combination As indicated in Part II, EL believed that of the possible acquisitions proffered by it, the most attractive would be a combination of its main line with LV's line from Waverly, New York, to Northern New Jersey. LV's witnesses agreed that this also was the best disposition for itself. We agree with these positions. The advantages of this combination would have been great. As discussed elsewhere, see Part II, both EL and LV had lines running from the NJ Terminal area to Buffalo. EL's main line continued on from Hornell, New York, to Chicago. Almost 85 percent of LV's traffic originated or terminated *1347 on its lines between Sayre, Pennsylvania, near Waverly, New York, and the New Jersey terminal area, with only 15 percent of its originated or terminated traffic on the 172.6 mile line from Sayre to Buffalo. EL's line from Binghamton, New York, through Scranton, Pennsylvania, to the New Jersey terminal area generated less traffic than its line through Port Jervis and Suffern, New York, and almost all that was generated on the Scranton line could have been diverted to LV's parallel line from Waverly to Oak Island Yard near Newark. By abandoning EL's Scranton line and LV's route from Ithaca, New York, to Buffalo and routing the traffic of EL's Scranton line over LV's line and LV's traffic to and from points west of Waverly over EL's lines to Buffalo and Chicago, an acquirer could create a rail system with a compact network of rail lines, a relatively simple route structure, well clustered industry, and extremely high traffic densities. Adik (S) (EL) (Dec. 31, 1978) at 4. Both Adik and Shannon testified that potential acquirers of EL's and LV's lines would have viewed this routing as extremely desirable and no GPs' witness disputed them on this point. See id.; Shannon (S) (EL) (Dec. 31, 1978) at 39. Accordingly, our consideration will be devoted to a Western or Chessie acquisition of the EL-LV combination, which would include the sale of freight lines in Northern New Jersey in the package.[279] Our discussion will deal with four topics: (1) Whether or not there would have been competitive bidding for the EL-LV combination. (2) The validity of the EL-LV earnings estimates. (3) EL's claim for a $100 million enhancement in price because of New York's alleged willingness to make a grant of that amount to the purchaser of its route across the Southern Tier. (4) The effect of our rejection of the JT proposal. 1. The Existence of Competitive Bidding For reasons developed in Part V, we must face at the outset the question whether there would have been competitive bidding for the EL-LV combination. We find that there would have been. The rock on which EL's case with respect to competitive bidding was built was that the roads running west from the principal gateways of Chicago[280] and St. Louis would not have sat by and witnessed the disappearance of the only railroads that gave them satisfactory access to New York, Northern New Jersey, New England, and Southeastern Pennsylvania. The Chessie's route from Chicago to New York was exceedingly circuitous; the Norfolk & Western had none. The situation with respect to the Philadelphia area was similar, although Chessie's disadvantage was not so great.[281] Once one Western line moved toward an acquisition, whether of EL or, perhaps more likely, of a PC west-east IFL, others would follow. Chessie and Norfolk & Western, dismayed at the Western invasion of the East, would have entered the fray to blunt its effects so far as possible; indeed, they might have moved earlier in an effort to forestall the invasion. EL offered two important pieces of evidence in support of this scenario. One was a study made on his own initiative by John *1348 Darling, Acting Director of Cost Analysis and Research of the Santa Fe.[282] The study is reproduced at Shannon (S) (EL) (Dec. 31, 1978) Ex. 1. He noted that EL and Santa Fe shared a number of common characteristics, including small amounts of branch mileage, reputation as high speed railroads, and major emphasis on trailer on flat car (TOFC) traffic. Darling's preliminary report concluded that after a six-year rehabilitation period, EL would contribute annual net after-tax income of $28,488,000 to the consolidated net income of the Santa Fe. This was projected to grow at an annual average rate of 15.4 percent until it leveled off to an annual growth rate of 5.1 percent ten years after the acquisition. These assumptions were predicated on an average inflation rate of 4.6 percent per year with no significant EL traffic growth. By becoming the first transcontinental railway system Santa Fe would have had important opportunities to gain new traffic, although these were not taken into specific account. While EL had been undermaintained, Darling did not think the road had deteriorated to the point reached by PC or some other bankrupts. The report was distributed to members of Santa Fe's executive department, to Lee Hudson, Vice President of Traffic and to John Barriger, the Assistant to the President-Staff Studies, who expressed concern about retaliation by other railroads. A further study, made by Darling and signed by Robert Keyes, his supervisor, concluded that "the retaliation of any one or group of railroads in the northeast could not result in the long run diversion of sufficient traffic to negate the revenue gains of the traffic obtained by entering the northeast." Darling (D) (Jan. 31, 1978) at 30, EL App. A at 99. Santa Fe concluded not to pursue the matter, apparently because of fear of opposition by other railroads.[283] However, Hudson contacted a Peter Lutken of Murchison Brothers. Hudson reiterated Darling's conclusion that EL could be operated profitably and Santa Fe's concern that its interchange business, particularly the TOFC business, "would be jeopardized if Erie should fall into the hands of less friendly or less competent owners." Lutken (S) (Dec. 31, 1978) at 2. With Murchison's approval, Lutken contacted Maxwell, President of EL. No offer materialized. According to EL, this was because Maxwell was obliged to fix a two week deadline in light of the date set for Chessie's final acceptance of USRA's offer of part of EL under the FSP, id. A much more elaborate study of EL was made by Chessie when, pursuant to the FSP, USRA offered part of EL to Chessie in 1975. This study involved several thousand man-hours of effort, with EL giving full cooperation and providing raw data that would not ordinarily have been made available. Two reports, presented to Chessie's Board of Directors on October 27, 1975, and November 19, 1975, showed that an EL acquisition would contribute net pre-tax income of $36.3 million and $30.7 million to Chessie's earnings. The difference is largely attributable to a reassessment of the operating costs and revenue gains from the additional EL traffic operating over Chessie's lines. The reports are reproduced at Shannon (S) (EL) (Dec. 31, 1978) Ex. 12, 13. These reports were prepared not to assist Chessie executives in negotiating with EL but to justify the acceptance of USRA's bargain offer.[284] Indeed, at a meeting of *1349 Chessie executives held prior to submission to the directors, serious question was raised whether the estimates of per diem charges were not too high by as much as $11 million a year and it was noted that all budgets appeared to be based on a very conservative approach, to the point that Chessie's President, H. T. Watkins, asked "Have we gone too far?" Watkins' attitude, and his direction that the Chessie study be kept confidential within top management, see White (D) (July 18, 1977) Ex. 13, EL App. A at 109-13, stand in considerable contrast to the comment of James White, author of the study, much cited by the GPs, that "there was great sentiment [in Chessie] that said `We are not interested in Erie, it's a dog'". 3 GP at H 56-57 n.22. The GPs naturally sought to downplay the force of EL's evidence. With respect to Western line acquisitions, they rely not only on the testimony of Barriger; see note 283, supra, and accompanying text, but on statements of Frank Barnett, former Chairman of the Union Pacific, Alan Boyd, former President of Illinois Central Gulf, and Thomas Lamphier, then President of Burlington Northern. Although Barnett, Barnett (S) (Nov. 1, 1979) at 3, strongly stated his belief that the Union Pacific "would [not] have been interested in acquiring any of the Penn Central's lines, or for that matter any of the lines of the Northeastern bankrupts (including Erie-Lackawanna's Chicago to New York line)", his two major reasons, massive rehabilitation costs and ICC-imposed labor protection costs, might not have been such important factors with respect to EL, see Parts VIII and IX.[285] Moreover, his professed disinterest in any of the east-west lines of the bankrupts contrasts strongly with his dire warnings in testimony to a Senate committee in 1973 about the effect on the national economy in general and the Western railroads in particular if Congress failed to enact the Rail Act, Shannon (S) (EL) (Dec. 31, 1978) Ex. 6; we cannot escape the belief that his avowed unwillingness to do anything to pull the Union Pacific's valuable chestnuts out of the fire stemmed from his belief that the government would and should perform the task instead. Boyd, as President of a north-south line, was naturally less preoccupied by the threat of a shutdown of the bankrupts. Moreover, his comments were addressed primarily to the IFLs. Lamphier, who had not read any of the evidence in the case, conceded that a shutdown of the PC lines from Chicago would have harmed the Burlington Northern; that if a competitive bid were made for a PC line, Burlington Northern would have been faced with a choice of attempting to block the offer before the ICC — a rather unlikely prospect — or submitting a competing bid. Cross-examination developed a memorandum from Lamphier's predecessor as President to the then Chairman of Burlington Northern, Lamphier (D) (June 19, 1980) Ex. 1 (GPs), EL App. H at 78-81, making clear that acquisition of any PC or EL line by a competitor would have seriously impaired Burlington *1350 Northern revenues; that Burlington Northern would have gained revenues by acquiring one of those lines; and that EL's lines would have been among those considered. Altogether we cannot conclude that, with the number of Western lines available as purchasers, not one would have made a bid for EL. On the contrary, we think it highly likely that at least Union Pacific, Santa Fe, and Burlington Northern would have been bidders for the three available Chicago-New York lines, and that Southern Pacific, Missouri Pacific, and perhaps Union Pacific would have bid for PC's IFLs running east from St. Louis, especially the St. Louis-Harrisburg line, which afforded access to New York either via Amtrak through Philadelphia or via R and CNJ, and to Philadelphia through R. Since Missouri Pacific could not have been sure of obtaining PC's St. Louis-Harrisburg line, it might also have bid on one of the lines running east from Chicago to which it had access by its subsidiary, the Chicago & Eastern Illinois.[286] It seems highly likely that with this much prospect of Western lines entering the east, Chessie would have bid either for PC's water-level route between Chicago and New York, or for EL's route or the EL-LV combination between Sterling, Ohio, and the Northern New Jersey terminals. While the water-level route was doubtless the best route between Chicago and New York, Chessie would have encountered stiff competition from Western carriers on such a purchase and the EL-LV combination would have been less expensive since Chessie would not have been obliged to acquire the 372.7 miles between Chicago and Akron, Ohio.[287] In attacking the prospect of a Chessie bid for EL, the GPs begin, 3 GP at H 40, with abortive meetings between Chessie and EL officials during 1974, after the Rail Act had been passed but before EL had come under it, in which Chessie said that it "would not consider acquiring the Erie Lackawanna, in whole or in part, under a Section 77 proceedings [sic]." We find this of little significance. For reasons indicated above, the Rail Act had ended the threat of a Western invasion, and if Chessie were going to acquire EL, it would obviously be preferable to do so under the Rail Act, with Government financing of labor protection, rather than without it. The GPs also argue that the collapse of USRA's offer to Chessie in the FSP, 1 FSP at 14, of a "package" consisting of "most EL freight services east of Sterling (Akron), Ohio, most of Reading services [plus] the Charleston, W.Va., market of the Penn Central and access to some CNJ traffic in the Elizabethport/Perth Amboy, N.J. areas", demonstrates that Chessie had no interest in acquiring EL. In fact, the collapse was due to the inability of Chessie to agree with certain labor unions that the protective conditions for workers on the acquired lines should be the same as those prevailing on Chessie rather than the much more liberal ones provided in § 505 of the Rail Act. The GPs also rely on figures developed during the negotiations. USRA had originally asked $114.1 million for the package, based on the net liquidation value or going concern worth of the various parts of the package, whichever was greater. USRA allocated only $14.9 million to EL as its liquidation value, since going concern value was estimated as negative. Chessie accepted on condition that the price be paid 20 percent in cash and 80 percent in B&O bonds, which USRA considered as a cash equivalent of approximately $62 million. *1351 Both sides having held fast for some time, Chessie moved to break the log-jam by offering, on October 28, 1975, $26.4 million for the R and PC Charleston properties but said that it had found EL to be "essentially without value" and was "concerned that EL's earning power was marginal or negative." Ex. GPS-PP-53 at 1. When USRA insisted on the sale of the package, Chessie raised its offer to $52.61 million. Agreement was ultimately reached at $54.5 million for the package, with Chessie emphasizing that "the maximum amount we can pay is net liquidation value". Id. 55 at 1. We find little force in all of this. Chessie was toughly negotiating to pay as little as possible. How USRA could have come out with a net liquidation value of only $14.9 million for EL is beyond our comprehension; the FSP, I FSP at 142, had estimated this at $60.27 million. When Chessie was asserting to USRA that EL had no earnings value, it had in hand detailed confidential studies made by its own employees showing anticipated annual pre-tax earnings of $30.7 and $36.3 million — studies which its management thought might be too conservative. In light of the foregoing, we find that in all likelihood there would have been competitive bidding for EL or a combination of EL and LV by Chessie, and by more than one Western line in the absence of the Rail Act. As for the Norfolk & Western, while we believe that EL has established that Norfolk & Western would have been interested in purchasing EL, EL has failed to establish how interested Norfolk & Western would have been. We cannot simply assume that the earnings studies which EL presented for the EL and EL-LV Chessie acquisitions can be applied directly to purchase by Norfolk & Western. Accordingly, we find that Norfolk & Western would not have been a factor in the bidding. A practical consequence of all this is that we need not overly concern ourselves with the value of an EL-LV Chessie acquisition since Chessie would pay at most only slightly more than a Western acquirer. See note 151, supra. 2. The Earnings Projections for EL-LV The GPs summarize their belief that EL's earnings projections lack credibility when they say in their reply brief, 1 GPRB at F 19, that, although EL lost $4,729,000 in freight operations in 1973 (and had done worse in 1972 and was to do worse in 1974), the earnings projections assume first year net railway operating income ranging from $18 million to $37 million.[288] The GPs contend that EL offered no sufficient explanation for this "remarkable rebirth of earnings power"; EL and LV claim they did. The EL and EL-LV earnings projections rested on three essential building blocks. These are: (a) Studies by Laurie for EL and Adik for LV of the traffic and revenues which the restructured properties could expect either as part of a transcontinental railway system or as part of the Chessie; (b) Studies by Edward Jennison[289] and Adik of the cost of carrying this traffic; (c) Adjustments by Shannon to reflect additional revenue or lessened expenses not considered in (a) and (b). (a) Laurie's and Adik's Traffic and Revenue Studies We find it unnecessary to detail the methodology used by Laurie and Adik in making their traffic studies of the EL and combined EL-LV acquisitions since we are unimpressed with the GPs' criticisms that these did not adequately "evaluate the entire northeastern rail network of the bankrupts, including changes in ownership". 3 GP at H 78-79.[290] It would have been *1352 impossible to predict just what these changes would have been.[291] We are satisfied that Laurie and Adik did as well as could have been done with the information available to them, and, more importantly, as well as an analyst for a prospective purchaser would have done, and that any overstatements of added traffic due to reroutings are more than counterbalanced by the probability of other traffic gains that were not quantified, see, pp. 1358-1359, infra. The attack on Laurie's and Adik's studies that does require serious consideration is that the "complete reliance upon a single year, 1973, which was an abnormally good year in the rail industry, created a significant upward bias in the results of EL's traffic studies." 3 GP at H 79. Elaborating on this, the GPs seek to refute Laurie's justification, namely, that during the 1963-73 period EL's traffic had increased at a rate of approximately 2 percent per year, Laurie (D) (July 31, 1979) at 867-69, 13 GP Supp. App. at 491-93; Laurie (D) (July 31, 1979) Ex. 37, 13 GP Supp. App. at 502, by pointing out that the growth had occurred between 1963 and 1969, after which traffic declined each year until 1973.[292] The GPs also relied on the facts that in their profitability analysis of EL, R, and PC properties performed during 1973, Chessie officials concluded that 1973 had been "a very good year, and not representative", White (D) (July 22, 1977) at 548, 21 GP Supp. App. at 625, and considered traffic and revenue data from 1973, 1974, and the first five months of 1975, id. (July 18, 1977) at 74, 21 GP Supp. App. at 572; id. (July 19, 1977) at 198-200, 21 GP Supp. App. at 585-87; and that Chessie ultimately rejected 1973 data and used only 1974 and part-year 1975 traffic data, id. (July 19, 1977) at 199-200, 21 GP Supp. App. at 586-87; id. (July 21, 1977) at 405-06, 21 GP Supp. App. at 609-10. They relied also on the fact that Darling, in his study for Santa Fe, concluded that EL's traffic levels in 1975 and 1976 would be below those in 1973. Shannon (S) (EL) (Dec. 31, 1978) Ex. 1 at 59-60. However, they failed to inform us that Darling also thought that for the whole period 1975-84 EL freight traffic would "remain constant" at the 1973 level. Id. The GPs also relied on a forecast of the combined traffic of PC, EL, R, LV, CNJ, and AA prepared by GPs witness Robert Waldner, an employee of Temple, Barker & Sloane, the consultant in forecasting used by the GPs in this proceeding, from a "perspective of early 1976". 3 GP at H 89, et seq. Waldner found that over the 1976-85 period, traffic of the bankrupts would have somewhat increased, but that, for a number of commodities, comprising 48 percent of EL's freight traffic, volume would have declined from 1973 to 1978 by figures ranging from 2.9 percent to 15.5 percent. EL points out that the drop from 1973 to 1974 was not particularly drastic, especially considering that 1974 was a year of deep recession, and seeks to account for the sharp decline in 1975 on the basis that shippers were uncertain of EL's future in view of the uncertainty, not dissipated until our decision of April 11, 1975, In the Matter of Erie Lackawanna, 404 F.Supp. 954, whether *1353 EL would be allowed to reorganize under the Rail Act. However, EL's stronger answer is that matters should be viewed from the perspective of early 1974, by which time decisions on purchases would have been made, rather than with the hindsight of later years. Here EL draws great comfort from the fact that the GPs' own forecasting consultant, Temple, Barker & Sloane, in late 1974 and 1975 predicted in one study that EL's traffic level would exceed the 1973 level in the years 1977-80 and in another study that it would exceed that level in 1985, see Temple, Barker & Sloane, Forecast of Traffic and Revenues, 1974-1980, Volume III, Forecast of Conrail Traffic and Revenues, 1985 (Dec. 1974) at Ex. 1-1, EL App. O at 287. We do not believe EL is to be faulted for using 1973 traffic and revenue data. As indicated in Part X of this opinion, discussions of the future of the bankrupts would have begun in 1973 and the purchase price would have been established in early 1974, even though ICC approval would not have been obtained that soon. Nineteen seventy-three would thus have been the last full year for which data would have been available, and developments in early 1974 would not have indicated that it was untrustworthy. (b) Jennison's and Adik's Cost Studies Jennison's and Adik's expense estimates were prepared as follows: EL and LV freight costs were broken down into "unit costs" representing costs per car originated or terminated, costs per car interchanged, and costs per loaded car mile. These were then multiplied by the related counts of service units. In the studies relating to EL alone, the only costs eliminated were: (1) The expenses of operating and maintaining the EL docks at Hoboken and the EL car floats that had operated in New York harbor. This was done because Shannon had concluded that neither a Western line nor Chessie would have acquired these facilities. (2) Property taxes to New York State and its municipaiities. This was done on the basis of testimony by Louis Rossi of the New York Department of Transportation that the State and the municipalities would have committed themselves to exempt EL properties from taxation as an inducement to the acquisition of EL by a profitable railroad. (3) A portion of the property taxes in New Jersey and of the payroll taxes attributed to the New Jersey freight service. (4) Maintenance of way expenses attributable to the EL suburban lines in New Jersey, which would have been acquired by the State. (5) Maintenance of way expenses attributable to non-acquired branch lines. In the EL-LV studies Jennison eliminated all New Jersey state taxes, and certain other expenses for the handling of EL and LV traffic in New Jersey on the ground that the EL and LV lines would have become part of the NJ/NYTerm, and substituted the trackage rights fees and switching charges calculated in the JTs study, Jennison (S) (EL) (Dec. 31, 1978) at 15, 17. In the EL-LV studies, it was recognized that because of the much heavier movement over LV's line resulting from the abandonment of EL's Scranton line, its 1973 maintenance of way expenses would have been inadequate. Hence the experts eliminated these expenses and the EL maintenance of way expenses for traffic that would now move over LV's Waverly-Northern New Jersey line, and Adik developed a detailed study of normalized maintenance requirements adequate to handle the increased traffic. These did not include overhead items since Shannon concluded that most of such items would be eliminated if LV became a part of EL and made separate provisions for those that would not be. Jennison also eliminated LV dock and car float expenses and New York and certain New Jersey property taxes for the same reasons indicated with respect to his EL studies. In the case of the Chessie acquisition, costs of the movement of EL-LV traffic over the B&O line west of Sterling were computed on the basis of the B&O 1973 ratio of variable costs to systems revenue, to wit, 77.6 percent. *1354 The first principal attack by the GPs on the expense estimates is that EL's earnings studies substantially understated the switching charges that would have been charged by the public bodies in Northern New Jersey. However, since we reject the JTs case and find that the acquirer of the EL-LV main line would have acquired the EL-LV JT properties as well, we do not have to deal with this problem. As already discussed, the GPs further claim and we agree that the trackage rights fee estimated by EL for use of the passenger lines was too low. See pp. 1342-1346, supra. The GPs also claim that Jennison improperly calculated the cost of handling: (1) former LV traffic considered to move over EL's lines west of Waverly, N.Y., in the assumed EL-LV combination acquisition, and (2) former EL and LV traffic considered to move over B&O's line between Sterling, Ohio, and Chicago, in the assumed Chessie acquisitions of EL and LV lines. 3 GP at H 155-56. Jennison's theory was that "modest" increases in traffic of up to "20 or 25 percent" could be handled on an incremental cost basis, which in EL's case he estimated to be 71.4 percent of revenues, Scheer (S) (Jan. 30, 1980) at 15-16, as opposed to a full freight cost to revenue level of approximately 97.3 percent, EL 1973 Form R-1, Schedule 310; Jennison (S) (EL) (Dec. 31, 1978) App. 1, at 4; id., App. 3 at 2. For traffic increases over 25 percent, Jennison applied a "variable" cost-revenue ratio, which was an average of EL's full and incremental cost-revenue ratios. See pp. 242-43 & n.272, supra. The GPs assert, 3 GP at H 156 n.66, that application of B&O's variable cost ratio of 77.6 percent rather than its full cost ratio of 93.5 percent to Chessie's acquisition of the EL-LV package accounted for an increase to annual net income of $10,730,000. A principal difficulty facing the GPs is that they produced no expert witness to testify that full costs should have been charged. Common sense tells us that unless a railroad is operating at capacity, some additional traffic can be accommodated at less than full cost. See p. 1340, supra. There is no suggestion that the LV line between Waverly and Northern New Jersey was operating at anything like capacity; indeed, in another context, the GPs tell us that the LV was "starving for traffic". 3 GP at J 18. There is also no evidence that the B&O line between Sterling and Chicago was operating at capacity. Chessie in its own study applied incremental costs to EL traffic that would move over its lines as a result of rerouting, White (D) (July 20, 1977) at 347, EL App. N at 716; Sparrow (D) (Mar. 17, 1978) at 135-49, EL App. N at 731-45. Application of incremental or variable costs to added traffic was no invention of Jennison's; it is in general use by railroad experts throughout the country, see, e.g., Rules to Govern the Assembling and Presenting of Cost Evidence, 337 I.C.C. 298, 305-09 (1970). We see no reason to question Jennison's use of variable and incremental costs.[293] The GPs also attack Jennison's use of a system average switching cost for work performed in the allegedly higher cost Northern New Jersey area. Although the GPs do not explicitly say so, this argument has little relevance to a Western line acquisition since all major yards were included and any understatement of New Jersey *1355 switching costs resulting from the use of a system average would be compensated by overstatement of such costs elsewhere. However, the GPs make a superficially appealing argument on this subject with respect to a Chessie acquisition. Since, in that proffer, the Chessie did not acquire EL lines west of Sterling, Ohio, where switching costs are assumedly lower than those in the Northern New Jersey area, the use of a system average switch cost that included the allegedly efficient yards west of Sterling would tend to understate costs on the Eastern portions of the EL's lines. EL responds, ELRB at 122-24, we think correctly, that the GPs' argument rests on the assumption that because switching costs in Northern New Jersey are higher than average, switching costs west of Sterling must be lower. The only testimony in the record on this point shows that the GPs' assumption is wrong. Jennison testified that he believed the Marion Division (west of Sterling) switching costs were close to the system average, Jennison (S) (EL) (Dec. 31, 1978) at 14; Shannon testified that in his opinion Marion Division switching costs were higher than the system average because of inefficient operation in the deteriorated and antiquated Marion Yard. Shannon (S) (EL) (Dec. 31, 1978) at 12-13.[294] The GPs offered no evidence on this point. (c) Shannon's Further Adjustments Shannon projected that by the fifth year after acquisition, net railway operating income would be substantially improved as a result of operating savings.[295] The principal items were: (1) Savings in operating expenses from the rehabilitation program; (2) Savings from the modernization of EL's classification yard at Marion, Ohio;[296] (3) Savings from rebuilding EL's classification yard at Croxton, New Jersey;[297] (4) Savings from single tracking EL's main line between Binghamton, New York, and Northern New Jersey via Port Jervis, New York, and in the case of acquisition of EL alone, parts of EL's line between Binghamton and Northern New Jersey via Scranton, Pennsylvania. While not denying the probability of some savings with respect to each of these items, the GPs argue that the amounts estimated by Shannon were so large hat a purchaser would have viewed them with suspicion, especially since these items represented significant portions of total net railway operating income. Specifically the GPs say that: (1) It was error for Shannon to estimate the savings in overall expenses from rehabilitation and increased annual maintenance would be 5 percent of these sums, rather than to make a detailed examination of individual expense categories; (2) A potential acquirer would have viewed with suspicion Shannon's shift from an estimate of $3.4 million in annual savings from reconstruction of the Marion Yard made in the spring *1356 of 1973, Shannon (D) (May 25, 1979) Ex. 32 at 1, to an estimate of $6.4 million made in the fall of 1974, Shannon (D) (May 21, 1979) Ex. 1 at workpaper 206405, and would have wondered why a project with such stupendous savings, which EL admitted had been under continuous study since the mid-60's had not been completed long ago; (3) The plans for upgrading Croxton Yard were so incomplete that it was impossible to make any reasonable estimate of annual savings; (4) The plans with respect to single-tracking the Port Jervis line and parts of the Scranton line were likewise too indefinite for a purchaser to take savings from them into account. We find little force in the GPs criticism of Shannon's estimate of a savings in operating expenses of 5 percent of the total amount expended for rehabilitation and added maintenance. He explained that his long experience in the railroad industry had taught him that companies generally figure on a 10 percent return on such expenditures, half of which takes the form of adding to or avoiding loss of revenues and half the form of reduced expenses. The benefits from rehabilitation and improved maintenance will also show up in the catastrophe accounts with consequent savings in overtime to train and yard crews and maintenance of way labor. These accounts amounted to between $18 and $19 million in 1973, and Shannon's projected savings represented only a 7 percent reduction in these expenses. We see no reason to believe that a purchaser would have doubted that a large program of rehabilitation and added maintenance would have produced this much of a saving in operating expenses; certainly the GPs adduced no credible evidence, expert or otherwise, to that effect. The GPs criticism of the Marion Yard figures are also lacking in force. This was a well-developed project, which the EL Trustee and management had submitted to the bankruptcy court and for which banks had committed to lend the funds. The commitment had been obtained on the strength of two reports by the independent engineering firm of Ford, Bacon & Davis. Both the cost of the project and the anticipated savings had been increased in the second, more detailed, report. We perceive no reason why a prospective buyer would have become suspicious over Shannon's use of these figures. Omitting the Croxton Yard project, which would not have figured in the acquisition of an EL-LV combination, and the single tracking of parts of EL's Scranton line, which would have been abandoned in such a combination, we find equally little force in the GPs' objection that Shannon estimated an excessive "return" on the single-tracking of the Port Jervis line. The only substantial cost in such a project is the labor in picking up abandoned track, a cost that is largely offset by its salvage value. This work, as pointed out in a report by EL's operations planning department, Shannon (D) (May 21, 1979) Ex. 1 at workpaper 206344-206400, EL App. N at 140-99, can be performed over any convenient period of time. Nevertheless the savings begin immediately once the decision is made to stop maintaining one set of tracks. The GPs heaviest attack was reserved for Shannon's estimate of consolidation savings. These, they point out, also produce a very large share of projected net railway operating income. They claimed that Shannon relied on nothing more than judgment and the application of a supposedly universal but undisclosed "Wyer Dick formula". The EL's opening freight brief, 1 ELF at 98-105, contained a detailed explanation of Shannon's method of estimating consolidation savings which showed that the GPs had either seriously misunderstood or mischaracterized Shannon's testimony. Railroad mergers fall into two general categories, parallel and end-to-end, although any particular merger is likely to have some characteristics of each. Parallel mergers provide an opportunity for savings by elimination of duplicative facilities and concentration of traffic on fewer lines. Organizational savings, which can be realized in both sorts of mergers, occur because the combined *1357 railroads do not need as many executives, lawyers, accountants, and traffic salesmen as had the two separate companies. Accounts particularly susceptible to reduction are the supervisory accounts, including the heads of the maintenance of way, maintenance of equipment, and transportation departments and their staffs; traffic department expenses, including the salaries of sales executives and field personnel and rental expenses for sales offices; and general expenses, including the executive, accounting, law, and medical departments. Employing the same general method as Wyer Dick had used in many railroad mergers since 1955, Shannon projected that either a Western purchaser or Chessie could have achieved savings of 50 percent of the expenses in the maintenance of way, maintenance of equipment, and transportation overhead accounts, 90 percent of the traffic department overhead,[298] and 70 percent of the general expenses. Shannon (S) (EL) (Dec. 31, 1978) at 59. He testified that he had checked his results against those obtained by use of a different formula developed by John Barriger, another recognized expert in estimating merger savings, and found the results to be consistent. Id. at 59-60. He said also that he had compared his results with those of the Chessie study and found Chessie's figures also to be generally consistent. Id. at 84. He further concluded that in the EL-LV acquisition, almost all LV overhead expenses could be eliminated — a reasonable conclusion since, for all practical purposes, the acquired LV line was simply replacing EL's Binghamton-Scranton-Northern New Jersey lines and the overhead incident to the latter should be adequate to support the former. The GPs presented the testimony of Albert Moon,[299] who stated that he "would not advise an investor to take such estimates [as Shannon's] into account in arriving at a proposed purchase price." Moon (S) (Jan. 30, 1980) at 88. Without going into exhaustive detail, EL's opening brief placed Moon in a state of nearly total eclipse. No investor had ever asked Moon's advice on such a question. He had no opinion whether Shannon's results were right or wrong; his sole point went to Shannon's methodology, which he erroneously conceived as being a mechanical application of the average results of Wyer Dick predictions of savings in five other railroad consolidations. The GPs' reply brief made scant attempt to remove the shadow. Rather it dwelt upon the somewhat semantic question whether Shannon had or had not used a "formula" based on earlier Wyer Dick studies and the need for a sufficiently "property-specific" study.[300] Beyond this the GPs rely on statements by their financial witnesses, Jones and Smith, both of First Boston Corp., as to the uncertainties of projected merger savings, Smith (S) (Aug. 26, 1980) at 2166-67 (by Jones), 19 GP Supp. App. at 433-34; Smith (D) (Aug. 4, 1980) at 100 et seq., 19 GP Supp. App. at 5, and 19-51, and to more extreme positions taken in extracts from a number of official sources. However, as pointed out in EL's reply brief, ELRB at 106, these extracts do not make clear whether they were directed at savings of overhead, which constituted the bulk of Shannon's estimate, as distinguished *1358 from operating savings, which admittedly are more difficult to predict with accuracy. In support of Shannon's estimates, EL brought out, 1 ELF at 349-53, that in the two of the three mergers that had been consummated substantially in the form that Wyer Dick had studied,[301] the projected consolidation savings had actually occurred. Lamphier, President of the Burlington Northern, stated that Wyer Dick's estimate of the consolidation savings in the Burlington-Great Northern-Northern Pacific merger had been achieved, and indeed surpassed. Lamphier (D) (June 19, 1980) Ex. 12, EL App. H at 76-77. Likewise, Shannon testified that Wyer Dick's estimate of the savings on the Erie-Lackawanna merger had ultimately been surpassed. The record with respect to the Seaboard Airline — Atlantic Coast Line merger was not quite so clear. However, the impact of the ICC report relied upon by the GPs, in which the ICC reduced Wyer Dick's estimate of annual consolidation savings by 50 percent, Seaboard Airline R. Co. — Merger — Atlantic Coast Line, 320 I.C.C. 122, 154-59, supra, was effectively blunted when EL brought out that the Commission's refusal to take account of some of Wyer Dick's estimated savings did not concern consolidation savings but rather savings from abandonments for which the carrier had ultimately decided not to seek authority. Id. Except for these the Commission noted that "most of the prospective savings appear to be fairly realistic and generally attainable". Id. at 156.[302] We do not say that an acquirer of EL would have accepted Shannon's estimate of consolidation savings at face value, to be taken as the word of God without debate. Very likely, in the real world, an acquirer would have attempted to make its own estimates if EL had been willing to cooperate and time permitted.[303] Consolidation savings, like everything else in the estimates, would have been a subject of bargaining. What we do conclude is that a purchaser would have regarded Shannon's findings as a reasonable estimate of consolidation savings. Before leaving the general subject of EL's revenue and expense studies and turning to two other important aspects of EL's case, we think it desirable to emphasize one factor that gives us considerable confidence in Shannon's results. This is the number of positive factors that Shannon did not seek to quantify and include in his pro forma statements. He listed twelve of these, Shannon (S) (EL) (Dec. 31, 1978) at 71; we quote only the first seven: (1) Traffic the acquirer would lose if EL and the other bankrupts were shut down; (2) Traffic the acquirer would lose if the acquirer did not acquire EL, or a comparable Penn Central line, and competitors of the acquirer made such acquisitions; (3) In the case of the Western line acquisitions, traffic diversions from other carriers to either the lines acquired or the lines of the acquirer or both; (4) In the case of the Chessie acquisitions, diversions of existing Chessie interchange traffic with lines other than the acquired lines; (5) In the case of the Chessie acquisitions, the benefits to Chessie of possibly withdrawing from Canada; *1359 (6) Reductions in EL and EL-LV car hire deficits resulting from investment in additional freight cars, improved utilization, and increased length of haul; (7) Net revenue gains on traffic formerly handled by Penn Central but rerouted over EL and LV because of Penn Central's projected abandonments; ... Item (6) is of particular importance. It was a recognition of this problem, it will be recalled, see p. 1347 & n.282, supra, that attracted Darling's attention to EL and led him to make his study for Santa Fe. EL amplified these points in its opening freight brief, 1 ELF at 109-14, and we find no sufficient answer to them in the GPs' reply. 3. The Addition of $100,000,000 to Reflect a Grant by the State of New York EL contends that the price payable for its freight lines for continued rail use should reflect a $100 million grant that the State of New York allegedly would have been willing to give to a purchaser in order to assure the continuation of EL's freight service across the Southern Tier of that State. Its witness, Shannon, included the $100 million in his first year's projected cash flow. Shannon (S) (EL) (Dec. 31, 1978) App. 1B, 2B, 3B, 4B. Although EL's contention was made in the context of attempting to "bridge the gap" between the capitalized earnings of the various EL or EL-LV proposals and the higher net salvage value claimed by EL, EL urges more generally that a purchaser on a capitalized earnings basis would bid higher if part of its payment would be financed by a third party without cost to it. Mere statement of the argument reveals its legal difficulties. EL is seeking, in effect, to have the prospect of financial aid from a state to a purchaser considered as property of the seller within the scope of the just compensation clause. It seems questionable from a constitutional standpoint whether a state's display of generosity can increase the bill of the United States in a federal condemnation. Beyond this since New York's payment, as recognized by EL's witness Young, Young (D) (July 19, 1979) at 161-62, 21 GP Supp. App. at 739-40, would be for the "social value" of EL's rail service in New York, it is arguable that recognition of the grant as enhancing the purchase price would violate the rule that value to the taker is not a proper element for consideration. We find it unnecessary to answer these intriguing legal questions, however, since analysis of the facts shows that New York's offer was by no means so unequivocal as EL contends. Cf. note 220, supra. EL's case rests on the testimony of its witness, Louis Rossi, Director of the Railroad Division of the New York State Department of Transportation. This must be considered along with the testimony of GP witness Raymond Schuler, who at the relevant period was New York Commissioner of Transportation and thus Rossi's superior. Rossi was Vice Chairman of a task force which New York created in 1972 to deal with the developing rail crisis. Prior to the discussions leading to the Rail Act, the task force was discussing with profitable railroads the possibility that they might acquire with state assistance the two principal lines across New York, PC's (formerly New York Central's) water level route and EL's route across the Southern Tier, both of which the State deemed essential. The prospect of a federal solution, beginning early in 1973, put a temporary stop to these discussions. However, the possibility that New York might make some contribution to preserving the Southern Tier route was renewed when the EL reorganization court decided in 1974 that EL should not be reorganized under the Rail Act, In re Erie Lackawanna Ry., 393 F.Supp. 352 (N.D.Ohio 1974), a decision that was not appealed to this court. The task force opposed a plan of the EL Trustees to improve EL's Scranton line and to single-track and downgrade the Port Jervis line. A principal reason for this position was the belief that the Port Jervis line was essential to a reopening of the Maybrook gateway and rebuilding of the Poughkeepsie Bridge so as to permit competitive service between the west and New England, and also with New York City by land rather than by car float. Since the Port Jervis line alone would be inadequate, the task force advocated use of the LV line *1360 from Waverly and abandonment of EL's Scranton line. Rossi (S) (Dec. 31, 1978) at 4. On April 1, 1974, Governor Malcolm Wilson approved the Rail Preservation Bond Act of 1974, 1974 N.Y. Laws, c. 118, N.Y. Transp. Law App. 6. The Act provided for a funded $30 million essential rail services preservation program and a $250 million bond sale, subject to approval by a referendum, which was obtained in the November elections. Later the legislature appropriated $107,500,000 of these funds, $62.5 million of which was for the improvement of upstate freight services. Rossi (D) (Mar. 9, 1979) at 92, EL App. A at 154. In April 1975 the task force made known to USRA New York's willingness to offer substantial financial inducements to profitable railroads to acquire EL's Southern Tier route. Rossi (S) (Dec. 31, 1978) Ex. 6. It also renewed its earlier contacts with potential acquirers of the Southern Tier route, particularly Chessie. Chessie's first concern was with state and local property taxes. It wished to be sure that the 80 percent reduction that had been available to EL because of poor earnings would also be available to it; legislation to that end was adopted, Rossi (S) (Dec. 31, 1978) at 7. Chessie also wished exemption from the remaining 20 percent, and Southern Tier mayors and legislative leaders agreed to support this. Id. Finally, according to Rossi, "we made clear that the State had available under the Rail Preservation Bond Act substantial funds to facilitate acquisition of the Southern Tier route", which "would have been in the form of grants and could have been used for acquisition costs as well as rehabilitation". Id. Rossi asserted that "[o]n several occasions we specifically offered up to $50 million if Chessie were to make the acquisition proposed in the Final System Plan and an additional $50 million if Chessie were to gain access to New York City". Id.[304] Schuler's testimony has a quite different flavor. While not denying that New York might have furnished financial aid to an acquirer of EL's Southern Tier line, he stated that he did not consider making a direct grant of $100 million, or any other amount, to finance acquisition costs. Schuler (S) (Nov. 1, 1979) at 25-26. Rather, any moneys advanced would have been for "a program of projects and not a direct grant for acquisition". Id. at 25. These advances might have been traditional loans, loans repayable out of revenues generated by the project, or outright grants. Id. at 26. Whatever the form, "we would have insisted that the Port Jervis line remain as a main east-west freight line and that the projects to be financed were to supplement and not to substitute for the basic maintenance and rehabilitation that a purchaser would have undertaken on its own."[305] Insofar as the testimony of the two witnesses, differs, we must follow that of Schuler, who was Rossi's superior and would have had to approve any action taken. EL has attributed to the State of New York a certainty that was not there. There was no guarantee that New York would supply $100,000,000 or even $50,000,000, see note 304, supra, to a purchaser of EL's line through the Southern Tier. The most that can be said is that an acquirer of that line could have looked forward to a warm reception from New York's Department of Transportation and could have expected favorable treatment if it sought financial aid *1361 for new capital projects in the State. This is a far cry from EL's argument that, because of a New York "grant", an acquirer would have bid $100 million more than it otherwise would. We thus have no occasion to address the legal problems listed at the outset of this discussion. 4. Consequences to the EL-LV Earnings Case of Our Rejection of the Joint Terminals Concept Our holding that public bodies would not have purchased the EL's and LV's JT properties does not necessarily entail rejection of their earnings cases. As already noted, both EL and LV submitted evidence that assumed that private acquirers would have purchased their properties in NJ/NYTerm. EL, for example, claimed that the value of its properties alone to a Western acquirer would have been $225 million, Young (S) (Dec. 31, 1978) at 22, and LV, assuming an acquisition by a large Eastern carrier, claimed a value of $75 million for its lines. LV at 56. Moreover, EL and LV witnesses testified that the viability of a combined EL-LV configuration would not have been dependent on the existence of the NJ/NY Term. Shannon (S) (EL) (Dec. 31, 1978) at 39-40; Adik (S) (EL) (Dec. 31, 1978) at 8, and the two Ts have so contended in their briefs and arguments. This leads, however, to another problem. EL's evidentiary submission did not explicitly calculate the earnings value of an EL-LV combination based on the assumption that the purchaser would have acquired the EL's and LV's lines in the NJ/NYTerm in addition to their main freight lines (hereinafter the EL-LV without JT). In its June 1, 1981, Statement, EL did give a rough indication of the earning power of an EL-LV without JT proffer. The EL calculated that "annual traffic study net operating income" would increase by $6,777,000 in the case of a Western line, and by $6,748,000 in the case of a Chessie acquisition of EL-LV. These figures do not account for rehabilitation costs. The EL claimed that the asking price for the EL-LV would have increased by $78 million if their JT properties were included. This figure is based on the claimed salvage value of the EL-LV properties in the JT rather than on the capitalized value of the increase in earnings attributed to the EL-LV without JT proffer. Statement of the Erie Lackawanna Trustees in Response to the Court's May 21, 1981, Memorandum (June 1, 1981) at 8-9. Sp. Ct. Rep. at N 35807-08. In a letter to the parties dated June 19, 1981, Sp. Ct. Rep. at N 36204, the court requested EL to give a detailed breakdown of the projected earnings and cash flows of the EL-LV without JT proffer including the costs of rehabilitation. This statement shows that, including rehabilitation expenses as projected by EL's and LV's witnesses, the net gain to the acquirer of the EL-LV without JT combination would be roughly $4,000,000 annually.[306] In response to the EL's filing, the GPs take the position that the court should be "loathe" to rely on the EL's new figures. They first contend that: The transferors' submissions implicitly and explicitly advance many entirely new factual contentions which are not subject to judicial notice. They are not supported by any testifying witness, nor have they been subjected to cross-examination or to evidentiary rebuttal .... It would thus be unfair for the Court to rely importantly on [these] new contentions.... Government Parties' Response to July 10, 1981, Filings of EL and R Regarding Earnings Values Without JTs, (July 24, 1981) at 20-21. Sp. Ct. Rep. at N 36364-65. The court finds little support for this argument. First, as has already been noted, the idea of an EL-LV without JT configuration did not come out of thin air — two EL witnesses discussed this configuration in their statements. What did not appear in those statements are numerical estimates of the earnings power of the EL-LV without JT proffers. But a review of the workpapers underlying EL's July 10, 1981, submission readily reveals that the calculation of the earnings value of the EL-LV without *1362 JT configuration was a straightforward development of figures that are in the record and which have already been subjected to discovery, cross-examination, and rebuttal testimony. For example, the key component in EL's analysis is the substitution of EL's and LV's own operating costs in the JT area for those derived in the JTs study. EL's and LV's own operating costs were taken directly from Jennison's studies of the independent proffers. The GPs had ample opportunity to challenge these figures.[307] What admittedly is different in the EL-LV without JT case is the way in which the figures in the record were used. But, again, we do not think that the GPs were prejudiced by this use. Given that the court believes that an EL-LV without JT acquisition would have been the most likely outcome in the alternative scenario and that all of the numbers that the court would have needed to adjust the EL-LV with JT acquisition to eliminate the effects of exclusion of EL's and LV's JT freight properties can be found in the record, there is no reason why the court could not have taken it upon itself to attempt to make the appropriate adjustments. If the court had chosen this course, it would still have had to rely on the parties to provide it with information from the workpapers. Undoubtedly, we would have heard objections, either from the GPs, EL-LV, or both, that the court erred in its calculations. Rather than pursue this course, the court chose to seek the aid of the parties in its calculations. We cannot see how the GPs can legitimately complain that they were prejudiced by this procedure.[308] Apart from attacking the propriety of our considering the EL-LV without JT configuration, the GPs also argue that EL's contention is, as a matter of fact, simply implausible. At the heart of EL's claim that the earnings value of the EL-LV without JT proffer would be greater than that with NJ/NYTerm is evidence showing that EL's and LV's own historic costs of handling traffic in the terminal areas were lower than the costs projected in the JTs study.[309] Yet, as the GPs point out, the entire purpose of the JTs was to achieve major savings through the consolidation of *1363 redundant facilities. Citing a 1973 PC report that concluded that "terminals will continue to be a major expense and a major operating headache" and that relief could only come from the expenditure of "massive funds" and the rearrangement of facilities, the GPs state that "[t]he transferor's contention now that unconsolidated operations could have been run at (or below) the costs of consolidated terminals is nonsensical", 2 GPRB at L 42. At oral argument counsel for the JTTs attempted to explain this seeming anomaly. First he contended that because the projected JTs costs were based on averages for all of the carriers in the terminal areas and because EL's traffic was of a type that was generally cheaper to handle, when the averages were computed "the Erie came out somewhere on the wrong end of the stick", Tr. at 532-33. He also argued that EL's costs for its non-JT traffic in the EL-LV without JT acquisitions improperly reflected the arguably higher costs in Northern New Jersey because the non-JT traffic costs were calculated on the basis of system-wide average costs, which included the cost of traffic in the North Jersey area. Id. at 526-32. While these two factors certainly could explain the seeming anomaly, the court is reluctant to base its conclusions on what, after all, are speculations. Although we could independently explore the record further to confirm or refute these theories, this would be a most difficult task without the aid of counsel. This does not mean, however, that the total cost figures for the EL-LV without JT acquisition cannot be accepted. Indeed, we think that, subject to a few important exceptions noted in this opinion, we can be rely on them for the simple reason that the total cost figures for the EL-LV without JT acquisitions are derived directly from the system-wide, historical average cost figures calculated by Jennison for EL and LV. Apart from a few problems, which will be discussed later, the GPs give no reason why Jennison's figures cannot be used in this way. Indeed the GPs' major criticism of the entire JTs study — that it is an artificial construct which ignored the historical operations and costs in the JT areas — is a strong argument in favor of using Jennison's figures. Jennison made his calculations directly from actual expenses incurred by EL and LV. And as already discussed, see pp. 1353-1355, supra, we reject the limited criticisms levelled by the GPs at Jennison's study. The GPs do, however, make one relatively major criticism of the actual calculations contained in EL's July 10 submission. The NJ/NYTerm operating plan provided that CNJ's main line, which previously carried both freight and passenger traffic, would have been dedicated solely to passenger use. CNJ's freight traffic was thus rerouted over the NJ/NYTerm freight lines. Whatever the merits this routing decision may have had assuming that NJ/NYTerm would have been created, when that assumption is abandoned any possible justification disappears. Without a wholesale public acquisition of all of the Ts' Northern New Jersey properties, there is no reason to suppose that a private acquirer of EL-LV would have been able to divert to its own line all the traffic historically carried by CNJ. Accordingly, in a letter to counsel dated September 4, 1981, Sp. Ct. Rep. at N 36853, the court asked the parties to adjust the figures contained in EL's July 10 submission to reflect this.[310] *1364 B. Ann Arbor As previously stated, Part II, p. 25, supra, AA, a 99+ percent owned subsidiary of D,T&I, itself a 99 + percent owned subsidiary of PC, as operated in 1973, extended from Toledo, Ohio, in a generally northwest direction, to Ann Arbor, Michigan, and thence across Michigan to Frankfort on Lake Michigan, where car ferries were operated to Manitowoc and Kewaunee, Wisconsin.[311] The 54-mile Toledo-Ann Arbor segment (sometimes hereafter "the segment") was believed to be profitable, the balance of the line to be a producer of deficits. The FSP designated the Toledo-Ann Arbor segment[312] to Conrail and an 86.5-mile stretch of line between Durand and Ashley, Michigan, to the Grand Trunk Western. The balance of the AA was not then designated for conveyance. Prior to promulgation of the FSP an Inter-Agency Task Force appointed by the Governor of Michigan in May 1973 had expressed concern in a report dated January 2, 1975, over expected loss of rail service as a result of pending abandonments and the proposals with respect to AA probable in the FSP, and had advocated the State's assuming a positive attitude. The report recommended that main lines such as AA's should be considered the backbone of the State's railroad system and expressed concern that the expected designation of the Toledo-Ann Arbor segment to Conrail in the FSP would preclude the State's using profits from that segment to cross-subsidize the losing line north of Ann Arbor. The report led the Michigan legislature of 1975 to adopt a number of statutes giving the State extensive power to purchase and operate railroads. Among the 1976 amendments to the Rail Act was a provision, § 208(d)(2), authorizing USRA, upon petition of any State, to make further designations under the FSP for conveyance, inter alia, to a State. On February 11, 1976, Michigan petitioned USRA to redesignate all of AA's properties for conveyance to the State. USRA granted this petition to the extent of redesignating the Toledo-Ann Arbor segment to Michigan instead of Conrail and also designating to the State the segment between Ashley and Cadillac. USRA also adopted a resolution recommending that the Grand Trunk Western grant trackage rights to the State over the segment designated to it, see Michigan Interstate Co. v. Grand Trunk Railroad Company, 459 F.Supp. 1008, 1010 n.2 (Sp.Ct. R.R.R.A.1978). In 1980 Michigan purchased all of AA's remaining properties. The trustee of AA claims that Michigan would have purchased the Toledo-Ann Arbor segment for $9,690,193. Although this figure must be subjected to downward adjustment, we conclude that the Toledo-Ann Arbor segment would have had substantial capitalized earnings value to at least two railroads and would have been purchased in the absence of the Rail Act either by the State of Michigan or by one of the two railroads at a price slightly higher than the lower of the two capitalized earnings values. *1365 1. The Position of the State of Michigan Although in our view it does not particularly matter whether the purchaser of the segment would have been the State or a railroad, we begin by discussing arguments of the GPs that the State would not have been an acquirer. We have thought it preferable to treat this separately from our general discussion of public body purchases, Part XI, because the situation differs in a number of important ways. First, only freight lines are at issue. Second, since AA's earnings value studies for its allegedly profitable Ann Arbor-Toledo segment did not include any traffic originating on the unprofitable northern segment, the GPs' arguments we accepted in concluding that the JTs' freight lines would not have been purchased by public bodies and that prices paid for passenger lines would reflect deductions for the earnings those lines contributed to the privately acquired lines are inapplicable.[313] Third, AA claims no premium in excess of X for its conveyed Ashley-Cadillac segment. Finally, as will appear, there would have been a number of railroads that would have bid for AA's Ann Arbor-Toledo segment. As a result, all we need here consider is whether Michigan would have had the inclination and financial capacity to purchase AA's Ann Arbor-Toledo segment or, perhaps, since Michigan's interest lay in the preservation of the whole of Ann Arbor, the entire Ann Arbor rail line, in the absence of the Rail Act.[314] Relatively little evidence on the position of the State of Michigan in the absence of the Rail Act was or perhaps could be offered by either party. The bulk of AA's argument is based upon what actually transpired in the presence of the Rail Act, under which Michigan ultimately acquired AA's entire system. In the absence of the Rail Act, Michigan also would have been concerned about the future of PC's lines in the State. On the other hand, acquisition of the Ann Arbor-Toledo segment in the alternative scenario would have been at a price AA contends to be substantially higher than the actual price paid by the State under the Rail Act, and the purchase would have had to have been effected without federal assistance that explicitly was part of the federal response to the rail crisis. See 2 GPRB at Q 5 n.7 (use of federal assistance in State's purchase); pp. 1323-1324, supra, (limits upon consideration of federal finance). Also, the profits from the southern segment would not have been available to cover losses on the rest of the system if Michigan, rather than obtaining the southern segment at a bargain price under the Rail Act, would have had to pay the full capitalized earnings value to avoid its acquisition by prospective solvent acquirers.[315] The GPs are probably correct in arguing that Michigan would not have acquired the entire AA unless there were no alternatives. 4 GP at R 165-66. But, aside from the argument that the ICC would have conditioned the sale of the southern segment upon inclusion of part or all of the lines to the north, a question we have above decided in the negative, Part X, pp. 1306-1308, supra, the GPs indicate no other possible acquirer of the northern lines. Michigan's efforts in the presence of the Rail Act to get Conrail or solvent railroads to bear the burden, 4 GP at R 167-72, are hardly surprising but are not necessarily indicative of what Michigan's actions would have been in the absence of the Rail Act. The State's unwillingness to acquire or operate rail lines prior to enactment of the Rail Act, 4 *1366 GP at R 174-75, is likewise not probative since this does not indicate its probable response when faced with a rail crisis. After all, Michigan did enact legislation permitting acquisition and operation of railroads when confronted with the possibility that lines in the State would not be included in Conrail, AA at 6-16, and its study of the State's railroad network and prospective abandonments began in June 1973, prior to enactment of the Rail Act. Petition for the Amendment of the Final System Plan, VII 17 at 2. The importance of the AA to the state was adequately documented. See id. at 5-11. Our review of these materials indicates to us an attitude toward the State's acquisition of AA entirely different from that of the northeastern states whose heavy financial involvement in supporting commuter traffic made them allergic to further expenditures for freight lines, whether for capital or operating expenses, see Part XI, p. 1316, supra.[316] Michigan's financial capacity is not seriously questioned. AA claims that the Motor Vehicle Highway Fund would have been available, for example, by using it to retire special obligation indebtedness, which, by contrast to general obligation bonds, did not require electoral approval. The GPs correctly indicate that the Motor Vehicle Highway Fund was to "be used exclusively for highway purposes as defined by law". Mich.Const.1963, Art. 9, § 9. But in 1977, the Michigan Supreme Court upheld the legislature's definition of "highway purposes" as including the acquisition of rail properties. Advisory Opinion on Constitutionality of 1976 PA 295 and 297, 401 Mich. 686, 707, 259 N.W.2d 129, 137 (1977). There is no reason to doubt that the legislature's definition would have been accepted by the same court in the absence of the Rail Act. In any event, the GPs offer no persuasive evidence suggesting that Michigan would have been unable to raise the necessary funds from any of a number of traditional revenue sources.[317] We thus conclude, although again emphasizing that the issue is not of real importance, that Michigan would have sought to acquire the Ann Arbor-Toledo segment and, indeed, the whole of AA. 2. Competitive Bidding Shannon, testifying as a witness for AA, considered D,T&I as the most likely railroad to have acquired the Ann Arbor-Toledo segment. The primary reason he gave was that the acquisition would have afforded D,T&I an improved route into Toledo, which, despite its name, it did not reach directly but only by trackage rights. Shannon (S) (AA) (Dec. 31, 1978) at 4. However, as the discussion of off-segment revenues developed, improved access of D,T&I to Toledo emerged as a rather minor factor; the important point became that if the Ann Arbor-Toledo segment were acquired by another railroad, D,T&I would have been subjected to substantial traffic diversion. When AA was under D,T&I control, traffic was routed to give the maximum haul over the AA-D,T&I system even if this shorthauled AA. A solvent railroad purchasing of AA's Ann Arbor-Toledo segment would have followed the same policy of obtaining the long haul for its own combined operation, with serious adverse consequences to D,T&I.[318] *1367 The GPs make several answers to the claim that D,T&I would have been a bidder, but we do not find these persuasive. One is the testimony of the AA Trustee that when D,T&I placed AA in bankruptcy, a D,T&I official stated that "they had put their last penny in the Ann Arbor and they were not going to take any further bath." Chase (D) (June 29, 1979) at 9, 9 GP Supp.App. at 509. This testimony, however, related to what D,T&I would do for AA as a system, subject to the claims of creditors, not to what it would do by way of acquiring the Toledo-Ann Arbor segment free of debt, particularly if it were faced with prospects of the latter's abandonment or sale to a competitor. As we shall see later, a purchaser of all the IFLs or an acquirer of the Detroit-Cincinnati IFL, which might be Southern, Louisville & Nashville, Chessie, or Norfolk & Western, Shannon (S) (PC) at 36-37,[319] or Norfolk & Western even if it did not acquire the Detroit-Cincinnati IFL, Shannon (AA) (S) at 4, would have been possible acquirers. The GPs third ground for attacking the premise that D,T&I would have feared diversion is that "carriers cannot dictate routings". Sullivan (S) (Traffic on the AA) (Nov. 1, 1979) at 5. While it is true that if a shipper explicitly directs that his traffic be transported over an established through route, the carrier is legally obligated to follow the shipper's instructions, 49 U.S.C. § 10763, AA argues convincingly that the carrier serving a shipper's plant is peculiarly able to influence the shipper, assuming cost and service are equal, to route traffic so as to give that carrier the long haul or leave the routing to the carrier, which would have the same result, AA at 61-65.[320] Doing this costs the shipper nothing and enhances the carrier's zeal to give good service and respond to complaints.[321] It is true that an acquiring carrier will not ordinarily be able to divert 100 percent of the traffic susceptible to diversion, but D,T&I would have been sufficiently disturbed over the threat of substantial diversion to wish to keep the Ann Arbor-Toledo segment for itself. AA also made a convincing case that there would have been at least one other bidder, namely, an acquirer of one of PC's IFLs. We deal initially with the GPs' claim that it is impermissible for us to consider the acquirer of any or all the IFLs as a potential bidder since there is no assurance that any of the IFLs would have been acquired. Apart from the considerations set forth in Part III of this opinion relating to the GPs' refusal to take a stand with respect to the continuance of any IFL in rail use, see pp. 28-29, supra, we believe there is sufficient evidence in the record to show, at least in the absence of contradictory proof by the GPs, that the IFLs would have remained in service, although they might not necessarily have commanded the prices proposed *1368 by PC. In addition to the general statements by James McClellan, USRA's former Director of Strategic Planning; Frank Barnett, former Chief Executive Officer of Union Pacific and a vigorous sponsor of the Rail Act; USRA's consultants, Temple, Barker & Sloane; and the Rail Services Planning Office of the ICC, see AARB at 15-17, the evidence presented by PC, notably the testimony of witnesses Hallingby, Jennison, and Myers, sufficed in the present context to meet the burden imposed on page 5 of our order of December 5, 1980, immediately after PC's settlement, that "to the extent that any of [the Ts'] proposals for continued operation of their own lines hinge on continued operation of lines of PC, they have the burden of showing that such lines would in fact have continued in rail use". Sp. Ct. Rep. at 23424. When we turn to the facts, we find that such difficulties as AA faces are largely semantic. AA witness Laurie[322] said in his initial statement: Penn Central was recognized as a likely purchaser of the segment by reason of its Chicago-Detroit route, Detroit-Cincinnati route, and its interchanges with the Ann Arbor at Toledo and at the City of Ann Arbor. Although a sale of the Penn Central system to a single buyer assuming no Rail Act would have been extremely doubtful, I nevertheless concur with Mr. Charles Shannon's testimony that a buyer of the Detroit-Chicago route or the Detroit-Cincinnati route or both would indeed have been a likely purchaser of the Toledo to Ann Arbor segment. Laurie (S) (AA) (Dec. 31, 1978) at 2. The difficulty springs from the reference to the Chicago-Detroit route. Although acquisition of AA by an acquirer of that route alone presumably would have afforded some opportunity to divert traffic between points on the Ann Arbor-Toledo segment and the west which AA had interchanged with Norfolk & Western at Milan, Michigan, apparently the bulk of the diversion would have been of southbound traffic previously exchanged with D,T&I at Diann, see note 318, and would have enured to the benefit of IFL lines east of Toledo to Potomac Yard; an acquirer of the Chicago-Detroit IFL would have derived no benefit from this although the assumed acquirer of all the IFLs would have. In his rebuttal testimony, Laurie submitted a Study of Southern Railway Acquisition of the Toledo-Ann Arbor Segment of the Ann Arbor Railroad Company. Laurie (S) (Oct. 15, 1980) Ex. 5. In an introductory section he explained that since his earlier report, which had assumed acquisition of all the IFLs by a single purchaser, "it has been determined that PC's scenario contemplates segmentizing that railroad for sale to various potential purchasers." He chose for study the Detroit-Toledo-Cincinnati "zone"; assumed that, as Shannon had testified, Shannon (S) (PC) (Dec. 1, 1978) at 36-37, Southern would have been a likely purchaser of that "zone", and further assumed that, having made that purchase, Southern would have been a bidder for the Ann Arbor-Toledo segment.[323] This showed even higher earnings than the study made on the assumption that the bidder would have been a single acquirer of all the IFLs since in the Southern study, off-segment revenues included not only those on the acquired IFL, as had been the case with his initial study, but on Southern's own lines south of Cincinnati. *1369 We thus conclude that, assuming for the time being the correctness of Laurie's figures, there would have been two railroad bidders for the segment and that Michigan would have had to pay somewhat more than the lower of the two capitalized earnings values. This last statement warrants elaboration. It is clear, from what has been said in Part V, that if two railroads were the only bidders for a T, the road to which the T would be more valuable would need to bid only somewhat more than the other; hence the lower of the two capitalized earnings values determines the outcome. AA seems to argue that its case is different because there were three potential bidders, not two. If the third bidder were a railroad, AA would be right; the third railroad would have had to bid more than the capitalized earnings value to the higher of the two others, not the lower. But here Michigan had the power to condemn all the Ann Arbor-Toledo segment, except for the few miles between the Ohio-Michigan boundary and Toledo, which AA surely would have been willing to sell at a reasonable price if the rest of the segment were condemned. In a negotiated sale, therefore, Michigan would not have paid more than what it would have had to pay if it resorted to condemnation. See Part XI, pp. 1334-1337, supra. In such a condemnation, the market value would have been only what AA could have obtained from the railroad having the greater interest in the segment and that, as we have shown, would have been only somewhat more than what would have been paid by the next lower. 3. The Extent to Which Off-Segment Earnings Should be Taken into Account in Determining Capitalized Earnings Value Laurie's estimate of the value of the segment to Southern depended heavily on added earnings Southern would achieve on the acquired Detroit-Toledo-Cincinnati IFL and on its own lines from traffic originating or terminating on the segment that had not heretofore moved over these lines. His estimate of the value of the segment to D,T&I depended heavily on the retention of earnings on D,T&I that would be diverted if the segment passed into the hands of a purchaser of all the IFLs. The GPs object to inclusion of such amounts on a number of grounds. We have already largely disposed of the first of these, namely, that traffic cannot be diverted since routing rests in the control of the shipper. What remains to be determined is what proportion of the divertible traffic would in fact have been diverted. Recognizing that "[t]he assumption that PC could control 100 per cent of the interline competitive traffic may not be entirely realistic", Laurie (S) (AA) (Dec. 31, 1978) Ex. 1 at 5, Laurie's study of the value of AA to a purchaser of all the IFLs made separate estimates on the basis of 100, 90, 85, and 75 percent control. He followed the same course in his Southern study. There was no need for doing this in the D,T&I study since this was based on historic movement of the traffic. AA's translation of the earnings into capitalized value used the 100 percent figure. We have no real basis for choosing among the figures except that we join Laurie in believing that 100 percent is too high. Since some figure must be chosen, we take the most conservative and opt for 75 percent. The GPs argue also that earnings should never be counted save on the line on which they are earned. This would be true enough if we were dealing with a study of the overall level of railroad earnings or with segregation studies of the sort that figured prominently in railroad reorganizations of the 1940's, since earnings should not be double-counted. However, we see no reason why if a railroad believes that an acquisition will funnel new profitable traffic over its lines, it would not take that profit into account in making a bid. What must be avoided is not counting off-segment earnings but double-counting them. The GPs assert that the latter has occurred, but we find AA's response convincing. Southern would not have been interested in acquiring the Ann Arbor-Toledo segment unless it was sure it had acquired *1370 the Detroit-Cincinnati IFL, for which, as pointed out, it could well have faced competitive bidding from the Louisville & Nashville. In bidding for the Detroit-Cincinnati line, it would not have known whether it would succeed in acquiring the Ann Arbor-Toledo segment since it would have faced bidding from D,T&I and the State of Michigan. Hence it would have been irrational for Southern to include added earnings from acquisition of the segment in its bid for the Detroit-Cincinnati line.[324] Beyond this Laurie testified in rebuttal, Laurie (S) (Oct. 15, 1980) at 9, that, as suggested by the GPs' witnesses, Allen & Kozu (D) (Aug. 21, 1980) at 1040-41, 1 AA Supp.App. at 37-38, he had contacted PC's valuation witness Wilmot, who advised him that when PC submitted its earnings estimates on December 1, 1978, a month before Laurie's testimony was filed, it had taken no account of any additional traffic that might be received from the Ann Arbor-Toledo segment.[325] 4. American Motors Jeeps Many pages of brief have been devoted to the question whether AA would have continued to receive revenue from jeep traffic from an American Motors plant on Stickney Avenue in Toledo. The jeeps were manufactured at a plant located on Cove Avenue, which was served by PC, but PC's tracks were incapable of handling the type of bi-level cars needed to ship the jeeps. Hence these were trucked to the American Motors parts plant on Stickney Avenue, which was adjacent to AA's southern terminus at Ottawa Yard. The Stickney Avenue plant was served by six tracks, all set up by AA and to all of which it had access. PC had access to Tracks 5 and 6, but these were suitable only for the delivery of coal to the plant. Tracks 3 and 4 were used for loading boxed jeeps for export and Tracks 1 and 2 were used for loading jeeps aboard bi-level freight cars for domestic shipment. GPs witness Sullivan testified, Sullivan (S) (Traffic on the AA) (Nov. 1, 1979) at 6-7, that neither a PC acquirer nor D,T&I would have paid for revenues from the Stickney Avenue plant since AA had no control over the traffic and could have lost it at any time. He also said, id., that a PC acquirer could have adjusted its tracks serving the plant so as to give it full access to Tracks 1 and 2 and serve the jeep traffic. GPs witness Klick added, Klick (S) (Revenues Associated with AA) (Jan. 30, 1980) at 5, that D,T&I could have obtained access to the plant through reciprocal switching over PC if AA went out of business. AA's broadcast answer is that if these changes could be made now, they could have been made years ago but were not,[326] and that this indicates there must have been a good reason for not doing so. It finds the reason in American Motors' desire to have all movements of jeeps at the Stickney Avenue plant under the control of one carrier, to wit, AA. It claims this situation would continue so long as it stayed in business and points out that the GPs have not claimed that the Ann Arbor-Toledo segment had no earnings value. Hence, it argues, American Motors would not have consented to the "adjustment" of the PC *1371 tracks, which were located on American Motors' property, or in any way encouraged PC's competition for the jeep traffic.[327] With respect to D,T&I, AA calls attention to the limited nature of Klick's testimony, namely, that D,T&I could have obtained access by reciprocal switching over PC's lines if AA had ceased operation, which AA claims would not have happened, and not that D,T&I could have obtained access by reciprocal switching over AA. It also complains about the GPs' failure to introduce the reciprocal switching tariff in evidence, and questions the value of access by D,T&I to the Stickney Avenue plant since D,T&I's only access to Toledo was by trackage rights over AA. The situation is a confused one, all the more so because of the failure of anyone to call any official of American Motors. The claim that traffic of a particular shipper would be wholly or partly lost to a T that had long handled it, solely because other carriers would increase their efforts to get it, is surely one on which the GPs had the burden of proof. We think they did not carry it. 5. The Cost of Handling Off-Segment Traffic Having determined what off-segment traffic would accrue to an acquirer of the segment, Laurie then had to estimate the costs of handling it. For this he utilized a study prepared by Edward Jennison[328] with respect to acquisition of AA by either a single acquirer of the IFLs or by D,T&I.[329] Based on his judgment, Jennison set figures for variability in the 1973 operating accounts, expenses for net rents, costs of capital (at 8 percent), and taxes of PC and D,T&I. This resulted in a cost-revenue ratio of 75.58 percent, Laurie (S) (AA) (Dec. 31, 1978) Ex. 1, Table 5, Sheet 4, and of 71.68 percent for D,T&I. Id. Table 8, Sheet 4.[330] In his Southern study, Laurie used the 75.58 percent ratio he had developed for PC for traffic carried on the acquired line north of Cincinnati. Laurie (S) (Oct. 15, 1980) Ex. 5 at 5. However, for the traffic carried on Southern's historic lines he used a figure of 54.56 percent, which was said to "reflect" a special study of the five Class I railroads in the Southern Railway System in 1973. Id. The GPs attack all these percentages as too low, particularly when used to develop earnings that are expected to last forever. They rely not on a detailed criticism of the percentages of variability applied to the particular accounts but to testimony given by the PC's and GPs' opportunity cost witnesses, Kornhauser and Schneider, on the degree to which solvent railroads could recoup revenues lost from PC's abandonment by reducing expenses, including the opinion of some economists that the ratio of long-run variable costs to total costs might be as high as 90-99 percent, which would produce a ratio of long-run variable costs to revenue in the range of 85 percent, 3 GP at K 34-36. In the absence of direct evidence to the contrary, the view that PC could have spread 20,000 added cars of AA traffic over all the IFLs at a ratio of 75.58 percent of added costs to revenues does not seem unreasonable. However, that conclusion does not contribute greatly to a solution since *1372 the proffer of a single acquirer of the IFLs as purchaser of the Ann Arbor-Toledo segment has been withdrawn. With respect to D,T&I, the ratio of the amount of traffic being considered to the railroad's total revenue is, of course, much greater. We would think that on a short run basis, D,T&I would have done exceedingly well to contain the loss of what had been a substantial portion of its traffic to 28.32 percent of revenues. It might have done better in the long run but — apart from Keynes' epigram, which AA refreshingly notes, AARB at 91-92 — long-run savings would be diminished by discount. Also Laurie, in apparent error, failed to take into account that acquisition of the segment by D,T&I would save the latter $168,772 a year in trackage fees, AA at 57 n.2. All things considered, we are not disposed to alter, on the grounds presently under consideration, Laurie's estimate of the capitalized earnings value of the segment to D,T&I. We have more difficulty with Laurie's estimates with respect to Southern. As a matter of common sense one would suppose that if it would have cost an acquirer of all the IFLs 75.58 percent of added revenues to handle the added traffic produced by the segment, including the long haul of much of the traffic to Potomac Yard, it would have cost an acquirer of the Toledo-Cincinnati IFL a higher percentage of revenues to haul the traffic over that relatively short railroad, with the added traffic constituting a larger proportion of the total. As against this Southern has had a particularly low freight operating ratio, ranking third (on the basis of 1969-73 averages) of the twenty railroads studied by GPs witness Smith, Smith (S) (Jan. 30, 1980) Ex. 17 at 14, with a 1969-73 average of 71.1 percent. On this basis the 75.78 percent figure of added costs to revenues over the Detroit-Cincinnati IFL operated as part of Southern does not seem unreasonable. With respect to the estimate of a 54.56 percent ratio of costs to revenues for added traffic over Southern's own line, the "special study" on which Laurie relied apparently was not produced, but the GPs did not ask that it be. In a sense, moreover, the precise correctness of Laurie's estimate of earnings for Southern acquisition is immaterial. His estimated capitalized earnings value for this acquisition ranged upward from $11,091,999, assuming a 75 percent degree of control over segment originated or terminated traffic and a 15 percent capitalization rate, Laurie (S) (Oct. 15, 1980) Ex. 5 at 6, as compared with $7,044,280 for acquisition of D,T&I. Laurie (S) (AA) (Dec. 31, 1978) at 11. For reasons heretofore explained, Southern would have had only to top D,T&I's bid. 6. Use of 1973 Traffic as the Base for Projecting Revenue The GPs opening brief did not directly challenge the propriety of using AA's 1973 traffic as a basis for calculating revenues, except insofar as such a challenge could be inferred from a quotation from a paragraph of the testimony of AA witness Weisman, 3 GP at K 7, that "traffic began to decline again in 1967 and continued to do so for the years thereafter". AA's reply brief is convincing that this testimony referred only to the car ferry traffic, none of which was included in Laurie's estimated revenues, AARB at 77-79. However, during the cross-examination of AA's rebuttal witnesses, Weisman and Laurie, the GPs introduced, over AA's objection,[331] identical exhibits purporting to show that cement shipments on AA declined substantially from 1973 to 1974 and 1975. In its reply brief, AARB at 80, AA noted that "[a]lthough the Government Parties did not raise the matter at all in their Opening Brief, they may attempt to argue in their reply brief that a decline in traffic would have reduced the amount that a purchaser of the Toledo-Ann Arbor segment would have paid if, as the Government Parties have suggested, the disposition of the segment were to have been delayed several years". *1373 AA's anticipation proved correct; the GPs' reply brief contains several pages, 1 GPRB at I 9-15, dealing with the subject and expanding the argument to include automobile originations from American Motors' jeep plant, as to which evidence similar to that with respect to the cement shipments had been used in the cross-examination of AA's rebuttal witnesses, and auto parts, mainly from Ford's Salina plant, figures for which had not been presented.[332] AA objects to consideration of this evidence on a number of grounds. It claims that this material should have been used at the initial cross-examination of Weisman and Laurie, which would have given AA a chance to respond in rebuttal or that, at the least, it should be restricted to Laurie's Southern study which, was the only new earnings material in his rebuttal testimony. It also questions whether the statistics were properly authenticated under F.R.E. 901 and 902.[333] It surely would have been preferable for the GPs to have made their point on Laurie's initial cross-examination so that AA could have developed explanatory evidence in rebuttal. So far as we can determine, the record contains no figures as to the total traffic on the segment on 1974 as against 1973; for all we know, the drop in cement and auto parts shipments may have been made up by increases in other traffic. Moreover 1974 and 1973 were years of recession, and, AA claims, AARB at 87, that if the contention had been timely raised, it could have proved that shipments increased when the recession ended. All things considered, we are not disposed to give more than slight weight to the GPs' contention. We thus conclude that D,T&I would have been willing to pay a sum close to Laurie's estimate subject to the downward adjustments noted herein and required by the discussion in Parts VII, VIII, and IX, and that the acquirer of PC's Detroit-Cincinnati segment would have been willing to pay sufficiently more to insure the acquisition.[334] C. Lehigh & New England We have described the configuration of this small road and the claims of its receiver in Part II H. We now hold that L&NE did not discharge its burden of showing that its Tamaqua branch would have continued in rail use. With respect to its Bethlehem branch, we hold that while this might well have been purchased by Chessie as an acquirer of R's class (a) lines, there is insufficient evidence that Chessie would have faced a competitive bidder. Hence this property also is to be valued at the appropriate version of X. The sole connection of the Tamaqua branch was with R's Little Schuylkill line. R did not proffer this for continued rail use and classified it as a class (c) line whose future R was unable to predict. It is true, as L&NE argues, that this was not a concession *1374 of abandonment on the part of L&NE. However, in its Statement of Position dated May 10, 1978, at 1, Sp. Ct. Rep. at N 10926, L&NE acknowledged that the Tamaqua branch would not have been sold for continued rail use. L&NE had the burden of showing that a connecting line would have existed and did not discharge its burden. The position with respect to the Bethlehem branch is different. This branch connected with R and LV at Bethlehem, Pennsylvania, and with EL at Bath, Pennsylvania. However, EL's line from Bath to Portland, Pennsylvania, was not designated for conveyance and it connected with a line of EL that would be abandoned in the event of an EL-LV combination. The only possible acquirers thus would have been Chessie as an acquirer of R's class (a) lines and the acquirer of LV.[335] The figures make it plain that neither of these acquirers would have bid for L&NE on the basis of the earnings on the L&NE's own line, which had become negative. The question thus becomes whether L&NE's computation of off-segment revenues is such as to show that competitive bidding would have existed. L&NE witness Storm estimated total off-line revenues of $1,268,434. He assumed the incremental cost of carrying this traffic would be between 70 and 80 percent. Although in our discussion of AA we have approved recognition of off-segment revenues that could have been diverted (or preserved) through an acquisition, Storm's figures do not show how much off-line revenues that had historically accrued to R could be diverted by a LV acquirer, or vice versa. There was no occasion for him to do so since he cast Chessie as the acquirer of both LV and R. There is thus no record basis for holding that the acquirers of the LV and R would have bid against each other for the purpose of changing L&NE traffic from its historic patterns. We thus conclude that there would have been no competitive bidding for the Bethlehem branch and that, even if it continued in rail use, the sales price would not have been significantly more than the applicable version of X. XIII. Central Railroad of New Jersey CNJ's contentions in this phase of the case are unlike those of any other T. CNJ alleges that, but for the Rail Act, it would have sold its entire rail system to New Jersey for approximately $100 million. It has two ways of reaching that result, however, and it presents what amounts to two separate rail use cases. CNJ's first ("real life") case is unique in this litigation: it attempts to prove a hypothetical sale to the State, based not on speculative contentions about what the State's interests and actions would have been had the Act never been passed, but rather on what the State actually did in the presence of the Act. According to CNJ, it and New Jersey carried on active negotiations that would have resulted in a sale but for the interference of this court in the 180-Day Appeals, 384 F.Supp. 895. In the event that its "real life" case should fail, however, CNJ has backed it up with a more conventional alternative scenario case — one that assumes that in the *1375 absence of the Act the State would have bought CNJ's lines for $100 million. A. Central Railroad of New Jersey's "Real Life" Sale to New Jersey In 1974 representatives of the State of New Jersey, the federal government, the estate in bankruptcy of CNJ, and some CNJ bondholders discussed on many occasions the prospect of a purchase of some or all of CNJ's properties by the State. According to CNJ, these negotiations made great progress toward an actual agreement during the course of the year, until they were aborted by this court's decision, in the 180-Day Appeals, that CNJ would be reorganized under the Rail Act. CNJ alleges that, but for that decision, it would have reached accord with New Jersey for a sale of all of CNJ's rail assets for a price in the range of $100 million. Accordingly, CNJ argues, we should assign CNJ a net liquidation value of $100 million, based on an actual, probable sales opportunity foreclosed by the intervention of the Act. The GPs mount two attacks on this contention, one factual and one legal. Factually, they argue that New Jersey never made any offer or decision to buy CNJ, for $100 million or for any other price. The negotiations CNJ describes, the GPs say, were merely exploratory on the State's part — an attempt to discover the possible size and shape of one among several options open to it. Legally, the GPs contend that even if CNJ's version of the facts is correct, the court should not take cognizance of them because the events in question all occurred in the presence of the Rail Act. The passage of the Act and the resulting infusion of federal activity and money into the Northeast rail system, they argue, must have had a substantial influence on New Jersey's alleged decision to focus its efforts on CNJ. We need not reach the legal issue, however, because we are persuaded by the GPs' factual argument. CNJ went into bankruptcy in 1967. Its trustee in bankruptcy, R. D. Timpany, who took office in January 1971, was in sporadic communication with New Jersey State officials throughout his tenure concerning such matters as the State's subsidy of CNJ passenger operations, a possible lease of passenger facilities to the State, and a proposed extension of the Port Authority's operations near Newark. Serious negotiations for a possible State purchase of CNJ did not begin until the election of Brendan Byrne as Governor in November 1973. The principal negotiator for the State was Alan Sagner, Governor Byrne's Commissioner of Transportation. Sagner spoke or corresponded with Timpany and, separately, with representatives of a committee of bondholders on several occasions in February, March, and April 1974. The parties agree, however, that these talks were merely exploratory as to any State purchase. On March 28, Sagner told Jacob Goodstein, a bondholders' representative, that he saw no possibility of a purchase in the immediate future. By April 3, when CNJ's reorganization court held a hearing to determine whether CNJ would be reorganized under the Rail Act or under section 77 of the Bankruptcy Act, the State had decided that it preferred a Rail Act solution. Nevertheless, Sagner continued into early summer to discuss the possibility of a purchase. He met with the bondholders and with the bankruptcy judge, and he requested advice from the Port Authority on how an acquisition might be structured. In late June 1974, a series of related events set off a small crisis and led to the resumption of active negotiations. CNJ's contract with the State for subsidies on its passenger operations was scheduled to expire on June 30. On June 27, CNJ petitioned USRA for permission to cease all rail operations. The following day the reorganization court issued its decision that CNJ would not be reorganized under the Rail Act. Faced with the possibility of a shutdown of CNJ, Sagner met with Timpany and federal officials to arrange a short-term solution to maintain the status quo while New Jersey and other parties appealed the district court's ruling to this court. The resulting agreement, approved by the district court on July 15, provided for modification of the State's passenger subsidies *1376 and for extensive State expenditures to rehabilitate CNJ's main line. The United States Department of Transportation agreed to provide an additional $3 million advance under § 213 of the Act. In return, CNJ agreed to a stay of the district court's decision, and it agreed to continue passenger and freight operations through November 30, 1979. The parties also agreed to enter into good faith negotiations looking toward a permanent viable solution for CNJ. Apparently taking this last point to heart, Sagner attended two meetings on July 18. The first, in Washington, was with several federal transportation officials. Sagner mentioned the possibility of a State purchase of CNJ, but pointed out that the transaction would require substantial federal financial help. The federal officials were noncommittal. The second meeting of the day, in New York, is a key element of CNJ's factual case. Sagner met with several representatives of the bankholders' group to discuss a possible purchase. It was at this meeting that, according to CNJ, Sagner, as New Jersey's Commissioner of Transportation, made an offer to buy the railroad. Timothy Crane, a member of the bondholders' committee, stated, both in a contemporary file memorandum and in later testimony, that Sagner presented a tentative offer to buy all of CNJ's properties, except for three parcels of off-line realty and certain court-controlled accounts. The State would assume an estimated $94 million in obligations and would pay $25 million in cash. The bondholder representatives, Crane said, reacted favorably to Sagner's proposal. Crane (S) (Dec. 31, 1978) Ex. 1, CNJ Reply App. at 11. Sagner, by contrast, stated in his file memorandum that the proposal Crane described was in fact presented by Jacob Goodstein, and that Sagner made no commitment on any price. He testified that "what I intended to convey to all of the people involved is that there were no proposals, there were no offers". Sagner (D) (Mar. 21, 1979) at 74, I CNJ App. at 180. Goodstein testified simply that "we discussed a proposed deal" matching the description given by Crane. Goodstein (S) (Dec. 31, 1978) at 7, I CNJ App. at 95. There are numerous reasons, even in CNJ's own evidence, why we must credit Sagner's recollection of the July 18 meeting over Crane's. On the next day, according to an aide's contemporary memorandum, Sagner told his staff that he was interested in the "possibility" of a State purchase of CNJ, but also considered the possibility that a Conrail solution would be better. Ex. CNJ-3, I CNJ App. at 147-49. On July 26, Sagner's office requested two independent valuations of CNJ — one by Sagner's own department, and the other by L. E. Peabody & Associates, a consulting firm. The instructions for the latter study show that Sagner was apparently still undecided as to the form of the transaction (purchase of assets vs. tender offer for bonds),[336] as to whether the State might itself operate the railroad after acquisition, and as to the proper basis for determining the purchase price. At a meeting with Goodstein on July 30 (according to Goodstein's account), Sagner again raised the question of the form of the transaction; he also revealed that he had not yet decided whether off-line realty should be included in a purchase, although he leaned toward exclusion. He raised possible political difficulties if the bondholders were paid in full. Finally, he mentioned that he had commissioned the Peabody valuation study and stated that any proposals he might make would be based on that study. Goodstein (S) (Dec. 31, 1978) at 9, I CNJ App. at 97. As late as August 23, Sagner expressed his expectation that *1377 CNJ's net liquidation value,[337] in his opinion, the only proper basis for a price, would be in the range of $70 to $80 million. In the face of this evidence, it is incredible that on July 18 Sagner committed himself, even "tentatively", to a detailed (and unwritten) offer of purchase at a price of $119 million. On the contrary, CNJ's own evidence bears out the inference that, if Sagner raised the "proposed deal" at all, he did so hypothetically in order that he might estimate the bondholders' position. Of course, it is not necessarily fatal to CNJ's case that no offer was made on July 18, for the record shows that Sagner continued through September to pursue the possibility of purchase. We have already mentioned his meetings with his own staff on July 19, with the Peabody consultants on July 26, and with Goodstein on July 30. On July 24, he met with Timpany and requested that Timpany propose an offer to sell that Timpany would be willing to submit to the reorganization court. Timpany complied on July 26; his offer proposed a sale, effective on September 1, 1974, of all CNJ properties except court-controlled accounts and three parcels of off-line realty. The State would pay the first $80 million in claims against the estate, plus $25 million in cash less a credit for any reduction the State might procure in claims held by the federal government against CNJ. Timpany (S) (Dec. 31, 1978) Ex. 18. A month later, at an August 23 meeting with federal officials, Sagner reportedly remarked that New Jersey then "contemplated" buying the CNJ and was preparing a bond issue for that purpose. On the other hand, it was at this meeting that Sagner said he expected to pay only $70 to 80 million, and that he learned that he could not obtain any waiver of federal claims against CNJ. Jacoby (S) (Dec. 31, 1978) Ex. 60 at 2. On September 12, Sagner again met with Timpany to review the problem. Sagner expressed concern about paying more than book value for CNJ's assets, but Timpany told him that no data on book value existed. Sagner also asked Timpany's advice on how the State ought to run the railroad if it acquired it. On September 25, Sagner met with USRA officials; they said that a State acquisition was probably better than a market solution, and that partial federal funding might be available under the Rail Act. They pointed out, however, that a federal acquisition was still among USRA's options. Three days later negotiations effectively stopped when this court decided that CNJ should be reorganized under the Act, 180-Day Appeals, 384 F.Supp. at 954-55, 987-89. This train of events substantiates CNJ's assertion that, up until that decision, New Jersey had shown great and increasing interest in purchasing CNJ.[338] Despite this, it is far from certain that the State would have actually carried out such a project. On the contrary, neither Sagner nor any one else in State government was ready to make any final decision on the matter. Although Sagner apparently had Governor Byrne's authority to "explore" the possibility of a purchase, there is no suggestion that he was ever authorized to close a deal, or even to make a tentative proposal. Nor is there any evidence that Sagner or his staff had drawn up any particular plans for acquisition or operation of the railroad. On July 26, Sagner asked the Peabody firm for advice on such basic questions as whether the State should buy freight-only properties in addition to passenger lines, whether it should buy CNJ's nonrail properties, whether New Jersey should operate rail service itself or contract to have private railroads *1378 do it, and whether it should make a tender offer for CNJ's publicly held bonds. With the possible exception of the last item, there is no record that Sagner ever reached any decision on these points. Nor did he ever request, let alone receive, any sort of study or estimate of the financial burden the State would incur in operating CNJ — surely a major factor in any decision to buy. It is hardly surprising, then, that Sagner never presented any definite proposal to buy CNJ to the Governor, members of the legislature, the State Attorney General or Treasurer, the Port Authority, the United States Department of Transportation, USRA, or any of the other public agencies that one would expect to participate in such a decision. Moreover there was no consensus, either within Sagner's department or between Sagner and the various CNJ parties, as to an appropriate price. The existence of substantial uncertainty as to the sale price and properties to be included in the transaction casts further doubt on CNJ's claim that an agreement had nearly been reached. CNJ asserts that the price would have been "in the $100 million range", CNJ at 3, but it appears that that range extended quite far on either side of the mark. We have found ten or more figures in the record, ranging from a low of $70 million to a high of nearly $120 million, representing the proposals, suggestions, predictions, and after-the-fact estimates of the many persons involved as to what the sale price would be.[339] CNJ's suggested figure of $100 million, then, could be 15 percent too low, or it might be 43 percent too high; there is precious little in the record to tell us where within this $48 million range of possibilities the final result would have fallen. Even that does not reflect the full degree of uncertainty we face. In the first place, there is the problem as to how much of CNJ's property the State would have bought. Perhaps more important is that not all of the many prices mentioned (for example, neither of the proposals by Goodstein or Timpany) can readily be adjusted to exclude the value of CNJ's assets not conveyed to USRA. In short, although the record suggests that Sagner pursued and probably favored a State purchase, it does not bear out the allegation that New Jersey was on the verge of buying CNJ's entire system. On the contrary, we cannot conclude with any confidence that any State purchase would have occurred at all. There are simply too many unknowns, too many decisions not yet made, too many important actors still absent from the stage, to permit that. *1379 B. Central Railroad of New Jersey's Alternative Scenario CNJ's alternative scenario differs little from the supposed "real life" deal we have just discussed: once again, CNJ alleges a sale of all of its rail properties to New Jersey for $100 million. The most important factual difference between the two sets of contentions concerns timing. According to CNJ, its "real life" sale would have taken place sometime late in 1974. In the absence of the Rail Act, however, CNJ contends that the sale would have occurred a year earlier, in the fall of 1973. 1. Purchase for Rail Use We accept CNJ's contention that, in the absence of the Rail Act, New Jersey would have bought all of CNJ's conveyed properties. CNJ's evidence shows that its commuter lines including the New York & Long Branch Railroad, which CNJ owned jointly with PC, carried 25,000 passengers a day into the New York metropolitan area. New Jersey had already recognized the public importance of this service, providing many millions of dollars in subsidy payments and indirect financial assistance. In 1972 the State had negotiated seriously for a possible lease of CNJ's passenger facilities. CNJ at 28. The State similarly recognized the importance of CNJ's freight service. For example, in 1971, the Commuter Operating Agency found that service "essential", remarking that its loss would be a "serious hardship" to shippers and their employees and communities. Timpany (S) (Dec. 31, 1978) Ex. 6 at 5-6. These State findings were confirmed by a 1971 decision of the United States Secretary of Transportation, Timpany (S) (Dec. 31, 1978) Ex. 3. CNJ served an important function as a sort of state-wide freight terminal carrier. In 1973 it carried 200,000 carloads (mostly originating or terminating traffic), serving some 1,400 customers both on- and off-line. Much of this traffic relied exclusively on CNJ. In the absence of any prospect of private acquisition,[340] then, we find it likely that New Jersey would have intervened to save CNJ's passenger and freight service by purchasing CNJ's rail properties for continued rail use. The GPs make no attempt to show that any particular CNJ line or facility would not have been acquired. See Part III, p. 25, supra. Their main attacks on CNJ's alternative scenario are two that we have already addressed and rejected — the contentions that New Jersey lacked the legal authority and financial capacity to make large rail purchases. The GPs also point out material discrepancies between CNJ's proffer and those of other Ts, especially the JTTs.[341] To point out inconsistencies, however, is not to resolve them in GPs' favor. In the absence of any evidence that any particular line would not have been acquired, we hold that GPs have not rebutted CNJ's showing that its entire system would have been purchased for rail use. 2. Timing The boldest feature of CNJ's alternative scenario is its assertion that the transaction would have been completed in the fall of 1973. We have already noted the key difficulty in claims of early, independent dispositions: the likelihood that any sale must await the outcome of a comprehensive regional rail planning process. As we said in the CUE Opinion, "we think it unlikely that the ICC would have tied its hands early in the proceeding by permitting abandonment or sale of major lines even before it knew what shape the surviving system might take". 439 F.Supp. at 1375. CNJ seeks to escape this mire by arguing that it would have entered the abandonment process so early, and completed its transaction so *1380 quickly, that all would have been over before anyone could have seen the need to coordinate CNJ's disposition with those of other bankrupt carriers. We find this contention implausible. CNJ alleges that it would have been the first T to seek abandonment.[342] CNJ had been in reorganization since 1967, longer than any other T. It had not reported an annual profit since 1957; by 1973, it showed a negative net worth. It was in arrears on interline balances and on scheduled wage increases for its employees. It had already begun disposing of some of its facilities and operations; in 1971-72, it halted its New York harbor lightering operations, its Jersey City coal dumping activities, its entire Pennsylvania freight operation, and its service on several small branch lines. In late 1972 Trustee Timpany staved off the latest of several attempts by the bondholders to force a shutdown of all rail operations. At the same time, however, Timpany himself petitioned the reorganization court for permission to cancel CNJ's subsidy agreement with the State and to terminate all passenger service if the State would not agree to a more generous subsidy. The district court granted that relief in December 1972, but the Court of Appeals for the Third Circuit, sitting en banc, reversed, In re Central Railroad of New Jersey, 485 F.2d 208 (3 Cir. 1973), cert. denied, 414 U.S. 1131, 94 S.Ct. 870, 38 L.Ed.2d 755 (1974). In the absence of the Rail Act, according to Timpany, he would have responded to this reversal by filing immediately in the district court for permission to abandon all operations. CNJ contends that this petition would have been virtually unopposed, and would have been granted with very little delay. Accordingly, New Jersey would have stepped in with an immediate purchase to avoid cessation of service. We find several serious flaws in this thesis. The first is CNJ's assertion that it would have been the first of the bankrupts to seek liquidation. Indeed, for CNJ to prevail, it must show not only that it would have been the first, but also that it would have led by a margin of at least several months — long enough to allow it substantially to complete its disposition before other parties realized the need for an overall regional solution. CNJ has foreseen our conclusion that PC and other Ts would have filed for abandonment by February 1973, but it contends that this is more a matter of concession by the GPs than of factual proof. CNJ argues that it is entitled, if it can, to show as a matter of fact that the other bankrupts would not have sought to close down in 1973. We need not pass on CNJ's characterization of our conclusions, however, because there is a glaring gap in CNJ's own attempted proof. Although CNJ presents detailed argument as to whether LV, EL, and R would have sought liquidation before CNJ did, it makes no effort at all to show that PC would not *1381 have done so.[343] PC was the key to the Northeastern rail problem; it was by far the most important carrier, both in the region as a whole and in New Jersey in particular. In the absence of any evidence, or even any square allegation, that PC would not have applied for abandonment, we cannot credit CNJ's contention that it would have received authority for abandonment before any other major liquidations commenced. Even if CNJ would have been the first to seek liquidation, we find no warrant for its contentions concerning the rapidity with which it could have completed the abandonment hearing and the consequent sale to New Jersey. The key to the argument is CNJ's allegation that it could have obtained speedy abandonment authority from the reorganization court without having to go to the ICC or any New Jersey regulatory agency. This seems unlikely in view of the Third Circuit's en banc opinion dealing with CNJ's petition to abandon its passenger operations. In re Central Railroad of New Jersey, supra. The court applied the requirement of section 77(a) of the then-current Bankruptcy Act, 11 U.S.C. § 205(a) (1970), that a reorganization court may approve a trustee's petition for permission to abandon all or part of a line "only with the approval and authorization of the Commission when required by [the Interstate Commerce Act]". The court held that CNJ, if subject to the ICC's jurisdiction, could not abandon service without ICC approval. 485 F.2d at 211-12. Even absent ICC jurisdiction, the court held, CNJ must seek approval from the Commuter Operating Authority, a State agency. Id. at 212-13, citing Palmer v. Massachusetts, 308 U.S. 79, 60 S.Ct. 34, 84 L.Ed. 93 (1939). The court rejected CNJ's argument that it was constitutionally entitled to abandon without seeking regulatory approval because of financial erosion. Since CNJ had not even applied to any agency, the court reasoned, it was not allowed to attempt to show that its application would meet undue delay or frustration. The court distinguished Brooks-Scanlon Co. v. Railroad Commission, 251 U.S. 396, 40 S.Ct. 183, 64 L.Ed. 323 (1920), and Railroad Commission v. Eastern Texas R.R., 264 U.S. 79, 44 S.Ct. 247, 68 L.Ed. 569 (1924). In Eastern Texas, the issue was the carrier's substantive right to abandon, not the procedures it must follow. In Brooks-Scanlon, the procedural issue arose as a collateral point, but the Supreme Court disposed of it by noting that the regulatory agency had already indicated its refusal to grant the requested relief. 251 U.S. at 400, 40 S.Ct. at 184. Instead, the CNJ court relied on Palmer v. Massachusetts to hold that the bankruptcy court could not bypass the administrative agencies. 485 F.2d at 213-15. In the face of this holding, which we approved in the CUE Opinion, 439 F.Supp. at 1369, 1372 n.34, it is hard to indulge the assumption that CNJ's reorganization court would have repeated its error by allowing a total abandonment without referring the petition to the ICC and the Commuter Operating Authority. CNJ tries to evade the Third Circuit's holding by suggesting that its petition to abandon all service would have been unopposed in the district court. This is a non sequitur; we can not assume that the consent of the parties would have led the district judge to disregard his statutory duty. Even granting that point, however, we find CNJ's allegation to be unsupported. CNJ relies on evidence that purportedly shows that an attorney representing the ICC and the Justice Department conceded *1382 the district court's jurisdiction to decide the bondholders' motion in December 1972 to shut down all of CNJ's operations. Timpany (S) (Dec. 31, 1978) at 83-84. From the excerpt CNJ presented, however, it is hard to tell just what the attorney was conceding; he may have meant only that the district court should hear the bondholders' constitutional claims. Moreover, whatever he conceded, he conceded reluctantly; almost in the same breath, he asked the court to delay any shutdown to allow the ICC and other agencies to make alternative service arrangements. It is unlikely that the ICC would continue to concede the court's jurisdiction totally after the Third Circuit's ringing affirmation of the ICC's authority — especially since the need for a comprehensive regional solution could only have become clearer in the intervening months. CNJ also points to the ICC's statement in October 1973 that it was not the proper forum to decide issues of unconstitutional takings. CNJ at 119 n.* Again, however, that is not a concession of all jurisdiction; rather, it states the obvious point that the reorganization court retains authority to act in cases of undue delay, frustration, cashlessness, or similar emergencies. Even had the ICC been as accommodating as CNJ would like, it is improbable that New Jersey would have been the same. The State did fight the district court's decision to bypass the ICC and the Commuter Operating Authority in December 1972; indeed, it was a successful appellant in the Third Circuit's en banc reversal of that decision. There are several reasons to suppose that New Jersey would have continued to press the public interest in continued rail service, both before the ICC and in the district court.[344] By September 1973 the State, like the ICC, would have perceived the need for comprehensive planning; even if the ICC had not acted, New Jersey would have sought time to develop its own overall solution within the State. The need for caution would have been multiplied by the lame-duck incumbency of Governor Cahill between his primary defeat in June 1973 and Governor Byrne's inauguration in January 1974. Several witnesses, including Timpany, testified that Cahill's policy was to leave all major policy decisions to the successor administration. 4 GP at P 141-42. Furthermore, even if New Jersey had decided in September 1973 to buy CNJ, it does not follow that it would have agreed to a court order relieving CNJ of its duty to operate. The State surely would not have irrevocably committed itself to a purchase until it had reached agreement on a purchase price — a time-consuming matter. There is no reason why it would have volunteered to accept all interim operating losses during negotiations, if it could have imposed those losses on CNJ's stockholders and creditors by delaying or defeating CNJ's abandonment petition. The conclusion we reach is that CNJ could not have escaped entanglement in a complex regional planning process. 3. Price All parties agree that any sale of CNJ to New Jersey would have taken place at a price equal to what we have called X. Predictably, however, they disagree as to what X is. Of more immediate importance, they disagree as to what issues we should decide concerning it in this phase of the case. CNJ urges us to decide that the proper measure is "stand-alone" net salvage value, which disregards any possible effect of glut resulting from liquidation of several northeastern railroads at about the same time. That is not all, however; CNJ further urges *1383 us to fix an actual dollar amount for its sales price, based on several studies of CNJ done between 1969 and 1974.[345] For reasons developed in Part III of this opinion, we agree with the GPs that these questions must be left for decision in the nonrail use phase of the case. Petitions for reconsideration may be served and filed on or before 21 days from the filing of this opinion.[346] While it is not setting any page limit, the court requests that such petitions be limited to instances where the court is thought to have misapprehended evidence or to have erred in computations and not repeat arguments that have been made at length in the opening briefs and reply briefs and in oral argument and have been fully considered. Answers to such petitions shall not be filed unless requested by the court. GLOSSARY AA Ann Arbor AAR Association of American Railroads AREA American Railway Engineering Association AWCC Average weighted cost of capital B&M Boston & Maine Railroad BTF Bridge track feet CERL Combined Erie, Reading and Lehigh Valley CMV Constitutional minimum value CNJ Central Railroad of New Jersey CRA Charles River Associates CUE Compensable unconstitutional erosion DCF Discounted cash flow D, T&I Detroit, Toledo & Ironton EBIT Earnings before interest and taxes EL Erie Lackawanna ELMPA Erie Lackawanna Merger Protection Agreement FAHA Federal Aid Highway Act FAUS Federal Aid — Urban System Program FRA Federal Railroad Administration FSP Final System Plan GPs Government Parties ICC Interstate Commerce Commission IFLs Intercity freight lines JT Joint Terminal JTTs Joint Terminals Transferors LIRR Long Island Railroad L&NE Lehigh & New England LPCs Labor Protection Costs LV Lehigh Valley MBTA Massachusetts Bay Transportation Authority MGT Millions of gross tons per year MOWE Maintenance of way expenses MTA Metropolitan Transportation Authority NEPA National Environmental Policy Act NJ/NYTerm New Jersey/New York Terminal PC Penn Central PCMPA Penn Central Merger Protection Agreement PDV Present discounted value Philaterm Philadelphia Terminal R Reading RC Rehabilitation costs SEPTA Southeastern Pennsylvania Transportation Authority S&P Sverdrup & Parcel and Associates TOFC Trailer on Flat Car T Transferor UMTA Urban Mass Transportation Act USRA United States Railway Association WJPA Washington Job Protection Agreement X See p. 1213 & n.23, supra. On Petitions for Reconsideration We deal in this opinion with petitions for reconsideration of our Rail Use Opinion dated November 24, 1981, which were filed on December 29, 1981, by the Trustee of the Lehigh Valley Railroad (LV) and the GPs.[1] We called on the adverse parties to answer these petitions on January 11, 1982. The LV Petition A. Valuation Methodology The LV petition raises three points concerning Part VII of the opinion entitled Valuation Methodology. These are (1) that "the conservative assumption in the EL-LV pro forma that there would be no growth in cash flows after the terminal year should *1384 result in a lower cost of equity than 20%"; (2) that the inflation factor used by the Court in deriving the discount rate was understated; and (3) that use of the DCF methodology requires that the EL-LV cash flows be restated to account for tax savings resulting from depreciation. We find no merit in any of these points. (1) LV's argument is that in using a 20% cost of equity for EL-LV, which was derived in part from PC witness F. Smith's figure of 22.5% for the IFLs, we failed to take account of the fact that whereas Smith's cash flows were projected to increase after the terminal year, to wit, 1985, at an annual rate of 2%, witness Shannon did not include any growth in the EL-LV traffic base after the fifth year.[2] This, LV argues, is an element of conservatism which should be reflected by a reduction in the cost of equity to 19%. Apart from other considerations, this claim shatters on the rock, pointed out by the GPs (Answer, pp. 5-7), that Shannon's projections for EL-LV were actually less conservative than those for the IFLs. The projections for the IFLs were stated in nominal dollars and, as LV correctly points out, contained the assumption that revenues, in nominal dollars, would grow at a 2% rate. When combined with the effect of long-term inflation, which the parties generally projected to be from 5-6%, see GPs' Answer, p. 6, the real cash flow was not projected to keep up with inflation. See Myers (S) (PC) (Dec. 1, 1978) at 15; F. Smith (D) (Aug. 22, 1980) at 1818, 19 GP Supp.App. at 346. In contrast, the EL-LV projections, which were stated in real dollars, assumed that cash flow would remain constant in real terms. Thus the EL-LV projections were in this respect riskier than those for PC since EL-LV assumed that cash flow would keep up with inflation. (2) LV's second point is that we erred (Opinion at 66) in using 5.58% as the factor for deflating the 15.22% nominal AWCC to derive a real AWCC of 9.13%. The 5.58% figure was the arithmetic average of the yearly inflation rates forecast by Dr. Olsen for the IFL projections. Smith (S) (GPs) (Jan. 30, 1980) at 154-55. These rates were 6% for the period 1974-80 and 5% for the period 1981-85. LV's argument is that since Shannon's projections were made for only five years, we should have used Dr. Olsen's 6% figure. This contention ignores the nature of a capitalization of earnings on the discounting of cash flows. Even when only the terminal year's figures are selected, the assumption is that these will continue. As the GPs put it (Answer, p. 11), "[c]apitalization of the fifth-year earnings for EL-LV has the obvious effect of extending the projections in perpetuity." (3) LV's third claim is that in translating the Shannon pro forma forecasts of earnings for the fifth year of operations into forecasts of cash flows, the court erred in using Shannon's estimates of depreciation rather than what LV contends would have been higher figures that would have been used in a cash flow analysis. The argument is that "both the buyer and seller of EL-LV, in determining the most they would be willing to pay and accept, respectively, would have determined the most that the buyer could get from the property" (Petition, p. 12); that in an earnings analysis this would be achieved by using the lowest depreciation rates consistent with generally accepted accounting principles; but that in a cash flow analysis the acquirer would adopt the highest rates permitted under the income tax laws since depreciation is not deducted from cash flow but higher depreciation will result in lower taxes. Also, under an earnings analysis, the purchase price will be allocated to nondepreciable assets so far as permitted by the contract and generally accepted accounting principles; per contra when the analysis is on a cash flow basis. LV has filed with its petition *1385 for reconsideration twelve pages of new calculations which purport to demonstrate that the latter adjustment alone would enhance the EL-LV cash flows for a five year period by $19,735,000. We agree with the GPs that this contention comes too late. The GPs argued in their initial submissions that cash flow rather than earnings constituted the proper valuation methodology, F. Smith (S) (Jan. 31, 1979) at 19-23, as indeed PC had also done, Hillingby (Dec. 1, 1978) at 21. This has always been a major issue in the case. See Opinion at 49-55. LV could not simply sit back and assume that the court would decide in favor of the capitalized earnings method. Indeed, our letter of June 19, 1981, Sp. Ct. Rep. N 36204, expressly requested EL and LV to provide us with cash flow projections based on their pro forma earnings statements, and they did. If LV considered that lower figures for depreciation ought to be used should the court opt for the cash flow method, it was bound both to raise the point and to quantify it by presenting calculations which the GPs could challenge. The GPs question many of the assumptions and computations which LV makes in its present effort to show that larger depreciation deductions would have greatly increased EL-LV's after tax cash flow. We do not reach such questions since, as a result of LV's failure adequately to raise the point that it would have had higher depreciation deductions and consequently lower income taxes on a cash flow method during the trial or even during the post-trial submissions (Opinion at 15), the record is not adequate for us to decide them. B. Labor Protection Costs LV also asks us to reconsider three aspects of Part IX of the opinion, relating to Labor Protection Costs. (1) Pre-Day One Attrition The first is our refusal to take account of "pre-day one attrition" (i.e., attrition prior to the conveyance date) beyond the six month period reflected in Table 35, Opinion at 156. This was a considered decision made partly in the interests of conservatism, Opinion at 163, and partly because, as stated in note 175, p. 1ii, "[t]he LPC studies made by the parties did not attempt to quantify the effect of a later acquisition date on LPCs" and the record was "devoid of evidence that would permit us to make such a calculation ...." We see no reason to depart from that decision. LV's own witness, W. Arthur Grotz, made no allowance for pre-day one attrition, Grotz (S) (Oct. 15, 1980) at 13-14; Grotz (D) (Jan. 22, 1981) at 186-87, in part because he thought that LV would need its quota of employees to run the railroad until the ICC permitted abandonments. Grotz (D) (Jan. 22, 1981) at 125-26, 723; Ex. GPS-Grotz-Labor (X)-16, 25 GP Supp.App. 122P, 251A. While LV might have been able to persuade a prospective purchaser that pre-day one attrition would have produced lower LPC's than those estimated in our Opinion, we find no basis in the record or in good sense for thinking that this would have been of an order of magnitude approaching the 50% which LV now asks us to adopt. (2) Transfer of Displaced LV Employees to Parallel EL Lines in New York LV urges that Grotz's estimates of displaced LV employees were based on a LV "stand-alone" hypothesis and require adjustment in light of our conclusion that the most likely outcome for LV was to sell its lines east of Waverly, N.Y., as part of an EL-LV purchase. The employees on the abandoned lines west of Waverly, LV alleges, would simply have moved over to the "parallel" EL lines in order to handle the traffic contributed to those lines by the LV lines east of Waverly. Shannon gave brief testimony to this effect. Shannon (S) (Oct. 15, 1980) at 15. In support of this LV points to our acceptance, Opinion at 141, of EL's contention that employees surplussed as a result of abandonment of its Scranton line would move over to LV's Waverly-Newark line. We note initially that, notwithstanding his unexplained statement to the contrary, *1386 Grotz's estimate of the number of positions that would have been required to operate the LV system on Day One is applicable to the EL-LV acquisition as well as to the LV stand-alone case. Grotz testified that he derived his estimate of the manning requirements of the LV system from "the actual employment of former LV personnel by the Trustee after conveyance to Conrail in 1976". Grotz (S) (Oct. 15, 1980) at 16. He justified this methodology on the ground that "LV contends that the same properties would have been kept in rail use absent the Rail Act as were in fact conveyed to Conrail". Id. (emphasis added). On this theory, however, Grotz's conclusions regarding manning requirements in the LV stand-alone case would be equally applicable to the EL-LV acquisition, since, as the GPs point out (Answer, p. 23), there was virtually no difference in the "properties [that] would have been kept in rail use". Thus, the Opinion's reliance on the Grotz study in calculating the number of employees that would have been required to operate LV's lines in the EL-LV acquisition was proper. There remains, however, the question whether the Court improperly relied upon Grotz's treatment of the possibility of the transfer of LV's New York State employees to EL's lines. Grotz did not think, as Shannon did, Shannon (S) (Oct. 15, 1980) at 15, that LV's New York State employees would be transferred wholesale to EL's New York lines, but instead testified that these employees would have been rehired into attrition-opened positions in the LV system, Grotz (S) (Oct. 15, 1980) at 16, 24, 31-32. On the other hand, Grotz's testimony was addressed to LV's "stand alone" case and there is no indication that he focused on an EL-LV package. LV's Petition thus raises a legitimate point. In most of the parties' discussions of LPCs, the Ts were treated largely as individual units. After we accepted the EL-LV package, however, the possibility arose of two Ts, i.e., LV and EL, cooperating to reduce one another's LPCs, and Shannon did testify, although only in a conclusory fashion, on the issue now raised by LV. We think that in EL-LV's negotiations with an acquirer some allowance would have been made for the possibility of cooperation between EL and LV to reduce total LPCs. The acquirer, of course, would not have accepted arguments based upon a premise that vacancies in one system would be filled by employees from both systems, which would be a form of double-counting. In addition, because of possible difficulties — relating, for example, to geographical restrictions, union restrictions and timing considerations — in effecting "inter-corporate" transfers, we think potential acquirers would have demanded an extremely conservative approach to this issue. Based on the foregoing considerations, we believe that only a limited number of LV's New York State employees would have been transferred to EL's parallel lines. We think an acquirer would have been willing to treat roughly a third of the 299 LV Maintenance of Way, Maintenance of Equipment and Train and Engine employees in New York State as able to find work on the EL lines. This would reflect an acquirer's uncertainty regarding "inter-corporate" transfers. Moreover, it would permit a rough equivalence of transfers between EL and LV, thus avoiding the need to "settle accounts" between EL and LV, since in the Opinion we decided that 90 EL employees would find work on LV's lines, Opinion at 141. We conclude therefore that 90 LV employees would also have found work on EL's lines. This requires reducing the figures for 4/76 in Table 37 from 281 to 191, and, following this change through Tables 38-42, results in subtracting separation payments, dismissal allowances, and fringe benefits attributable to these employees from our calculations of LV's LPCs. The parties should have no difficulty in making these calculations. If this conclusion still is less favorable to LV than our corresponding *1387 treatment of EL, a sufficient answer is that the latter was unchallenged.[3] (3) Dismissal Allowances The Opinion decided that the labor protection benefits that would have been accorded the Ts' employees would have been reached through negotiations and would have been based roughly on the pattern of the New Orleans conditions, Opinion at 1267-1270. LV's final point with respect to Part IX of the Opinion is that we "departed from [the New Orleans] pattern in certain significant respects". Petition, p. 26. Although LV's observation is correct, that departure was the result of a considered decision. As our discussion of the New Orleans conditions, Opinion at 1269, indicates, we were well aware that those conditions provided only for "residual" WJPA benefits in the fifth year following acquisition. However, as emphasized, Opinion at 1269, we said only that the labor protection bargaining process "would have settled on levels of labor protection roughly equivalent to the New Orleans conditions", (emphasis added), and that the parties would have reached a "compromise based roughly on historical levels of labor protection". Id. at 1268 (emphasis added). We similarly departed from or added to the New Orleans conditions with respect to non-union employees and an acquirer's transfer rights. Id. at 1369-1370. Thus, while occasionally we used the terms "New Orleans conditions" and "WJPA conditions" as shorthand references, our calculations were based upon an intentionally simplified form of the New Orleans conditions, e.g., Opinion at Tables 28 & 40. We recognized that this increased the Ts' LPCs, yet thought this consistent with the general historical trend towards increased labor protection benefits, discussed in the Opinion at 1267-1268. We see no reason to alter our original approach. C. Trackage Rights Fees In section XI.B.5 of the Opinion, we recalculated the trackage rights fees that would have been charged by the public acquirer of EL's New Jersey passenger lines to the private acquirer of the EL-LV combination, so as to change this from EL's figure of $733,000 per annum to $3,097,888. Opinion at 1343-1346. LV claims that our estimate of the maintenance component of the trackage rights fees overstates the total costs that would have been borne by a private acquirer. The initial step in LV's argument is that the Court erred in rejecting the $4.3 million maintenance cost figure contained in the study of EL passenger avoidable costs for 1973, Opinion at 249-50. In light of LV's explanations, which the GPs do not contest, we agree. It follows that Shannon's assignment of $4 million in maintenance costs to the public acquirer of the passenger lines was not excessive since New Jersey had already been paying $300,000 more. Since LV does not contest that passenger service would only have borne approximately 44% of the total maintenance costs for the passenger lines, a $4.3 million share for the state suggests that the total is approximately $9.8 million. Thus the maintenance costs properly chargeable to the freight acquirer through trackage rights fees would be the difference between $9.8 million and $4.3 million, approximately $5.5 million. At first blush, as contended by the GPs (Answer, pp. 31-32), this would justify an upward rather than a downward revision of the trackage rights fees. However, LV argues that the trackage rights fees should be revised downward because the difference *1388 between this higher total maintenance cost, $9.8 million, and the total maintenance cost assigned by Shannon for both freight and passenger operations over the passenger lines, approximately $4.8 million, has already been charged to the acquirer of EL-LV. The reason is that Shannon computed the maintenance costs over the freight lines by subtracting his $4.8 million estimate for maintenance costs over the passenger lines from the total reflected in EL's 1973 Form R-1 as adjusted for normalization. Thus, if maintenance costs over the passenger lines should be figured at $9.8 million, rather than $4.8 million, the deduction of the higher figure from Shannon's estimate would result in maintenance costs on the freight lines being $5 million lower than assessed to EL-LV. Any upward adjustment in trackage rights fees would thus be accompanied by an equivalent diminution of maintenance costs on the freight lines. LV's argument incorrectly assumes that in accepting EL's maintenance estimates for the freight lines over those of the GPs, Opinion, Part VIII, at 101, we endorsed the methodology behind EL's figures. We were not approving as adequate Shannon's adjustments in EL's 1973 maintenance costs to bring them to a "normalized" level, see Opinion at n.112, since the factual basis for computing the magnitude of those adjustments was not set forth in his testimony. Rather, the Opinion at 101 noted that the GPs' maintenance cost estimates, although far more reliable than their rehabilitation cost estimates, were overstated to some unquantifiable extent. Since the GPs' and EL's estimates were reasonably close, we accepted the latter. This conclusion derived support from the comparison of EL's rehabilitation and maintenance estimates with those contained in the Chessie Study, Opinion at 1249-1250. Although the Chessie study was not as detailed as that performed by EL, we regarded it as a sound basis for comparison because it was performed by a potential acquirer. We did not find the fact that EL's 10-year total for maintenance and rehabilitation cost was only 85% of Chessie's estimate to be a sufficient basis for discarding EL's figures. However, the GPs point out, Answer, p. 30, that the reduction in maintenance costs on the freight lines implicit in LV's new claims suggests that EL's 10-year total for rehabilitation and maintenance costs on the freight lines is only on the order of 68% of that suggested by the Chessie Study[4] — a divergence of over twice the magnitude originally contemplated. We surely would not have accepted an estimate so much lower than that in the Chessie study. In sum, we accepted Shannon's estimate for what it purported to be, a forecast of maintenance of way expenses on EL's freight lines — not, as LV would now have it, of such expenses plus an additional amount for freight service over the passenger lines. While we do not totally reject LV's suggestion that what appears to be a newly emergent consensus between the parties on a far higher estimate of maintenance costs on the passenger lines may not imply that maintenance costs which Shannon allocated to the freight lines were somewhat too high, it is too late in the day to undertake the development and analysis of new studies on this subject. We thus will not disturb the conclusion with respect to trackage rights fees reached in the Opinion. That estimate was based upon the computations of EL's own witness as to the respective maintenance costs of the passenger and freight lines, which LV adopted, see note 2, supra; we disagreed only as to the proportion of the costs on the passenger lines properly attributable to freight service. To accept LV's present argument would suggest that the calculations EL commissioned Shannon to make — in determining both the quantity of maintenance costs that should be attributed to the New Jersey passenger lines and the allocation of that quantity *1389 between passenger and freight use — were completely unnecessary since EL's 1973 maintenance costs could have been adjusted directly by deducting the $4.3 million figure from the passenger avoidable cost study for the State of New Jersey. We see no unfairness in holding LV to what is effectively its own study at this stage of the case in the absence of any reasonable alternative. D. New Jersey Taxes In n.310 of the Opinion we indicated our agreement with the GPs that EL-LV expenses should be adjusted to include additional New Jersey property taxes for the LV of $403,000. This figure represented the difference between $175,000 which the GPs asserted were included in the unit cost calculations for EL-LV and LV's total New Jersey taxes of $588,000.[5] The GPs found the $588,000, number in Exhibit LVR-Adik-1 at SR 45 and also in LV's 1973 Form R-1, and suggested that the only appropriate modification was "any amount applicable to the small portion of LV's New Jersey property not included in EL-LV." Comments of Government Parties on Erie Lackawanna's Response to the Court's Letter of September 4, 1981, at 12, Sp. Ct. Rep. at N 37063. LV contends there are figures in the record showing directly the taxes on the LV properties that were conveyed, LVR-Adik-1, App. A. This lists seven segments and states that "Apportioned NJ Taxes" totalled $232,101. Consequently, LV urges the added New Jersey taxes should be only $232,101 — $175,000, or $57,101. The difficulty with LV's contention is the lack of any explanation how the $232,101 figure was arrived at. In the absence of this we see no reason to depart from the figure reported in LV's 1973 Form R-1. It was LV's burden to show how much of this related to unconveyed properties; it did not discharge this. The GPs' Petition The GPs' Petition relates solely to our computation of LPCs for LV and AA. A. LPCs for the LV (1) Day One Dismissals The first point raised by the GPs' petition relates to the Court's calculation of the number of LV employees that would have been dismissed on "Day One" — the first day following disposition. The GPs contend that the Court inappropriately treated LV witness Grotz's estimate of the number of LV employees dismissed on Day One as if it represented the number of LV positions eliminated on Day One. Thus, the GPs conclude that the downward adjustment that the Court applied to Grotz's estimate in Table 37 to arrive at the number of "employees dismissed" was improper, because Grotz had already calculated "employees dismissed". At several points the Opinion did distinguish between "employees dismissed" and "positions eliminated". This distinction was intended as a shorthand reference to the Court's application of pre-day one attrition, post-disposition attrition, and rehire rates. "Employees dismissed" was equivalent to "positions eliminated" following application of these factors. We find no indication in Grotz's testimony that he attached a similar meaning to the phrases, although we recognize that the Opinion at 1286 can be read to the contrary. Most important, however, is the fact that Grotz's estimate of employees dismissed on Day One did not reflect adjustments for pre-day one attrition, post-disposition attrition, or rehiring. Rather, his estimate reflected only adjustments for unprotected employees and employees over age 64. Grotz (S) (Oct. 15, 1980) at Table 1. Thus, although the GPs might disagree with our decision to account for pre-day one attrition or with the timing of post-disposition rehiring, our methodology did not, as the GPs' petition suggests, result in any double-counting of these factors. *1390 Grotz's estimate that 389 LV employees would be dismissed on Day One was based upon his calculation that on Day One LV would have had 2,487 protected employees and only 2,098 positions available for these workers, Grotz (S) (Oct. 15, 1980) at Tables 1 & 2. Grotz did not adjust his estimate of the number of protected LV employees on Day One to reflect pre-day one attrition, id. at 13, a decision we did not accept, Opinion at 1288. Our adjustment for pre-day one attrition, which had the effect of reducing the number of protected LV employees on Day One, therefore involved no double-counting and was entirely appropriate. Grotz did attempt to reflect post-disposition attrition and rehiring of dismissed employees, Grotz (S) (Oct. 15, 1980) at 23-25 & Table 7, and adjusted his estimate of "employees dismissed" to reflect these factors, id. Although in calculating the LPCs of both EL and LV we accepted Grotz's general methodology, in both instances we treated the timing of attrition and rehires differently from the Grotz study, e.g., Opinion at 1286. As Tables 1-6 (regarding EL) and 34-37 (regarding LV) of the Opinion indicate, we computed dismissal allowances and lump sum payments on the basis of the number of employees eligible for benefits at the end of the period under consideration. We recognize that we could have treated the question differently. Indeed, we said in the Opinion at 1271 that "any one of numerous methods of calculating LPCs could be adopted." We do not think our considered choice of the timing of post-disposition attrition and rehiring was improper. Our treatment of attrition and rehiring took a consistently conservative view with respect to the usefulness of attrition, Opinion at 1274-1275, 1286-1287, 1290. The fact that one aspect of our methodology could have been made even less favorable to the Ts is no basis for rejecting the conclusions reached in the Opinion. Implicit in our acceptance of Grotz's methodology was our recognition that LV could have used attrition and rehiring to reduce its LPCs, and we think our treatment of the issue was appropriate. (2) Alleged Inconsistency Between Tables 36 and 43 for Period Ending 10/81 The GPs' second claim is that in calculating the LPCs attributable to the relocation of LV employees we inadvertently substituted the figure 39 in the final line of Table 43 when we intended to use the figure 79 which appears in Table 36. As our footnote to Table 43 indicates, the figures that appear in that table are drawn from Table 40 as well as Table 36. The figure 39, which reflects the number of employees that would have been rehired in the period beginning 10/81, is derived from the column headed "10/81" in Table 40. That column indicates that, although 79 positions were available for surplussed employees, see Table 36, there were only 39 LV employees still receiving dismissal payments, and thus only 39 employees could have been rehired. Accordingly, we reject the GPs' contention. (3) Displacement Allowance We find more merit to the GPs' claim that the court's estimate of the cost of displacement allowances for LV employees was understated because of our reliance on the FSP and Grotz's testimony that Penn Central paid only $76.28 to each displaced employee. Grotz (S) (Oct. 15, 1980) at 30. The GPs point out that, while Grotz cited the $76.28 per employee figure, he actually used a $533.28 figure in his calculations of LV's LPCs. They further note that no expert used the $76.28 figure, which reflected costs being incurred five and six years after the PC merger. LV responds that the Court properly relied on PC's historical experience, and directs us to Shannon's testimony that displacement allowances would have been minimal. We accept the GPs' point, both because we intended to accept Grotz's study on the issue, and because we think the $533.28 figure is far more reasonable than the other estimates in the record, such as Kasher's base figure of $2,717.04 per year. Kasher (S) (Nov. 1, 1979) at 32. As calculated by the GPs, this increases LV's LPCs attributable to displacement allowances from $19,757 to $138,120. *1391 (4) Timing Considerations The GPs attack our discounting LPCs to January 1, 1974, as inconsistent with our conclusion (Opinion at 1311, see also pp. 1277, 1287-1289, 1352-1353) that the sale of EL-LV to a solvent railroad would have occurred on April 1, 1976. They propose that LPCs be discounted only to the latter date. The court attempted (Opinion at 66) to determine what rate of discount a purchaser making its decision on January 1, 1974, would have applied to a cash flow it would not have begun to receive until April 1, 1976; we did not make any calculations of what this would mean in terms of purchase price. We used the same discount rate in connection with LPCs but there we applied the discount to reduce the costs to January 1, 1974, dollars. Consistency between these two items and with our conclusions with respect to timing would seem to be achieved best by discounting both the earnings stream and the LPCs to April 1, 1976, the conveyance date. However, LV suggests (Answer, p. 7 n.3), and we agree, that this point is too important to decide without further briefing. This can be had at the conclusion of the nonrail use phase of the case if the parties have not yet reached a settlement. We therefore leave this point undecided. B. Ann Arbor The GPs contend that our reference to a figure of $1,190,000 on p. 1293 was incorrect. This was a typographical error, but one which had no effect on our result. We in fact used Kasher's undiscounted figure of $1,374,000, reduced this to $687,000 because of income taxes, and discounted this to $493,200. The GPs also make the same point concerning discounting AA's LPCs to January 1, 1974, which they have advanced in respect of LV. We make the same disposition of it. The petitions for reconsideration are granted in part, and denied in part, and decision is reserved in part, as indicated in this opinion. NOTES [1] The phrase was not meant to imply that the GPs had the burden of proof on such issues but rather to refer to issues on which the GPs had been working and could continue to work effectively without having the Ts' direct cases in hand. The parties clearly understood this. [2] This, and other procedural innovations later described, resulted from informal exchanges between the court and its executive attorney, Richard E. Eriksen, Esq., who has rendered invaluable service throughout these proceedings, with the Steering Committees for the Government Parties and the Transferors and Intervenors. After substantial agreement had been reached, the court would issue a show cause order why certain procedures should not be adopted. On the principal procedural innovations, to wit, the submission of direct evidence in writing and the conduct of cross-examination by deposition, there was no objection and an appropriate order was entered. [3] There were remarkably few of these. Most of those requiring judicial action were disposed of by Judge Thomsen, as a single judge, under Rule 4, subd. A of this court. [4] PC's affiliated Ts include 38 wholly owned subsidiaries and 7 with shareholders outside the PC group. [5] The settlement for Pennsylvania-Reading Seashore Lines was approved on January 26, 1981, Sp. Ct. Rep. N 31838; for Dayton Union on February 12, 1981, id. N 32113; and for Lehigh & Hudson River and New York & Long Branch on February 26, 1981, id. N 32208, N 32214. [6] The settlement for Philadelphia, Germantown & Norristown was approved on October 23, 1981, Sp. Ct. Rep. N 37877, for Delaware & Bound Brook on October 19, 1981, id., at N 37855. The settlement for North Penn had not been approved at the time of publication of this opinion. [7] The proffered Joint Terminals would have included certain properties of the CNJ as well. However, CNJ has studiously avoided any participation in the JT proffer. [8] The route from Chicago through Cleveland and Buffalo to New York was basically the former New York Central "water level" route. The routes from Chicago to Pittsburgh and from St. Louis to Pittsburgh and Harrisburg were basically the former Pennsylvania Railroad's main lines. [9] Although PC joined in the JT proffer, it was not part of PC's primary case indicating what would have become of its lines in the absence of the Rail Act. The value of some of the IFLs did, however, depend upon the continued viability of freight lines in the terminal areas. See Memorandum on Behalf of PC Concerning JTs Evidentiary Submission (Dec. 15, 1978). [*] For a discussion of EL's use of the term "net salvage value", see note 16, infra. While we do not now decide the point, we confess to great difficulty in envisioning a sale to another railroad at "net salvage value" higher than earnings value if earnings value includes all the elements noted in Part V, pp. 1219-1220, infra, of this opinion. We do not believe that EL would disagree. [*] These earnings value figures are derived from a study of LV's value to Conrail rather than its value to hypothesized acquirers. [*] LV contends that, on this hypothesis, its line north from Sayre, Pennsylvania, through Ithaca, New York, to Ludlowville, New York, would have been acquired by shippers with assistance from the State of New York and a public regional corporation. [*] With the exception of 18 miles of LV track in the Geneva-Auburn, New York area, which connected to PC's "Water Level Route", from Chicago to New York, LV contends that none of its routes are dependent upon the continued existence of any IFLs. LV has also treated these 18 miles of track as "specialized properties". [10] Just prior to scheduling rebuttal evidence, EL, R, and LV moved to exclude from consideration evidence submitted by the GPs indicating what value the "Combined Erie, Reading, and Lehigh Valley" (CERL) would have had to the Chessie on April 1, 1976, Sp. Ct. Rep. N 19316. The Ts included in CERL contended that this evidence was irrelevant to the acquisitions they had proffered, that the GPs had no need to go beyond their directly responsive evidence for the purpose of establishing their claim that the Ts' properties had negative earnings values, and that consideration of the CERL evidence would require much further discovery and cross-examination by them. The court determined that the CERL evidence would not aid it in determining rail use values of the properties of the CERL transferors, and thus ordered that the evidence would not be considered, in the interest of avoiding unnecessary delay. Order Limiting Rail Use Issues Between the Government Parties and the Erie Lackawanna, Reading and Lehigh Valley Trustees (March 26, 1980), Sp. Ct. Rep. N 19685. The breadth of this order was defined in the parties' Stipulation Regarding Scope of the Order Limiting Rail Use Issues (March 26, 1980), filed on October 24, 1980, Sp. Ct. Rep. N 22028, which specified the portions of testimony to be excluded. See 4 GP at G 1-11. [*] This is subject to the important caveat with respect to the possibility of our having to depart from the "market value" rule under the unique facts of this case expressed in the CMV Opinion, 445 F.Supp. at 1030-31, 1045. [*] At least some of the Ts, notably EL, use "net salvage value" in the sense in which that term has been used in ICC abandonment proceedings. However, that does not get us far. EL tells us: Prior to the Rail Act era, no specific ICC rules existed concerning the determination of net salvage value. ELRB at 335. EL also tells us, id., without contradiction, "in practical application it generally came to mean the gross proceeds realizable upon dismantlement of rail lines and the sale of the removed materials (together with real estate, rolling stock, etc.) less the cost of such dismantlement" but without the deduction of other costs (e.g., expenses of a selling organization if needed, and reduction to present value), which the Supreme Court recognized to be appropriate in determining a fair and equitable price for a purchased railroad without earning power in the New Haven Inclusion Cases, 399 U.S. 392, 436, 90 S.Ct. 2054, 2081, 26 L.Ed.2d 691 (1970), or of "glut", i.e., the temporary presence of an extraordinarily large supply, accompanied by a lessening of demand, due to the assumed nearly simultaneous dismemberment of large portions of the Ts' lines. A petitioner for abandonment is required to include its estimate of net salvage value in its abandonment application. 49 CFR 1121.44(c). It thus seems that the ICC does not in fact "determine" net salvage value and also that the only significance of its specification by the petitioner is that the ICC customarily conditions approval of abandonments on an agreement of the petitioner to sell for continued rail use at a figure at or above net salvage value. The sole explanation for the omission of commonly recognized deductions in the determination of scrap value which we have found in the record is in the testimony of Shannon on behalf of EL: These conventions probably relate to the recognition that a buyer has perceived a rail use value in the line proposed for abandonment. If all possible theoretical deductions were made, the buyer would have paid nothing for the rail use value, since it could itself still liquidate the properties (assuming authority to do so) and receive the same price that it paid. Shannon (S) (Dec. 31, 1978) at 76. We confess our inability to understand this explanation. Nothing prevents a purchaser for continued rail use from bargaining for a figure lower than net salvage value, and the petitioner for abandonment would still prefer to accept the offer if it were even nominally more than what the petitioner would receive as scrap value after all proper deductions. Perhaps the answer is that since in small transactions the difference between net salvage value and a more nearly accurate scrap value figure is not likely to be significant, use of net salvage value saves the parties from the expense of protracted investigation and negotiation, and the petitioner's announcement of net salvage value assures a purchaser for continued rail use that a bid in that amount will be accepted unless other acquirers for rail use bid more. [11] This statement should be read with the qualification that we have not made a detailed examination of the massive evidence introduced by PC. [12] The best illustrations are found in EL's case. EL concedes that if its acquirer were to be the Chessie rather than a Western line, EL's main line west of Sterling, Ohio, to Chicago would have been abandoned as duplicative of Chessie's own line and that if its acquirer were to be Norfolk & Western, EL's main line west of Huntington, Indiana, to Chicago would have been abandoned as duplicative of Norfolk & Western's own line. Again, as indicated in Part II, EL and LV urge that LV's line from Waverly, New York, to Oak Island Yard near Newark, New Jersey, was a superior routing to EL's line through Scranton, Pennsylvania, to Croxton Yard near Weehawken, New Jersey; EL concedes that if an acquirer chose the EL-LV combination, the Lackawanna line from Binghamton via Scranton would have been abandoned except that certain eastern portions would have been sold to a public body, presumably the State of New Jersey, primarily for passenger service. [13] We say this despite the statement in our Order upon Motion of the Government Parties for Temporary Suspension of Trial Proceedings with respect to the Claims of all Transferor Parties other than the Penn Central Corporation (December 5, 1980) at 5, Sp. Ct. Rep. at N 23424, wherein we reminded the Ts "that to the extent that any of their proposals for continued operation hinge on continued operation of lines of PC, they have the burden of showing that such lines would in fact have been continued in rail use through sales to private entities or to public bodies other than the United States rather than being scrapped, although not of showing that PC would have received the price for which it contended". When an adversary declines a specific contest, as the GPs have on the issue here under discussion, the burden on a proponent is relatively slight. See also the discussion with respect to consideration of the acquirer of an IFL as the purchaser of another's railroad, pp. 1367-1369, infra. [14] This position was reiterated by counsel for the GPs at oral argument. Tr. at 412. [15] This avoids the seeming anomaly of pricing non-scrapped property on the basis that it would have been scrapped. The GPs say there is no anomaly since an owner whose property was without significant earnings value is entitled only to its next best use, to wit, sale for scrap, regardless of what in fact happens to the property, and that a purchaser would pay no more. The New Haven was valued primarily on that basis but without objection, New Haven Inclusion Cases, supra, 399 U.S. at 436, 90 S.Ct. at 2081. [16] A prime example is the contention that public bodies would have paid more than net salvage value for losing passenger lines. [17] We choose the letter X rather than "net salvage value" because of the danger that the latter term could mean either salvage value or scrap value, assuming there is a difference between them. In fact we may find different X's depending on whether a particular property, although without significant earnings value, would or would not be continued in rail use. [18] While we ruled in the CMV Opinion, 445 F.Supp. at 1013-16, that in valuing the Ts' properties the United States could not be considered as a prospective taker, we made clear that this did not mean that we must assume that the Government would have simply sat by with its hands folded while the nation moved toward the precipice of economic collapse, see id. at 1008-09 (consideration of interim aid to avoid shutdowns because of cashlessness and regulatory changes permissible in alternative scenario). [19] Shannon, a witness for EL, AA, and PC, is President and Chairman of Wyer, Dick & Co., a transportation consulting firm which for 40 years has specialized in railroad valuation and economics. He has been in the railroad industry since 1936 and has an extensive background in all phases of operations and management. After leaving the New Haven Railroad as Executive Vice President in 1960, he joined Wyer Dick where he has primarily counselled that firm's clients on mergers, acquisitions, and valuation of rail properties. Over the past eighteen years he has been involved in several detailed studies of EL. Shannon (S) (Dec. 31, 1978) at 1-8. [20] The GPs assert that we should discount the value of the Ts' proffered sales for the risk that they might not have materialized. 2 GPRB at Y 20-29. Taking this argument literally would require that we determine values not only for the most likely outcome, but for all possible (nonspeculative) outcomes, weighting each by the probability of its occurrence in the alternative scenario — hardly a feasible approach given the multitude of possible permutations. We think this point can be deferred until we attempt to arrive at definitive figures. [21] We shall use this phrase to include all the properties here at issue although many would generally be regarded as lying in the Midwest. ICC statistics for the Northeast region include New England, the states of Delaware, Virginia, Maryland, Pennsylvania, New York, and New Jersey, and the District of Columbia. [22] Kornhauser is Director of the Transportation Program for Princeton University and is an Associate Professor of Civil Engineering at the University. He has published extensively in the field of transportation planning and economics. [23] The most obvious example is savings in overhead through the elimination or curtailment of the acquired line's executives, and legal, accounting, and traffic departments. Obviously elimination of such departments of the seller would somewhat increase the costs of the corresponding departments of the acquirer; only the net decrease constitutes a saving. The GPs insist that buyers would demand a "share" of savings in overhead and would appropriate the entire savings on their own lines. We think rather that a capitalization of all such savings, if substantiated, would enter into the buyer's reservation price. Another type of saving would occur when substitution of a superior line of the seller would enable the buyer to downgrade or even abandon lines of its own. An example of this is EL's claim that an acquisition by Chessie would have permitted the latter to rationalize its Chicago to Buffalo service by abandoning its expensive cross-Canada route and to rationalize its Wisconsin to Buffalo route by abandoning its car-ferry operations to Michigan. [24] The parties refer to the latter as "opportunity costs". The most dramatic case would consist of earnings on traffic interchanged by the buyer with the seller which originated or terminated at points on the seller's line not served by any other railroad. Professor Kornhauser distinguished such "lost traffic" from "jeopardized traffic", to wit, traffic that had been handled with the selling railroad and originated or terminated at points served by roads that were continuing in business. He thought this was also significant although incapable of accurate quantification. The jeopardy would arise from the circuity of the new route, inability because of physical limitations (e.g., single tracking) of the alternative carrier to handle the new traffic efficiently, or policies of the alternative carrier to favor other connections. [25] On this point EL argues that the remaining solvent east-west carriers, primarily Chessie and Norfolk & Western, not only would not have afforded service to many points in the east but that many of their lines were unduly circuitous or could not have absorbed the added volume of traffic previously carried by the bankrupts. [26] We shall follow the parties in calling these off-segment earnings. For example, if Santa Fe acquired EL it could divert to its own lines west of Chicago a certain amount of traffic originating or terminating at points on EL which EL had previously exchanged with other Western carriers at the Chicago gateway. As against this, other Western carriers acquiring other eastern lines such as PC's water-level route from Chicago to New York or its route from Chicago to Pittsburgh and thence to Harrisburg, and on to Philadelphia and New York, could have diverted to themselves certain traffic that these Eastern lines had previously exchanged at Chicago with Santa Fe. EL regarded the balancing of these pluses and minuses as so speculative that it did not take this sort of diverted traffic into account in its revenue estimates for a Western line acquisition. On the other hand, in its estimates for a Chessie acquisition it did consider the value of revenues on EL's Sterling-Chicago line, which would be abandoned, with the traffic passing to the B&O. We discuss off-segment earnings most extensively in connection with AA, see pp. 1369-1370, infra. [27] The chief example of this was Shannon's claim that creation of one or more transcontinental railroads would have created entirely new rail traffic. Shannon (S) (PC) (Dec. 1, 1978) at 39. [*] Shannon (S) (EL) (Dec. 31, 1978) following 29. This compares with Santa Fe's total pretax net income of $144,400,000. Smith (S) (Jan. 30, 1980) Ex. 27 at 8. [28] Kornhauser estimated that the opportunity cost, see note 30, supra, to Santa Fe would have been $19,090,000 in 1973 revenues, $7,254,000 in pre-tax profits, and $3,627,000 in after-tax profits if the PC IFLs had ceased operations. In addition, $78,874,000 in revenues would have been jeopardized, see note 30, supra. Kornhauser (S) (Dec. 1, 1978) at 50, 53. [29] Numerous witnesses testified to the effect that once a railroad has been shut down, there are serious obstacles to a new start even though the tracks are still in place. See note 263, infra. [30] This is indicated by our prophecy that it was the GPs who would doubtless protest the idea of a consortium purchase. Instead they now advocate it. [31] EL argues that § 5(11) immunity would not have been available since the transaction would have been consummated under § 1(18). Passing the point whether the transactions here at issue could have been carried out under § 1(18), which is discussed in Part IX, we doubt whether the Ts would have had sufficient bargaining power to refuse to proceed under § 5 if purchasing railroads had insisted. [32] The GPs have argued that Chessie lacked the resources to finance all the transactions in which it has been named as a potential purchaser. But no one suggested that Chessie would have made all these purchases. [33] In addition to GP witness Jacobs, Shannon, testifying as a PC witness, discussed the possibility of purchases by a consortium but in terms of a consortium bidding against individual lines and having to pay more. Shannon (S) (PC) (Dec. 1, 1978) at 41. Kornhauser also looked with favor on the idea of a consortium purchase as a means of avoiding uncertainties with respect to existing traffic flows but did not consider the effect of this on price. Kornhauser (S) (Dec. 1, 1978) at 55-58. [34] Murphy is Executive Vice President of the Consulting Center, Inc. Prior to joining that company, he was Vice President of Finance and Operations at USRA and an official with the United States Department of Transportation. Both Murphy and PC witness Richard Olsen, a senior associate at the consulting firm of Charles River Associates who has an extensive background in transportation economics, developed multiple year cost projections through 1985 which specifically took into account the effects of inflation upon IFL costs and revenues. Murphy concluded that these effects would be distinctly adverse; Olsen concluded that by a combination of rate increases and higher productivity the IFLs would more than meet increased costs due to inflation. We make no attempt to resolve this conflict since no witness attempted to replicate the IFL inflation analysis for the Ts still in this case. [35] In addition Murphy found a 9 month lag between the increase in costs and the implementation of responsive rate increases. This was due to a combination of petition lag (the time between incurrence of increased costs and industry submission of a general rate increase petition to the ICC), Commission lag (the time between submission of a rate increase petition and a Commission ruling thereon), and accounting lag (the time between granting by the ICC of a rate increase authorization and implementation of increased rates pursuant thereto by the Railroads). 2 GP at E 103-04 n.58. [36] Testimony to that effect had been given by Shannon, when testifying on behalf of PC before the ICC on the viability of a shrunken PC system. See Shannon (D) (June 21, 1979) Ex. 6. [37] Smith is a Managing Director of The First Boston Corporation, has been involved in financial matters related to the railroad industry for more than a decade, and has had considerable experience in both raising capital and evaluating mergers for his railroad clients. [38] EL also cited testimony of F.L. Lee Jones, a retired First Boston official, that he had made seven merger studies on the basis of constant dollars. 2 ELF at 412 n.352. There was no indication of the nature of Smith's or Jones' constant dollar studies. If these were for the purpose of establishing a fair exchange ratio for stock in mergers, there would be no sound objection to constant dollar studies since inflation presumably would affect both merger partners equally. A study of what a buyer would pay in cash stands differently. [39] EL contends that the same picture exists if the entire period 1966-75 is analyzed. It claims that the historical data showed actual cost increases of $7,473 million as compared with $10,614 million presented in the AAR data and used by Murphy, with cumulative rate increases of $8,229 million — again an excess of rate increases over costs. ELRB at 175. [40] Ex parte 256, 332 I.C.C. 280, 287-88 (1968); Ex parte 281, 341 I.C.C. 290, 301 (1972); Ex parte 305, 349 I.C.C. 862, 873 (1975); Ex parte 310, 349 I.C.C. 555, 562-65 (1975). EL also points out that the rate increase granted in Ex parte 310 was $999 million, 349 I.C.C. at 581, rather than the $889 reported by the AAR. [41] Ex parte 265 & Ex Parte 267, 339 I.C.C. 125, 177 (1971); Ex parte 281, 341 I.C.C. 290, 302 (1972). [42] EL also relied on Murphy's concession that the impact of inflation differs from railroad to railroad. However, we have been pointed to no reason why the Ts here would be less affected than the industry in general. [43] The GPs' witness, Smith, defines DCF as the amount "of cash which would have remained after all of the lines' financial and operating requirements have been met and which, therefore, would have been available either for investment in other projects or, alternatively, for distribution to shareholders as dividends." Smith (S) (Jan. 31, 1979) at 19. [44] To calculate PDV, one multiplies the future payment by a discount factor derived from the interest rate. For example, assume one wants to derive the PDV in year O of a DCF of $2 a year realized over a three year period valued at a 10 percent interest rate. The discount factor for the first year is equal to one divided by one plus the interest rate. To calculate the discount factor for subsequent years one compounds the interest rate. One then multiplies the DCF by the discount factor and sums the results. The figures for the above example would thus be the following: Year Discount Factor DCF ($) PDV ($) 1 1 = .91 × 2 = 1.82 _________ (1 + .10) 2 1 = .825 × 2 = 1.65 _________ (1 + .10)2 3 1 = .75 × 2 = 1.50 _________ ______ (1 + .10)3 Total $4.97 [45] To capitalize earnings one simply divides EBIT by the interest rate. For example, an enterprise expected to earn $2 a year in perpetuity valued at a 10 percent interest rate would have a capitalized earnings value of $20-$2/.10. [46] See note 44, supra. [47] Assume that there are two streams of income expected to be realized over a ten year period. The first stream is a constant one of $100 a year. The second yields $0 a year the first five years and $200 a year in the second five. While both average out to $100 per year, the PDV of the first is greater than that of the second. The higher the interest rate, the greater the difference. Assuming a 15 percent interest rate, the PDV of the two streams (in year 1) would be calculated as follows: $n $n = Payment in year n PDV = ___________ i = Interest rate (1 + i)n -1 n = The year Year Stream 1 Stream 2 1st 100/1.0000 = 100.00 0 2nd 100/1.1500 = 86.96 0 3rd 100/1.3225 = 75.61 0 4th 100/1.5209 = 65.75 0 5th 100/1.7490 = 57.18 0 6th 100/2.0114 = 49.72 200/2.0114 = 99.44 7th 100/2.3131 = 43.23 200/2.3131 = 86.46 8th 100/2.6600 = 37.59 200/2.6600 = 75.19 9th 100/3.0590 = 32.69 200/3.0590 = 65.38 10th 100/3.5179 = 28.43 200/3.5179 = 56.85 __________________ __________________ TOTAL 577.17 383.32 [48] This calculation represents a mix of methodologies. To calculate terminal value one divides the normalized year's earnings by the appropriate discount rate — exactly the same as the calculation of capitalized earnings value discussed in note 52, supra. Next, one calculates the PDV of the terminal value in the same manner as discussed in note 51, supra. [49] In its own study, the LV assumed that its properties would be valued on a tax-free basis. However, in the EL-LV combination, which the court considers to be the more likely outcome, see Part XII, pp. 1346-1347, infra, federal and state taxes are taken into account. [50] Involved in the railroad business since 1929, Laurie eventually became a partner in the transportation consulting firm of Wm. Wyer & Co., the predecessor of Wyer, Dick & Co. Since 1961, Laurie has been an independent transportation consultant. Laurie also submitted evidence on behalf of EL. [51] As a factual matter, it is unlikely that D,T&I would have viewed all future earnings from an AA acquisition as being tax-free. Smith testified that tax loss carryforwards must be used within seven years from the year in which the loss was realized. Smith (D) (Aug. 26, 1980) at 2176, 31 GP Supp.App. at 257. At some point, then, unless new losses were generated, D,T&I's existing loss carryforwards would have been either used or forfeited. And if D,T&I were to generate new loss carryforwards it obviously would not be in a strong position to acquire any properties. Indeed, AA does not envision any loss carryforwards being generated by D,T&I itself since D,T&I was a profitable railroad. Rather the loss carryforwards come from D,T&I's parent company, PC. But again, if D,T&I expected that at some time in the future PC, as part of its reorganization, would have sold D,T&I, D,T&I would not have expected to remain in a tax-free status forever. [52] See Smith (D) (Aug. 26, 1980) at 2169-71, 2176-78, 31 GP Supp.App. at 254-59. To illustrate, suppose a company is choosing between two investments both yielding $100 a year before tax in perpetuity. Assuming a 10 percent interest rate and a 50 percent marginal tax rate, the market value of both investments would be $500: (.5 × $100) / .10. To a tax exempt acquirer, both investments would be worth $1000. If for some reason the price of one investment rose to $600, the company would choose the other investment even though the first investment would still be worth more than $600 to the investor. [53] For a discussion of a proper treatment of depreciation in valuing an enterprise, see Blum & Katz, Depreciation and Enterprise Valuation, 32 U.Chi.L.Rev. 236 (1965). [54] Continuing the above example from note 59, supra, if company C had annual earnings after depreciation of $100, depreciation charges of $30 a year, and capital expenditures needed to maintain income of only $20 a year, the capitalized value of the enterprise would be $1000 if one capitalized EBIT and $1100 if one capitalized DCF. Typically, however, both because capital replacement costs usually exceed the capital acquisition costs upon which depreciation is based due to inflation and because the PDV of allowable depreciation charges is less than acquisition costs, capitalized expenditures needed to maintain income are usually greater than depreciation charges, see Blum & Katz, Depreciation and Enterprise Valuation, supra. [55] L&NE argues that capital expenditures needed to maintain income are not properly chargeable against earnings since they are more in the nature of retained earnings. See L&NE at 34-35. L&NE misunderstands the GPs' point. The allowance for capital expenditures is not properly deemed retained earnings since there is no net addition to the earning power of the company. These are expenditures which are needed to maintain the earnings power of the enterprise, not to augment it. [56] Smith also included "additions to working capital" in his calculation of capital expenditures needed to maintain income. This item can safely be ignored, however, since it never exceeds even I percent of cash flow for the IFLs. See Jennison (S) (Cost and Revenue Analyses for the IFLs) (Dec. 1, 1978) Schedule C. [57] Storm is a Vice President of Wyer, Dick & Co. and has been involved in rail transportation during his entire professional career. He was employed by the New York Central and the Port of New York Authority prior to his joining Wyer, Dick eleven years ago. [58] Equipment trust certificates are loans secured by a first mortgage against the rolling stock purchased with the proceeds from the loan. [59] From 1931 until his retirement as General Investment Manager in 1972, Young specialized in the analysis of prospective and oversight of previously made railroad investments for his employer, The Prudential Insurance Company, and was personally involved in most of the railroad reorganizations of the 1930's and early 1940's. See Young (S) (Dec. 31, 1978) at 1-2. [60] Id. at 11-12. EL posited a July 1, 1974, disposition date. See Part X, infra. [61] Actually, Laurie used a "pre-tax" interest rate of 15 percent. Assuming a 50 percent marginal tax rate, this is the same as using a 7.5 percent after-tax rate. [62] Laurie explained only that a 15 percent pre-tax interest rate, the equivalent of a 7.5 percent after-tax rate, is "considered appropriate". Laurie (S) (AA) (Dec. 31, 1978) Ex. AA-JL-1 at 8. [63] Roll is a Professor of Finance at the Graduate School of Management of UCLA. He has published extensively and has served as Associate Editor of several journals in this field. See Roll (S) (Jan. 30, 1980) Ex. 1. [64] Richard Meyer is a Professor of Business Administration at the Harvard Business School where he has taught since 1965. His list of publications is both long and impressive. See Meyer (S) (Jan. 30, 1980) at App. 1. [65] Hallingby is Vice Chairman of Merrill Lynch and is former Chairman of White, Weld & Co. [66] Stewart Myers, a cross-town rival of Richard Meyer, is a Professor of Finance at the Sloan School of Management, Massachusetts Institute of Technology. Like Meyer, he has written extensively in this field. [67] The theory behind AWCC is the same as that behind the segmented earnings approach to valuation. See Gardner, The SEC and Valuation Under Chapter X, 91 U.Pa.L.Rev. 440, 460-64 (1943). [68] Even if we accepted Stewart Myers' approach in theory, we would have no reliable way of implementing it. Because railroad bonds, as opposed to stocks, are not frequently traded, Myers had to estimate their market values. He did this in two different ways and derived upper and lower bound estimates of their true values. See Myers (S) (Dec. 1, 1978) App.C. The GPs showed, however, that Myers' upper and lower bound estimates were too high and that, since he relied on this range of figures, his estimated debt ratio was also too high. See 8 GP at App. R 231-32. [69] Roll also argued that an acquirer would have used a 30 percent debt ratio. He contended that an acquirer of the IFLs would have used less debt than that used by the average railroad because it would not have been able to take full advantage of the tax benefits resulting from the IFLs' rehabilitation. Roll (S) (Jan. 30, 1980) at 113-17. Whatever the merit of this proposition, see Myers (S) (Oct. 15, 1980) at 41, the GPs have presented no evidence showing that acquirers of EL's, LV's, AA's, or L&NE's lines would have been in a similar position. [70] Stocks, Bonds, Bills and Inflation: The Past (1926-76) and the Future (1977-2000) (1977). [71] Smith initially derived this rate for the IFLs. See Smith (S) (Jan. 30, 1980) at 65, 133-35. Based on his judgment, he assumed that the level of risk associated with the proffers of the individual Ts would be the same as for the IFLs. Id. at 252-53 (EL), 285-87 (LV), 292-94 (AA), 308-10 (L&NE). [72] The importance of the disposition date is essentially limited to the inflation factor since the real rate of return is derived from long-range historical data. To factor in inflation, one multiplies the appropriate inflation factor by one plus the interest rate and then subtracts one. The nominal rate of return for an average risk investment as of April 1, 1976 would therefore be 15.69 percent — 1.092 × 1.0594 (the inflation factor) minus one. [73] For example, the fact that an industry is regulated does not necessarily imply that it is risky — utilities, among the most heavily regulated industries, historically have been regarded as being among the least risky. [74] EL makes much of its claim that it was different from the other bankrupts. In support of this, it points to the fact that its entry into bankruptcy was precipitated by a natural disaster and that "as late as 1974, EL was thought to be independently reorganizable". 2 ELF at 425. Yet the fact of the matter is that EL, like the other Ts, was ultimately found not to be independently reorganizable. [75] The GPs also presented testimony from five witnesses — Evan Allen, Senior Financial Analyst for Southern; Paul Zeis, Director of Research for Norfolk & Western; Frank Barnett, former Chairman of Union Pacific; Thomas Lamphier, current President of Burlington Northern; and Alan Boyd, former President of Illinois Central Gulf and current President of Amtrak — all of whom stated that an investment in the Ts' lines would have been viewed as riskier than investment in their own operations. [76] We conclude, infra, see p. 1305, that the parties would have reached an agreement on the prices for the Ts' properties ten to twelve months after negotiations began in February 1973. For purposes of valuation, we will assume that potential acquirers would have valued the Ts' properties on January 1, 1974. [77] The GPs' witness, Roll, developed formulas which allow one to estimate the yield on railroad debt by factoring in its rating and the yield on U.S. Government bonds of comparable maturity. See Roll (S) (January 30, 1980) at 84-101. Assuming a 20 year term, the yield on A-rated senior mortgage debt as of January 1, 1974 would have been 9.155 percent and the yield on Aa-rated equipment trust certificates would have been 9.03 percent. Leaving aside the overall reliability of Roll's equations, we think that they yield unreliable results in this case. First, they fail to predict the actual yields which we have noted. Furthermore, the differential between the A and Aa-rated bonds is implausibly small. [78] In S&P's two reports, the proposed acquisitions are referred to at times by numbers known as proffer codes and at other times by Roman numerals. The table below identifies the proffer codes of the Ts whose RC and MOWE are discussed in this opinion: Proffer Code Proposed Acquisition 33 AA (Toledo-Ann Arbor segment) 35 R 82 EL I (Western Acquisition) 83 EL II (Chessie Acquisition) 86 EL III (Western Acquisition) 87 EL IV (Chessie Acquisition) 88 LV 89 LV III, IV (Historical) 42 LV III, IV (Projected) 1980 Study at VIII 5. The EL III and EL IV proffer codes cover the EL lines in the joint EL-LV acquisitions. The LV III and LV IV proffer codes cover the LV lines in the joint systems. The estimates designated by proffer code 89 are based on actual 1973 traffic on those LV lines, while those under proffer code 42 are based on the projected higher traffic after a combination with EL. [79] For example, EL estimated RC on its lines in the proposed Chessie acquisition to be about $27.76 million in 1973 dollars; S&P's figure was $148.299 million. See Statement of the EL Trustees in Response to the Court's May 21, 1981, Memorandum (June 1, 1981) at 11, Sp. Ct. Rep. at N 35810. [80] The 1979 Study developed RC including overhead for all lines of each T, expressed in 1974 dollars and premised on an assumed 1974 disposition date. It did not, however, cover either the scheduling of rehabilitation work or the development of annual MOWE. The 1980 Study developed RC and normalized MOWE without overhead for each of the Ts' proposed acquisitions, set out a schedule for the rehabilitation program, and expressed costs in 1973 dollars. In June 1980, Morgan made changes to portions of the 1980 Study's RC and MOWE estimates to account for newly discovered data and errors pointed out by the Ts during Morgan's deposition. See H. Morgan Corrections (S) (June 6, 1980). [81] Before his retirement in 1974, Morgan spent 38 years in various managerial and maintenance positions with the Penn Central Transportation Company, the Pennsylvania Railroad, and the United States Army. H. Morgan (S) (Jan. 31, 1979) at 1-3. Gunderson, a licensed professional engineer, spent 26 years of his career directly and solely involved with bridges. He spent the remainder of his 43-year career supervising others who were directly involved in maintaining the facilities of the Western Maryland Railroad. Upon his retirement as Assistant Vice-President of Engineering in 1972, he served as a railroad consultant. Gunderson (S) (Jan. 31, 1979) at 1-3. [82] A work function is a quantification of the labor, equipment, and materials needed to perform a particular rehabilitation task. For example, one of the work functions defined the amount of labor, equipment, and materials needed to replace one mile of track with new 119-pound rail in one eight-hour work day. Most of the work functions developed by S&P are independent of one another and describe all the elements necessary to complete a particular rehabilitation task. Some work functions cannot be performed alone, however, and are combined with others to form a series of continuous operations. These groupings of work functions are called "work items". The methods used to derive the various work functions and work items are described in detail in the 1979 Study at V. For a listing of all the work functions and work items, see 1 1980 Study App. IV. [83] See 14 GP Supp.App. at 701-28. Because the figures on the pages cited do not reflect credits for the salvage value of replaced rails, the sum of all rail, turnout, and crossings rehabilitation for any particular proffer code will be higher than the figure for total track rehabilitation listed for that proffer in H. Morgan Corrections (S) (June 6, 1980) at Table VIII-2aA. Even if all the salvage credit were applicable to rail replacement, rail replacement would still be the dominant item of track rehabilitation. [84] In most instances, sound business practice dictates removal and relaying before safety considerations require removal, for if the rail is left in its first position too long, its life for some second-position uses will be too short to merit the cost of installation. As a result, the decision to remove and replace rail generally depends on a railroad's needs for used rail. If mainline rail were destined for use in branch lines, it would be removed earlier than if destined for yard use, where even excessively worn rail has an extended useful life. 1 ELF at 160 & n.118; H. Morgan (D) (Aug. 11, 1980) Ex. 134, EL App. E at 212. In basing his rail replacement decision on the American Railway Engineering Association headwear standards, Morgan did not consider the needs of the Ts or their potential acquirers for used rail. [85] The AREA table is reproduced in the 1979 Study at III 30. [86] Rail is manufactured in a number of different weights, expressed in pounds per yard of rail; the heavier the rail, the greater the allowable headwear. [87] See H. Morgan (D) (May 7, 1980) at 3036-41, EL App. D at 317-22. The S&P Study itself states, "Since the most economical rail use is to relay it in the next lower class, rail should be removed when the headwear approaches the value indicated in Table III-14". 1979 Study at III 31. [88] The S&P Study states, "The Table, of course, does not represent the actual minimum limits which indicate when rail must be removed from a track. It is, instead, a guide for indicating the class of track in which a given rail may be relaid." 1979 Study at III 27 to 28; see also 3 GPRB at App. A 42. [89] At Morgan's deposition, D. A. Diestelkamp, one of two assistants assigned to making rail replacement decisions, testified as follows: Q. Mr. Diestelkamp, when you had a piece of 140-pound rail in Class 1 which had a quarter-inch wear, would that be a strong indicator for you of removal? A. That would be one of the factors, yes. Q. Would that be a strong indicator, a major factor? A. It would depend on the other criteria that is [sic] listed in the Bechtel data. Q. I understand that. I think Mr. Morgan had said that wear is one of the important factors. A. That is correct. Q. In your work didn't you regard this quantitative measurement as a major indicator of a need to replace? A. That is correct. Q. And when you found out the wear reached a quarter of an inch in Class 1, in most cases you replaced that rail, did you not? A. That's true. Q. If we were to go through all of these forms, we would find out that you replaced virtually all of the ail that exceeded that limit; isn't that correct? A. Within the confines of what you just stated, yes. H. Morgan (D) (Apr. 14, 1980) at 202, EL App. C at 24-25. [90] Class IV lines, which are yard lines, were considered separately. [91] Morgan did not have Bechtel Study data for all the lines in the EL Western acquisition. See 1979 Study at IV 10-11. [92] Tables showing the FRA track classifications and limits on allowable end batter and mismatch can be found in the 1979 Study at III 5, III 26. The standards used by Morgan, extrapolated from the FRA figures, are shown id. at III 41. [93] To the extent that Morgan overstated rail replacement, he also substantially overstated total track RC. It is possible, of course, that, even if he had chosen not to replace particular stretches of rail, those stretches would have required some other forms of rehabilitation. The GPs make essentially this sort of argument, although in a different context, with respect to AA. See 3 GPRB at App. A 83. Nonetheless, the unit costs for the various work functions and work items related to rail replacement are so much larger than those for other forms of tracks rehabilitation that the choice of one of the latter instead of one of the former would have reduced track rehabilitation costs considerably. Compare 2 1980 Study App. at IV-1-1 to IV-1-10, IV-2-1 to IV-2-2 with id. at IV-1-11 to IV-1-17, IV-2-2 to IV-2-3. [94] The GPs attack EL's attempt to rely on the Bechtel Study rail replacement estimates as proof of the excessiveness of Morgan's rail replacement estimates. The Bechtel Study estimates were prepared by the engineering firm of DeLeuw, Cather & Co. under the technical direction of Bechtel, Inc. The GPs argue that the DeLeuw team incorrectly used the sample data they had obtained from visual inspections of EL's lines. See 1 GPRB at E 35-40. As the GPs acknowledge, their argument is based on a statement by EL's counsel about how DeLeuw used the sample data. As far as we can tell, neither party introduced evidence concerning the complete methodology DeLeuw used in making rail replacement decisions. Nevertheless, at least one piece of information about DeLeuw's methods is known. It is generally agreed that DeLeuw inspected samples of track at two-mile intervals. EL counsel asserted that, unlike Morgan (who used the DeLeuw data in his study), DeLeuw did not treat the sampled portion as necessarily representative of conditions throughout the two-mile interval. Instead, they looked at other track records to determine whether conditions on the unsampled portion (e.g., the rail's installation date, the curvature of the track if any, etc.) were similar to conditions on the sampled portion. The GPs argue that this method understated rail replacement because if the sampled portion showed deterioration, DeLeuw replaced only that part of the unsampled portion where conditions were similar to conditions on the sampled portion; on the other hand, where the sampled portion was good, DeLeuw did not replace any rail even though it is possible that only a short distance beyond the sampled portion deterioration was such that replacement was needed. We see no basis for that assumption about DeLeuw's rail replacement methodology. If DeLeuw supplemented its visual inspections with additional track records, it is just as likely that they could have determined that, because conditions on the sampled and unsampled portions were dissimilar, rail replacement was needed in the unsampled portion even though it was not needed in the sampled portion. Although we cannot know for sure how DeLeuw made its rail replacement decisions, if their method did take into account additional data besides that obtained from visual inspections, then it is likely to be more reliable than Morgan's method of treating each sample as representative of an entire two-mile stretch of rail. The GPs also point out that, even if DeLeuw's rail replacement percentages are significantly lower than Morgan's, DeLeuw's total rehabilitation estimate for all non-bridge items is significantly closer to Morgan's figures for those items than it is to EL's figures. As the Chessie Study points out, however, see Chessie Study at 14-22, EL App. B at 217-18, DeLeuw probably overstated rehabilitation in a number of ways. Indeed, Chessie felt that DeLeuw's estimate should be discounted by at least 40 percent. Chessie Study at 26, EL App. B at 220. As a result, we do not find the similarity between DeLeuw's and Morgan's rehabilitation figures for non-bridge items very probative. [95] Total costs for turnout and highway crossings replacement (before salvage credit) for each T are as follows: AA $287,015 R $1,517,512 EL I $10,658,020 EL II $8,994,707 EL III $9,607,427 EL IV $7,771,870 LV $4,147,221 LV III, IV $3,167,955 See H. Morgan (D) (Aug. 22, 1980) Ex. 280, EL App. E at 278-81. The Roman numerals for EL and LV signify the various alternative configurations they proposed depending on which railroad or public entity acquired their lines. See note 85, supra. [96] The Bechtel Manual, a guidebook provided to the six subcontractors who did the field work for the Bechtel Study, originally specified sampling rates as high as 50 percent for some types of bridges; sampling rates for other types of bridges ranged below that figure to as low as 5 percent. See Gunderson (D) (May 15, 1980) Ex. 10 at 6-5, EL App. G at 185. Apparently, these rates were later cut in half. Compare id. with 1 1979 Study App. at III-2-1. [97] The Bechtel field forms for bridges contained three possible designations for the condition of each item specified on the form: A (good), B (fair), R (repair). The Bechtel Manual defined the B category as follows: "The second answer [fair] should be used when the inspector believes that the item could develop a problem in the future and should be watched, but no repairs are merited at the present time." Gunderson (D) (May 15, 1980) Ex. 10 at 6-7, EL App. G at 187. Gunderson's interpretation of the B rating appears to have been at odds with this definition. Under his interpretation, even bridges with some B ratings and no R ratings were scheduled for repairs. See 1980 Study at IV 28; Gunderson (D) (May 16, 1980) at 458-60 (by R.A. Wokurka), EL App. F at 56-58. [98] Although there was enough data on PC yards to allow Morgan to use a separate sample to determine PC's yard rehabilitation estimate, Morgan did not have enough data to use separate samples for the other Ts. See 3 GPRB at App. A 27-29. [99] The number of miles in mainline track determined S&P's figures for each T's yard track mileage and yard turnouts. To obtain figures for each T's yard track mileage and yard turnouts, S&P used Conrail's systemwide ratios for the number of yard track miles to mainline miles and for the number of yard turnouts to yard track miles. H. Morgan (D) (Apr. 23, 1980) at 1611-13, EL App. C at 240-42. Some modifications to this approach were made in the 1980 Study; they did not affect EL, LV, and AA. H. Morgan (D) (Apr. 23, 1980) at 1614-16, EL App. C at 245. [100] For EL, replacement costs (before salvage credit) for small yard turnouts on each of its four proffers ranged from $4,061,333 to $6,371,844. For LV, the figure was $3,745,611. See 1 ELF at 212. [101] Section 215 of the Rail Act, 45 U.S.C. § 725, provided funds for interim maintenance and improvements pending conveyance of the Ts' properties to Conrail. [102] See note 25, supra. [103] Shannon (S) (EL) (Dec. 31, 1978) at 52-54. Shannon defined his standard as follows: Generally, one can consider that a railroad in "50%" condition is in good operating condition. For example, if the anticipated tie life is 35 years, at the end of 17½ years the ties are in 50% condition but can still adequately perform the function for which they were installed. Id. at 52. This standard appears to be the generally accepted interpretation of normal condition. See 3 GP at K 39 n.17. [104] Shannon (D) (July 9, 1979) at 1004-06, EL App. B at 1-3. In its brief EL incorrectly states that Chessie's figure was stated in 1975 dollars. 1 ELF at 135. Shannon testified however that, although the Chessie Study used 1975 freight rates, it used projected 1976 cost levels. Shannon (S) (EL) (Dec. 31, 1978) at 82. EL quotes Shannon on this point earlier in its brief. See 1 ELF at 46. The evidence cited id. at 46 n.37 is sufficient to prove that Shannon was correct. [105] The pro forma statements prepared by Shannon for each acquisition contained two maintenance accounts, — catch-up maintenance, and "normalized maintenance". There was no separate account for routine yearly maintenance. That cost was reflected in the figure for net operating income and was based on actual 1973 expenditures. "Normalized maintenance" represented the amount of additional maintenance expenditures that should have been incurred in 1973 to avoid deterioration. Put another way, "normalized maintenance" plus the sum actually expended for maintenance in 1973 would, in Shannon's opinion, keep EL in optimum operating condition once catch-up maintenance had been performed. Catch-up maintenance was merely the cost to rehabilitate the lines to optimum operating condition. [106] To use an example employed in LV's examination of Murphy, if a railroad decides to lay 119-pound rail instead of 136-pound rail in its rehabilitation program, its annual MOWE budget should be adjusted to account for the fact that the lighter rail wears faster and will have to be replaced sooner than the heavier rail. B. Murphy (D) (June 20, 1980) at 138-39, 2 LV App. at 262-63. [107] In the Statement of the Erie Lackawanna Trustees in Response to the Court's May 21, 1981, Memorandum (June 1, 1981) at 9-12, Sp. Ct. Rep. at N 35808-11, EL points out three errors in the GPs' calculations: (1) The Chessie figures used by the GPs covered a package of lines that is 5.8 percent larger than the package EL's estimate covered; (2) Chessie's rehabilitation figure did not contain an adjustment for shortfall that EL's estimate contained; and (3) Chessie's figures were not accurately restated in 1973 dollars because the GPs incorrectly assumed that Chessie's figures were stated in 1975 rather than 1976 dollars. This error by the GPs was perhaps induced by EL's opening brief which incorrectly stated that the Chessie study expressed rehabilitation in 1975 dollars. See note 111, supra. EL also takes issue with the figure the GPs assigned as EL's estimate of annual MOWE. Annual MOWE is not stated separately in Shannon's study and cannot be computed without complicated calculations. The GPs therefore used a rough approximation. 1 GPRB at E 26 n.80. EL's own rough approximation is about $1.6 million higher. Multiplied by ten, the difference in the two estimates is substantial. Nevertheless, even accepting the GPs' figure for the EL estimate, but correcting for the three other errors discussed above, the combined rehabilitation and ten-year MOWE figure using EL's estimate of the combined total is 79 percent of Chessie's estimate, while the GP's estimate is 160 percent of Chessie's estimate. [108] The GPs came up with the following figures for the total costs of rehabilitation plus ten years of routine annual MOWE: $431,300,000 using S&P's estimates, $318,643,000 using Chessie's estimates, $203,878,000 using Shannon's estimates. As modified to correct the errors EL found in the GPs' calculations, the figures look like this: $431,300,000 using S&P's figures, $269,118,000 using Chessie's figures, and $229,080,000 using EL's figures. [109] For example, if most wage increases occur around March, Shannon's January 1, 1975, cost levels would reflect the March 1974 increase. The AAR 1974 indices would therefore be an accurate indicator of the actual cost levels Shannon used since those indices provide year-end 1974 figures and Shannon's January 1, 1975, figures would contain no wage increases not contained in the AAR index. The 1973 AAR indices would reflect the 1973 post-March labor costs and thus would accurately reflect the costs obtaining during the maintenance season (March through October). The proper comparison would therefore be between the 1974 and 1973 AAR indices, a one-year adjustment rather than the two-year adjustment Shannon made. [110] See note 108, supra. [111] The Adik study that valued all LV lines conveyed to Conrail used a computerized costing model to determine revenues, expenses, and capitalized earnings value. The study does not provide discrete figures for RC and MOWE; rather, those costs are buried in a number of different accounts. The parties have not attempted to extract from the study the pertinent figures for RC and MOWE for the LV lines in the combined systems. We have not been able to determine how to extract them ourselves. See generally Adik (S) (December 31, 1978) Ex. 1. [112] LV's traditional procedures and equipment were older and less efficient than the procedures and equipment employed by the USRA in the § 215 program undertaken in 1975 and Adik believed that similar work techniques and machinery would also have been used by a solvent acquirer. [113] Murphy also testified that the figures presented by the S&P Study were a more reliable indicator of LV's needs. B. Murphy (S) (Jan. 30, 1980) at 14. We hesitate to rely on this statement, however. Murphy endorsed S&P's original estimate of $62.35 million for the RC of the LV lines in the EL-LV proffers. When Morgan and Gunderson subsequently revised their estimate of RC for LV to $52.99 million Murphy endorsed the reasonableness of that figure without even examining the errors that they corrected. B Murphy (D) (Aug. 29, 1980) at 193-207, 2 LV App. at 280-86. Murphy explained at his deposition that the two S&P figures represented a reasonable range of estimates. But his written testimony made no such qualification. And his deposition also revealed that he was generally unfamiliar with the way in which Morgan and Gunderson conducted their studies. B. Murphy (D) (Aug. 29, 1980) at 216-18, 222-31, 2 LV Supp. App. 295-97, 301-09. In light of all this, we do not find Murphy's endorsement of the S&P Study to be especially probative. [114] S&P placed rehabilitation on LV's lines in the combined EL-LV systems at $44.65 million and annual MOWE at $6.871 million. H. Morgan Corrections (S) (June 6, 1980) at Tables VIII-2aA, VIII-2bA. Adik placed rehabilitation at $7.775 million and annual MOWE at $9.766 million. Adik (S) (EL) (Dec. 31, 1978) Schedules 11 and 14. [115] The following figures are approximations of the cost per mainline track mile for EL and LV: EL — $11,000; LV — $14,204. The EL figure was developed by taking the figure for total rehabilitation (with § 215 adjustment) on EL's lines in the EL-LV Western acquisition as presented in Memorandum of the EL Trustees in Response to the Court's Letter of June 19, 1981 (July 10, 1981) at Table A (following p. 11), Sp. Ct. Rep. at N 36164, determining the proportionate share (with § 215 adjustment) attributable to the cost of rehabilitating JT lines, and dividing by the total mileage figure for Proffer Code 86 in the 1980 Study at VIII 5. LV's figure is the $7.775 million estimate from Adik (S) (EL) (Dec. 31, 1978) Schedule 14, divided by LV's mileage figure in the 1980 Study at VIII 5. [116] Hale did not analyze how a reduction in speed would affect any track rehabilitation estimates other than rail replacement. [117] The entire exchange on this point was as follows: Q. If you look at the computer printout for line 6903, you are showing a net total rehabilitation figure of $487,000, and this represents rehabilitation to a maximum speed of 40 miles per hour; correct? A. Correct. Q. Based upon the fact that, under the assumptions made by the Ann Arbor in the rail use case, this line would only be operated at 25 miles an hour, could you tell me — could you tell the difference in rehabilitation costs that would be provided solely for an FRA Class II railroad? A. Not offhand. Q. Would it be more than 50 percent? A. You mean 50 percent less? Q. 50 percent less. A. No. H. Morgan (D) (Aug. 19, 1980) at 4574, 28 GP Supp.App. at 346. [118] Indeed, given Morgan's tendency to overstate RC, it is likely that Hale's use of stricter standards overcompensated for his short-term perspective. [119] The figures for mainline ties and turnouts were calculated from Hale's estimates for the segments of AA from Toledo to Diann and from Diann to Whitmore Lake, making an adjustment because the Diann-Whitmore Lake segment extends 9.5 miles beyond the Ann Arbor city limits, Hale (S) (Dec. 31, 1978) at 3. Mainline Ties Toledo-Diann: Units Unit Cost 400 $25 $10,000 Diann-Whitmore: Units Unit Cost 8500 × (36.5 - 9.5) miles ________ 36.5 miles = 6288 $20 $125,760 ________ Total $135,760 Turnouts Toledo-Diann: Units 0 $0 Diann-Whitmore: Units 10 × (36.5 - 9.5) miles _________ 36.5 miles = 8 $7,500 $60,000 _______ Total $60,000 See Laughlin (S) (Dec. 31, 1978) Ex. AA-MOL-1 at B 2-3. [120] See p. 86, supra. [121] The EL ratio, as calculated by EL from its Schedule 411 data, is 46 percent, while Conrail's systemwide ratio is 53.4 percent. 1 ELF at 284, 219. EL's ratio is about 14 percent lower than Conrail's. EL incorrectly quantified the difference as 15 percent. See 1 ELF at 284. A 14 percent reduction in yard maintenance on EL's Western acquisition amounts to $1.25 million. See H. Morgan Corrections (S) (June 6, 1980) at Table VIII-2bA. EL also contends that Morgan overstated mileage by treating some lines as double track lines instead of single track lines. 1 ELF at 284. EL did not quantify the amount of overstatement, however, so we presume that this error, like the 43-mile error relating to a line maintained by B&O, id., was not significant. [122] EL had no lines with densities above 20 MGT, see 3 GPRB at App. A 46-47, so according to the 1980 Study, none of its rail had an annual replacement rate over 2.5%, and some had an even lower rate. The 2.5 percent figure contemplates a 40-year rail life because at that rate all the rail in a given line will have been replaced at an average of once every forty years. [123] E.g., Susquehanna & N.Y.R.R. — Abandonment, 252 I.C.C. 81 (1942); Cisco & N.E. Ry.— Abandonment, 267 I.C.C. 139 (1947); Black Mountain Ry. — Abandonment, 290 I.C.C. 238 (1954); Manistee & Repton R. R. — Abandonment, 324 I.C.C. 489 (1964); Okmulgee Northern Ry. — Abandonment, 320 I.C.C. 637 (1964); Tennessee Central Ry. — Abandonment of Operations, 333 I.C.C. 443 (1968), modified, 334 I.C.C. 235 (1969); Bush Terminal R.R. — Entire Line Abandonment, 342 I.C.C. 34 (1971); Northampton & Bath R.R. — Abandonment, 354 I.C.C. 784 (1978). The ICC, in situations similar to those in the foregoing cases, has imposed labor protection conditions, e.g., Seaboard Air Line R. R. Trackage Rights, 312 I.C.C. 797 (1962); Washington & Old Dominion R. R. Abandonment — Virginia, 331 I.C.C. 587, complaint dismissed sub nom. Washington & Old Dominion Users Ass'n v. United States, 287 F.Supp. 528 (E.D.Va.1968). Although the distinctions drawn by the Commission are not altogether clear, it appears that protective conditions have been imposed only in situations where a parent corporation derives substantial benefits from the abandonment of its subsidiary's lines. We think the ICC would have found this exception inapplicable if, in the alternative scenario, a T abandoned its entire system for scrap value. [124] See, e.g., In re Chicago, Rock Island & Pacific R. R., No. 75 B 2697 (N.D.Ill. June 5, 1980) Partial Transcript of Proceedings at 4 ("that Congress believes that it can legislate a $75 million labor protection burden on the assets of the Rock Island comes to me as a startling concept") (McGarr, J.), application for stay denied sub nom. Railway Labor Executives' Ass'n v. Gibbons, 448 U.S. 1301, 101 S.Ct. 2668, 65 L.Ed.2d 1094 (1980) (Stevens, J.) and 448 U.S. 909, 100 S.Ct. 3057, 65 L.Ed.2d 1140 (1980) decision on probable jurisdiction postponed and case consolidated with No. 80-1239, 451 U.S. 936, 101 S.Ct. 2014, 68 L.Ed.2d 322 (1981); and In re Chicago, Rock Island, & Pacific R. R., No. 75 B 2697 (N.D.Ill. Oct. 15 & 16, 1980), aff'd sub nom. Railway Labor Executives' Ass'n v. Gibbons, 645 F.2d 74 (7th Cir. 1980) (by an equally divided en banc court), probable jurisdiction noted and case consolidated with No. 80-415, 451 U.S. 936, 101 S.Ct. 2014, 68 L.Ed.2d 322 (1981). [125] The GPs argue briefly that the Penn Central Merger Protection Agreement (PCMPA) would have been applicable to LV and AA as direct or indirect subsidiaries of PC, although they do not attempt to calculate the costs of the benefits of the agreement. 5 GP at S 46. LV rightly points out that the PCMPA required PC, and not LV, to bear responsibility for labor protection payments. See Pennsylvania R. R. — Merger — New York Central R. R., 347 I.C.C. 536, 551 (1974). Moreover, our conclusions as to the uncertainty that would have surrounded the applicability of the Erie Lackawanna Merger Protection Agreement, see pp. 1263-1264, infra, also apply to the PCMPA. Accordingly, we accept LV's argument that the PCMPA would have had no effect on its labor protection costs. The same considerations apply to AA. The GPs apparently also contend, 5 GP at S 47, that the February 7, 1965 Job Stabilization Agreement would have applied to EL and LV in the circumstances of the proffered transactions, although the GPs again do not consider this agreement in their calculation of the Ts' labor protection costs. The agreement applied only to nonoperating employees employed by a T on October 1, 1962. Because employees generally would have been terminated in reverse order of seniority and most 1962 employees would have achieved relatively high seniority by 1973-74, the agreement would have had little practical effect on the Ts' labor protection costs. In any event, our conclusions regarding the uncertainty that would have surrounded the question whether the Erie Lackawanna Merger Protection Agreement was applicable to the proffered transactions, see pp. 1263-1264, infra, are equally applicable to the February 7, 1965 Job Stabilization Agreement. The GPs also argue briefly, 5 GP S at 253-56, that other agreements with third parties — such as equipment financing obligations, fuel contracts, and the like — would have reduced the price of the Ts' properties. As regards the Ts remaining in the case, virtually no evidence has been presented in support of this theory. Moreover, the GPs cite us to no transaction comparable to those proffered in which such agreements have affected the purchase price. Accordingly we reject the GPs' argument. [126] The GPs' argument, 2 GPRB S at 112 n.277, citing Minutes of May 13, 1936 at 5; 22 GP Supp.App. at 289 (statement of Mr. Harrison), that the WJPA negotiation minutes "reveal an intent to cover all attempts to unify `by a device'" is unpersuasive. The portions of the negotiations cited by the GPs reveal only an attempt to prevent circumventions of the WJPA, and not an explication of the scope of that agreement. [127] We are unpersuaded by the conclusory statements of the GPs' witness, Dr. Jacob Seidenberg, that a neutral arbitrator would have found the WJPA applicable in the circumstances of the alternative scenario. Seidenberg (S) (Jan. 31, 1979) at 36-37. As the GPs concede, Seidenberg did not consider the particular circumstances of the Ts' cases nor the arguments that the Ts might have advanced in arbitration. 5 GP at S 81; Seidenberg (D) (Apr. 10, 1980) at 122, EL App. M at 203. [128] EL's inclusion in the Norfolk & Western system was accomplished by Norfolk & Western's purchase of EL's stock, which Norfolk & Western regarded as a complete loss following EL's entering reorganization. See also Matter of Erie Lackawanna Ry., 404 F.Supp. 954 (Sp. Ct.R.R.R.A. 1975). [129] The most generous type of labor protection is "attrition protection", which guarantees employment or its equivalent in wages and fringe benefits until an employee dies, retires, resigns, becomes disabled, or refuses to accept a bona fide job opening. The ELMPA, the PCMPA, and other agreements provided attrition protection for covered employees. [130] See cases cited in note 130, supra. [131] We do not think that the refusal of labor unions to reach agreements with Chessie and Southern regarding acquisition of the EL system and PC's Delmarva lines indicates that labor would have blocked the proffered transactions. In the event that the Chessie and Southern acquisitions failed, the EL and PC lines were to be acquired by Conrail, and the employees on these lines would receive the extremely favorable labor protection benefits provided by the Rail Act. In the alternative scenario, labor would have had no such protection and would have faced the possibility of widespread shutdowns in the Northeast. [132] Amtrak C-1 benefits, negotiated pursuant to the Rail Service Passenger Act, 45 U.S.C. § 565, and incorporated in the 1971 Passenger Service Agreement between Amtrak and Penn Central, provide dismissed employees with a choice of a lump sum separation payment or a monthly dismissal allowance equal to 100 percent of their base period salary for up to six years following termination. Ex. GPS — Labor 27. We do not think this agreement is particularly instructive in our case. The Rail Service Passenger Act provided labor with a guaranteed level of benefits, 45 U.S.C. § 565(b), and the availability of federal intervention made it highly unlikely that the Ts would actually cease operations. [133] LV contended that it would have avoided any labor protection obligations, as did AA. Shannon took the view that labor would have "panicked". Shannon (D) (Jan. 15, 1981) at 652, 30 GP Supp.App. at 179. [134] The Rock Island and Milwaukee liquidations gave rise to agreements between labor and the potential purchasers of the lines involved regarding labor protection. These agreements — the "Miami accords" — provided 80 percent of base period compensation for up to three years for employees hired on the Rock Island and Milwaukee systems. As the GPs recognize, 2 GPRB at S 85 n.216, these terms were less than those labor historically had obtained. The GPs point out, however, that workers who failed to find employment on the Rock Island and Milwaukee lines were protected by special federal legislation. 2 GPRB at S 80 & n.201. The restructuring of the Rock Island and Milwaukee nonetheless involved labor "concessions ... in the amount and duration of salary guarantees", 2 GPRB at S 85 n.216, and, although coming several years after the transactions hypothesized in this case, belies the GPs' claims of an uninterrupted trend towards increased labor protection. See also R.D. Timpany, Trustee of the Central R. R. Co. of New Jersey — Abandonment of Operations between Phillipsburg, N.J. and Hudson, Pa., 342 I.C.C. 227, 301 (1972) (minimal labor protection benefits imposed in partial line abandonment). [135] If labor protection costs exceeded the excess value of rail use sales over scrap values, then it would be more profitable for a T to choose to scrap, thus avoiding labor protection costs. See p. 1260, supra. [136] Considerations that would have influenced the outcome of negotiations between the Ts and labor would have included the bargaining skill of the various parties, the parties' perceptions of the willingness of other parties to risk blocking rail use sales, and the parties' perceptions of the value of rail use sales. We do not attempt to quantify such intangibles. [137] The ICC devised the New Orleans conditions in response to the Supreme Court's direction in Railway Labor Executives' Ass'n v. United States, 339 U.S. 142, 70 S.Ct. 530, 94 L.Ed. 721 (1950), that it provide a "fair and equitable arrangement" pursuant to § 5(2)(f). See New Orleans Union Passenger Terminal Case, 282 I.C.C. 271, 281-82 (1952). [138] See examples cited in Ables, supra, at 136-37. [139] Kasher, a lawyer specializing in labor law, has held numerous positions in the railway labor field. He served as General Counsel on the National Mediation Board, as director of Conrail's Title V Administration, and has been employed as a consultant and labor arbitrator. Kasher (S) (Nov. 1, 1979) at 1-3. He had not, however, had any real experience in the computation of labor protection costs prior to this case. Kasher (D) (May 5, 1980) at 42, I LV App. I at 60. [140] A "nonagreement employee" is an employee unprotected by existing agreements. [141] Kasher estimated that 3,575 PC, 348 EL, and 139 LV nonagreement employees would have been terminated as a result of the proffered transactions. [142] If we were to conclude that more positions would have been lost following acquisition than the T's Pro Formas reflect, we would then be required to reduce the T's projected labor expenses. Because LPCs for an employee are less than the cost of retaining the employee in employment this would only benefit the Ts. [143] The PC "refined case" was devised by PC witness Edward Jennison, Jennison (S) (PC) (Dec. 1, 1979) at 29-31, simply to test the profitability of PC's IFLs assuming relatively extensive abandonments by the other Ts. No evidence was submitted that the "refined case" would have been the likely disposition of the Ts' lines in the absence of the Rail Act. [144] This result follows from our conclusion, p. 1351, infra, that there would have been several Western lines bidding for the EL-LV Western package and Chessie bidding for the EL-LV Chessie package. If a Western line bid would have been higher than the Chessie's bid then we obviously do not have to estimate LPCs for the Chessie. On the other hand, if the Chessie, taking into account the X of the Chicago-Sterling line, would have been willing to pay more than any Western acquirer we again need only consider the LPCs to a Western acquirer because Chessie would have only have had to pay a nominal amount more than that the Western acquirer would have bid. [145] See, e.g., Kasher (D) (Aug. 13, 1980) at 1884 (by Rybicki), EL App. M at 13; Kasher (D) (Aug. 13, 1980) at 1837 (by Willey), EL App. L at 304; Kasher (D) (May 6, 1980) at 262 (by Willey), EL App. L at 38; Kasher (D) (May 6, 1980) at 266 (by Willey), EL App. L at 39; Kasher (D) (Aug. 13, 1980) at 1860-63 (by Willey), EL App. L at 322-25; Kasher (D) (Aug. 13, 1980) at 1868-74 (by Willey), EL App. M at 1-7; Kasher (D) (Aug. 13, 1980) at 1892-98 (by Rybicki), EL App. M at 17-22; Kasher (D) (May 6, 1980) at 83 (by McDonough), EL App. L at 3. [146] See, e.g., Kasher (D) (Aug. 13, 1980) at 1876-78 (by Willey), EL App. M at 9-11; Kasher (D) (Aug. 13, 1980) at 1839-49, 1860-63 (by Willey), EL App. L at 306-16, 322-25. [147] See Kasher (D) (Aug. 13, 1980) at 1850 (by Willey), EL App. L at 317; Kasher (D) (Aug. 13, 1980) at 1889 (by Rybicki), EL App. M at 16; Kasher (D) (May 6, 1980) at 266 (by Willey & Rybicki), EL App. L at 39. See also Kasher (D) (May 6, 1980) at 256-58 (by Willey), EL App. L at 35-39; Kasher (D) (Aug. 13, 1980) at 1853-54, EL App. L at 318-19; Kasher (D) (May 6, 1980) at 226, EL App. L at 31. [148] A number of Shannon's assignments with Wyer, Dick & Co. involved labor and labor protection matters. Shannon (D) (June 21, 1979) Ex. 2, EL App. A at 1-6. See also note 25, supra. [149] The 1,819 figure was derived from Exhibit 16 based on the following methodology: addition of sheet 1, column 1, line 7 (14 positions); sheet 2, column 1, line 8 (166 positions); sheet 3, column 1, line 11 (39 positions); sheet 4, column 1, lines 7 & 28 (0 positions); sheet 5, column 1, line 6 (2 positions); and sheet 6, column 1, line 11 (97 positions), gives a total of 318 positions. Application of this method to the remaining years yields a total of 1,819. This figure includes the positions that Shannon concluded would have been eliminated from EL's rehabilitation and capital improvement programs. Although Shannon's workpapers net jobs abolished against those created, we have not done so, to reflect difficulties in relocating surplussed employees. For example, for year 2, addition of sheet 1, column 2, line 7 (41 positions) sheet 2, column 2, lines 8, 13, 14 & 16 (213 positions); sheet 3, column 2, line 11 (73 positions); sheet 4, column 2, lines 6 & 27 (5 positions); sheet 5, column 2, line 6 (15 positions); sheet 6, column 2, line 11 (196 positions) gives a total of 543 positions. [150] For purposes of ICC reporting, a carrier's workforce is divided into groups by type of employment: executive (Group I); professional and clerical (Group II); MOW (Group III); Maintenance of Equipment (Group IV); and Transportation (Groups V, VIa, and VIb). [151] See Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 1-6, EL App. N at 517-23; Kasher (S) (Nov. 1, 1979) at 26; Grotz (S) (Oct. 15, 1980) at 24-25; Pennsylvania R. R. — Merger — New York Central R. R., 327 I.C.C. 475, 689 (1966). [152] See note 156, supra. [153] Application of a 10.4 percent attrition rate to the total number of EL operating employees, as reflected in sheets 1-6 of Exhibit 16 to Shannon's deposition, produces the figures in Table 2. For example, addition of the figures (to the left of column 1) at line 10, sheet 1; line 23, sheet 2; line 14, sheet 3; lines 11 & 32, sheet 4; line 9, sheet 5; and line 20, sheet 6, produces 9232 total EL operating employees. Thus, 9232 × 10.4% = 961, as appears in Table 2. [154] The figures in Table 3 are derived from Exhibit 16, at sheets 1 and 6, EL App. N 517 & 23. For example, examination of column 1 in sheets 1-6 indicates that 616 new jobs would have been created in year 1 in ICC Group III, Shannon (D) (Feb. 11, 1981) Ex. 16 at sheet 2, EL App. N at 518. [155] We do not attempt to reflect the attrition of employees from the surplus pool due to death. This decision reflects both the fact that this factor would have had only a small impact and the fact that an acquirer would have made a conservative estimate of its LPC liability. [156] Although a considerably higher proportion of Conrail employees chose monthly dismissal payments under the Rail Act, this experience is of little relevance to this case. While the Rail Act's provisions regarding monthly dismissal allowances, Rail Act § 505, are more favorable than those of the New Orleans benefits, there is little difference between the lump sum provisions of the two types of protections. Moreover, we believe EL would have both announced and demonstrated its ability to police the outside earnings deduction, thus making monthly dismissal allowances even less attractive. Our calculations are based on Shannon (D) (Feb. 11, 1981) Ex. 16 at sheets 28-35, EL App. N at 544-50, and the figures for average yearly compensation contained therein. [157] Grotz's figures indicate that even the less senior LV workers tended to have worked for the railroad for longer than five years. Thus, even those employees opting to receive dismissal allowances would have fallen primarily into the 5 + year category. Grotz (S) (Oct. 15, 1980) at Table 4, 5. See also I FSP at 164 (employees with less than five years service constitute less than 15 percent of railroad workforces). [158] The figures in Table 21 are derived from Shannon's workpapers, XXXXXX-XX, EL App. N at 628-39, and testimony, Shannon (S) (Oct. 15, 1980) at 21-22. For example, with respect to MOWE, line 15 of 999020, EL App. N at 628, indicates a reduction in labor expenses of $240,525, which, when divided by the average MOW yearly salary, lines 7-8, reflects a reduction in employment positions in year 1 of 19 employees. Figures for Traffic and General departments derived from Shannon's testimony, id. [159] According to our calculations the ratio would have been 45 percent in year 1; 45 percent in year 2; 10 percent in year 3; and 0 in year 4. [160] Figures for total size of EL's nonoperating workforce derived from Shannon workpapers XXXXXX-XX, EL App. N at 628-39. For example, 127 MOW employees (999020, line 8), 93 maintenance of equipment employees (999023, line 7), 153 transportation employees (999027, line 6), 297 traffic employees (Shannon (S) (Oct. 15, 1980) at 21 (164 nonunion + 133 union = 297)), and 650 general employees (Shannon (S) (Oct. 15, 1980) at 21-22) provides a total of 1320, less 126 unprotected employees, leaves 1194. [161] Base period salaries derived from Shannon's workpapers 999020, line 30; 999023, line 29; and 999027, line 29, EL App. N at 628, 631, 635. Average protected employee's 1973 salary, based on these figures and adjusted upwards to reflect higher salaries of General and Traffic employees, see "Erie Lackawanna Railway Company, Black Book", Shannon (D) (Feb. 12, 1981) Ex. 26, EL App. N at 583, estimated to be $11,750. Adjustment for real increases in railroad wages, see Table 8, provides a 1975 salary of $12,320; a 1976 salary of $12,620; a 1977 salary of $12,920; and a 1978 salary of $13,230. Adjustments of these figures to reflect appropriate base period salaries, see Table 9, produces the following figures: 4/75-4/76, $12,419; 4/76-4/77, $12,719; 4/77-4/78, $13,022. No downward adjustment of salaries, as was used in Table 9, is applied in this context because the dismissal of a large number of nonoperating employees would have reduced the impact of seniority on the average salary of dismissed nonoperating employees. [162] Figures in Table 31 derived from Tables 24 (years 1-6) and 28 (year 7). [*] Grotz has held positions as Vice President of the Chase National Bank's Transportation Department and as President and Chief Executive Officer of Western Maryland Railway Company. Since 1969 he has been employed as a railroad consultant. Despite his extensive experience in the railroad business generally, he had never conducted a LPC study prior to submitting his statement in this case. [*] Our LPC calculations, which are based on Grotz's study, indicate that LV's LPCs pursuant to New Orleans benefits would be slightly larger than Kasher's projections for comparable benefit levels. This is attributable to the different assumptions adopted by Kasher and Grotz regarding the numbers of employees who would have been dismissed because of the proffered transaction. Kasher's other projections of LV's total LPCs were substantially higher than Grotz's. Grotz excluded employees aged 64 or older and nonagreement employees from his calculations. Grotz (D) (Jan. 24, 1981) at 462, 25 GP Supp.App. at 80. We accept his conclusion for several reasons. As in the case of EL, we think the number of nonagreement employees provides a fair estimate of the number of unprotected employees. Although LV would likely have incurred some costs as a result of requiring or encouraging, 2 GPRB at S 61, the retirement of employees aged 64 or older, we are aware of no evidence of the extent of these costs. In any event, we accept Grotz's treatment of this issue in an effort to reflect the reduction in LPCs that a 1976 acquisition date would have produced, see note 175, infra. [163] Alternatively, we think an acquirer would have applied a lower attrition rate — with much the same effect on LV's LPCs. For example, the average attrition rate for B&O in 1973 was 6.74 percent, Shannon (D) (Feb. 12, 1981) Ex. 23, EL App. N at 554, for C&O in 1973 it was 7.36 percent, Shannon (D) (Feb. 12, 1981) Ex. 24, EL App. N at 567, and for Norfolk & Western in 1973 it was 6.89 percent, Shannon (D) (Feb. 12, 1981) Ex. 25, EL App. N at 576. Either of these adjustments would have reflected an acquirer's reluctance to underestimate the difficulties and uncertainties, see pp. 1274-1275, supra, inherent in using attrition and rehiring to reduce LPCs. [164] The figures in Table 34 are derived as follows: 389 is taken from Grotz (S) (Oct. 15, 1980) at Table 2. The figure for year 2, 112, is obtained from Grotz (S) (Oct. 15, 1980) at Table 3, "Summary", lines 1-2: subtracting 2,015 from 2,161 leaves 146 and subtracting 34 unprotected employees, see id. at Table 3 n.6, assumed to come from earliest group possible, leaves 112. The same methodology has been applied to obtain the remaining figures in Table 34. [*] Distribution of employees choosing lump sum payments by years of service was obtained by calculating the percentage of employees in each of the WJPA categories, e.g., 4/195 = .0205, see Grotz (S) (Oct. 15, 1980) at Table 8, and .0205 × 112 = 2. Base period salaries used in Tables 39 & 41 derived from Grotz (S) (Oct. 15, 1980) at Table 6, and adjusted to reflect real increase in railroad wages and fact that recipients would earn somewhat less than average wage. E.g., for 1975 salary, $11,800 × 1.0242 = $12,373; $12,373 × 95% = $11,754. This figure is then adjusted to reflect 4/75-4/76 base period: $11,847. [*] The LPC studies prepared by the parties did not attempt to quantify the effect of a later acquisition date on LPCs. Moreover, the record is devoid of evidence that would permit us to attempt such calculations, although we seek to reflect the effect of this factor in the bargaining between the acquirers and the Ts and in our estimate of the number of LV employees adversely affected by acquisition, see note 171, supra. [*] In 1968 PC estimated that its LPCs under the PCMPA would total approximately $78 million in the eight years following its merger with the New York Central Railroad. Pennsylvania R.R. — Merger — New York Central R. R., supra, 327 I.C.C. at 686. After only five years, however, PC had incurred LPCs of $116.3 million. Ex. GPS — Labor — 4. [*] As noted earlier, p. 1260, supra, the ICC generally has refused to impose labor protection conditions in cases involving complete abandonments. See cases cited in note 130, supra. Decisions to the contrary have occurred only infrequently and in unusual circumstances, such as a sale of abandoned lines for considerably more than scrap value, e.g., Washington & Old Dominion R. R., supra, 331 I.C.C. at 600-02. [**] Our timing analysis must include abandonments, as well as sales, even though the former may be a nonrail use issue according to the letter of the Second Pre-Trial Order. Similarly, we decide the timing of the PC and R dispositions, despite the penultimate paragraph of the Sixth Pre-Trial Order, since the dispositions of all Ts are intertwined. None of the parties challenges this. [165] Under a "two-step" reorganization plan, each bankrupt would first dispose of its properties, with ICC approval, and then submit a plan for distribution of the proceeds, rather than submitting a unified plan in the first instance. See New Haven Inclusion Cases, 399 U.S. 392, 413 n.47, 90 S.Ct. 2054, 2069 n.47, 26 L.Ed.2d 691 (1970). See also Blanchette (D) (June 19, 1979) 289-92, 9 GP Supp.App. at 21-23. [166] R appears to suggest that its disposition might have taken place under § 77(o), rather than §§ 77(d) and (e). Section 77(o) authorizes "interim" sales and abandonments, while §§ 77(d) and (e) authorize sales and abandonments as part of a "final scheme of reorganization". For present purposes, the distinction between the two is not relevant, since the ICC would have had to scrutinize the sales and abandonments of rail use properties under either. [167] Sections 77(d) and (e) require that a reorganization plan be approved by the reorganization court after it is approved by the ICC. Blanchette testified that court approval of the decision to file with the ICC is also required. Blanchette (S) (Dec. 1, 1979) at 14. See also 2 ELF at 470 n.18. EL submitted, however, that court approval of the filing would be "virtually automatic". Id. Relying on this unrebutted assertion, we conclude that such approval, if necessary, would not have delayed the disposition process. [*] The ICC would, in fact, have evaluated the proposals under §§ 77(d) and (e) of the Bankruptcy Act, but these sections incorporated the standards of § 5 of the Interstate Commerce Act. New Haven Inclusion Cases, 399 U.S. 392, 425, 90 S.Ct. 2054, 2075, 26 L.Ed.2d 691 (1970). [168] Evan Allen, Director of Rail Asset Valuation for USRA since 1978, had previously served as Senior Financial Analyst for the Southern Railway System and as Master Budget Accountant for Lockheed Aircraft Corp. Allen (S) (Nov. 1, 1979) at 1-2. Allen testified that it would have taken potential purchasers of IFLs two years or more to prepare the financial projections necessary to negotiate a memorandum of understanding. Id. at 4. William Quinn, a Director of the National Railroad Passenger Corporation (Amtrak), had been Chairman of the Board of the Chicago, Milwaukee & St. Paul Railroad, as well as President of the Chicago, Burlington & Quincy Railroad. Quinn (S) (Nov. 1, 1979) at 1-2. He testified that studies sufficient to support even tentative offers could not have been completed before February 1975. Id. at 4. Maynard Parks, a labor expert for the GPs, had worked in labor-related capacities in the railroad industry for over 40 years. He had served as Chairman of the Western Carriers Conference Committee and as Vice-Chairman of the National Railway Labor Council, and had been involved in negotiations with unions concerning several major rail acquisitions. Parks (S) (Nov. 1, 1979) at 1-2. Parks testified that an in-depth labor study relating to one of the IFLs would have taken at least a year. If more than one IFL were studied, more time would have been required. Id. at 17. [*] See, e.g., Myles (S) (Nov. 1, 1979) at 14-26; Luna (S) (Nov. 1, 1979). Charles Luna is a Director of Amtrak and has been President of both the Brotherhood of Railroad Trainmen and the United Transportation Unions. Id. at 1. Alfred Myles is a former Assistant Vice-President of Labor Relations for Chicago & Northwestern. Myles (S) (Nov. 1, 1979) at 1. [*] E.g., Stangl (S) (Nov. 1, 1979); Altshuler (S) (Nov. 1, 1979); Schuler (S) (Nov. 1, 1979). Raymond Schuler is a former Commissioner of Transportation for the State of New York. Id. at 1. Alan Altshuler is Chairman of the Political Science Department and Professor of Urban Studies and Planning at MIT. Altshuler (S) (Jan. 31, 1979) at 1. Peter Stangl has been an Assistant Professor of Economics at Wheaton College, Assistant Transportation Administrator for New York City, and Assistant Commissioner for Public Transportation in the New Jersey Department of Transportation. Stangl (S) (Nov. 1, 1979) at 1-2. [**] E.g., Brooks (S) (Jan. 31, 1979); Corber (S) (Jan. 31, 1979). Robert Corber served as a member of the ICC from March 7, 1975, to December 1, 1976, and had twenty-four years of private legal practice in the transportation field. Id. at 1-3. Robert Brooks was employed by the ICC for twenty-one years, and served as Director of the Commission's Office of Proceedings from 1974 to 1978. Brooks (S) (Jan. 31, 1979) at 1-4. [*] See Blanchette (S) (Dec. 1, 1978) at 10-11, 40-42; Shannon (S) (EL) (Dec. 31, 1978) at 93-94; Israel (S) (Dec. 31, 1978) at 21-22; Hesse (S) (R's Probable Actions) (Dec. 31, 1978) at 21-22. Robert Blanchette, who testified for PC, was, first, a trustee, then Chairman of the Board of Trustees, and finally Chief Executive Officer of PC during the mid-1970's. He had previously been Counsel for the Trustees and, before that, General Counsel to the New York, New Haven & Hartford R. R. Blanchette (S) (Dec. 1, 1978) at 1. Phillip Israel, a witness for LV, had served with the ICC for eighteen years in various capacities. When he left the ICC in 1977, he was Deputy Director, Office of Proceedings, Section of Finance. Israel (S) (Dec. 31, 1978) at 1. Alfred Hesse, Jr., had been associated with R for more than thirty-nine years, primarily in legal and marketing capacities. He became the President of R in 1977. Hesse (S) (R's Probable Actions) (Dec. 31, 1978) at App. 1. [**] The GPs do not contest that these Ts would have become cashless on these dates, assuming the absence of financial assistance, nor do they contest AA's cashlessness date. [*] As R points out, however, one GP witness testified that unaffected shippers would have pressed for speed. 2 RF at 245-246, citing Corber (D) (July 8, 1980) at 344, 8 R.App.Ex. 77. [*] Indeed, the GPs themselves assert that the threat of cashlessness is the core of the Ts' argument that labor and public bodies would have accepted unfavorable terms in the alternative scenario. 6 GP at U 2; 1 GPRB at C 1-2. [*] The CUE Opinion actually held that in the Rail Act world the government was permitted to "keep the railroads active for such a reasonable time as was necessary to effect a solution". CUE Opinion, 439 F.Supp. at 1378. Because Congress was the prime actor in the solution, there was certainty that some solution could be found and that the Ts would be compensated for any damages in awaiting it. In the alternative scenario, however, where Congress by hypothesis could not itself take control of the Ts' lines, it is a fair extension of the CUE Opinion to allow the government to pursue a course of action that has a "high likelihood" of success within a reasonable time. [*] It might be argued that Congress could have avoided this danger by granting subsidies large enough to prevent erosion of creditors' rights. This argument is implausible, however, in view of the enormity of the cost. More important, such subsidies would create a stalemate, which Congress would eventually have to break. If continued indefinitely, they would have amounted to a purchase of Ts' properties by Congress under an installment plan — a forbidden "second-best Rail Act". Cf. RRB at 150 n.1. [169] Of course, postulating that Congress would have limited its financial assistance does not necessarily establish that reorganization courts would not have ordered some Ts shut down. Some courts, for example, might have considered the chance of successful sales too remote. We believe, however, that the sales would have gone through. [170] Aage Clausen is Professor of Political Science at Ohio State University. He is a member of the Council of the Inter-University Consortium of Political and Social Research, and Chairman of its Advisory Committee on Elite Studies. Clausen has published several articles on the policy decisions of members of Congress. Clausen (S) (Nov. 1, 1979) at 1-3. [171] Sections 213 and 215 of the Rail Act itself, 45 U.S.C. §§ 723, 725, provided aid, but it was limited in amount, and in duration as well (until the date of conveyance). Of the legislation cited by Clausen that actually involved the commitment of federal funds, both the Transportation Act of 1958, Pub.L.No. 86-625, 72 Stat. 568 (amending 49 U.S.C. §§ 1, 13, 13a, 15a, 203, 501-510), see Clausen (S) (Nov. 1, 1979) at 18-19, and the Emergency Rail Facilities Restoration Act ("Hurricane Agnes" Act), Pub.L.No. 92-591, 86 Stat. 1304 (Oct. 27, 1972), see Clausen (S) (Nov. 1, 1979) at 26-27, were one-shot responses to specific needs. The High Speed Ground Transportation Act of 1965, Pub. L.No. 89-220, 79 Stat. 893 (1965), see Clausen (S) (Nov. 1, 1979) at 19-21, provided funds for research and development by the Secretary of Commerce, not grants to carriers, and was limited as to both amount, Pub.L.No. 89-220 § 11, and duration, id. § 12. Although the Act was thrice extended, these extensions also carried time and expenditure limits. See Pub.L.No. 90-423, 82 Stat. 424, §§ 1(e), 1(f) (July 24, 1968); Pub.L.No. 91-444, 86 Stat. 915 (Oct. 13, 1970); Pub.L.No. 92-348 § 3, 86 Stat. 462 (July 13, 1972). The Rail Passenger Service Act of 1970 (Amtrak Act), Pub.L.No. 91-518, 84 Stat. 1327 (Oct. 30, 1970), see Clausen (S) (Nov. 1, 1979) at 21-23, provided guarantees on PC's obligations to stave off cashlessness. The authorized guarantees, however, were limited to $125,000,000. 45 U.S.C. § 662(e). While the Act could, nevertheless, be read to indicate a congressional willingness to pump needed cash into bankrupt railroads indefinitely, see id., this is not the best reading. More likely, the Act was a stopgap, passed hurriedly in the last days of Congress, leaving the next Congress to devise a permanent solution if necessary. See H.R.Rep.No. 1770, 91st Cong., 2d Sess., reprinted in [1970] U.S.Code Cong. & Ad.News 5983; id. at 5987-88 (minority view). See also 116 Cong.Rec.S 44066 (Dec. 30, 1970) (remarks of Sen. Cotton); id. at H 43334 (Dec. 22, 1970) (remarks of Rep. Randall). In addition, Congress may have thought that a permanent legislative solution would never be necessary, that the railroads could be successfully reorganized after a single infusion of federal funds. See id. at S 44065 (Dec. 30, 1970) (remarks of Sen. Hartke). [172] See Northeast Rail Transportation: Hearings on H.R. 6591, H.R. 4897, H.R. 5822, H.R. 5385, H.R. 6880, H.R. 7373, & H.J.Res. 50 before the Subcomm. on Transportation and Aeronautics of the House Comm. on Interstate and Foreign Commerce, 93d Cong., 1st Sess. 231, 238 (1973) (remarks of Secretary Brinegar); id. at 251 (statement of Jervis Langdon). [173] The GPs argue that consideration of such legislation in the alternative scenario is legally impermissible. 6 GP at T 219-222. They contend, first, that the Ts may not rely on changes in the law that were unforeseeable in February 1973. Second, they contend that the Ts may not add value to their property by claiming benefits that would have accrued from courses of action the government rejected or never considered, since the government did in fact enact a specific package of benefits, the Rail Act itself. Finally the GPs argue that the Ts may not assert that any of the procedural benefits of the Rail Act would also have been present in the alternative scenario. As to the first argument, it has been clear all along that interim aid was foreseeable. Congress would not have sat idly by while the Ts shut down. CMV Opinion, 445 F.Supp. at 1009. And it is inconsistent to claim, as the GPs appear to, that interim aid was foreseeable, but that congressional action to limit it was not. The other arguments are adequately addressed by noting that in the CMV Opinion, id., we specifically permitted regulatory reforms to be taken into account. The only congressional actions that we removed from consideration there were the purchase of the Ts' lines or the equivalent underwriting of their continued operation into the indefinite future. [*] It should be noted that the Bankruptcy Code and Milwaukee Act are also of general applicability, see Milwaukee Railroad Restructing Act, § 17, 45 U.S.C. § 915, and thus could apply to larger transactions. [174] The Chessie Study itself took only three months to complete. White (D) (July 18, 1977) at 69, EL App. K at 240. [*] EL offers comparisons to several mergers that are not in fact comparable. See 2 ELF at 468-77. R states that the evidence shows public bodies could act quickly when necessary, and cites testimony of witnesses Hesse and Ronan. 2 RF at 224. But the evidence they cite does not bear them out. [**] One GP witness, Clausen, analyzed several statutes and concluded that federal aid would have been in the form of loans. Clausen (S) (Nov. 1, 1979) at 33. But the precedents cited are equivocal. In any case, the closest analogy is the Rail Act itself. [**] As to traffic protection, the GPs offer no evidence showing what conditions would have been imposed or what effect such conditions would have had on the earnings values of the T's lines. [**] For similar reasons, we also find it implausible that the ICC would have conditioned its approval on any solvent's acquisition of predominantly passenger lines, leaving it to the public bodies to operate the passenger service on those lines, see pp. 1333-1334, infra. Substantial doubt about the ICC's willingness, or even its power, to impose such a condition would have induced the public bodies to negotiate in earnest for the purchase of the passenger lines and operations long before the ICC reached its decision. Having thus "tipped its hand", a public body would have been unable to persuade the ICC to require solvent acquirers to buy losing passenger lines. [**] LV argues that the ICC had for many years denied that it had power to restructure the Northeastern rail system. LV at 27. This is irrelevant, since the witness on whom LV relies admitted on cross-examination that the ICC's denial applied only to proceedings it initiated. Israel (D) (May 14, 1979) at 88-89, 25 GP Supp.App. at 591-92. [**] We do not, of course, hold that any of this opposition would necessarily have been successful. Much of it most likely would not. For present purposes it is enough that all the Ts' dispositions would have encountered opposition, which would have slowed the process. [*] LV presented the testimony of Phillip Israel, Deputy Director of the division of the ICC that would have handled LV's petition. Israel (S) (Dec. 31, 1978) at 1-2. Israel stated that the ICC would have recommended against consolidating LV's application with the disposition of PC, in part because such a consolidation would be beyond the Commission's power. As Israel himself admitted, however, this reasoning was flawed. See note 204, supra. Moreover, as the GPs correctly point out, Israel's expert testimony is no more credible than the contrary view taken by GP witness Brooks. See 2 GPRB at U 21-22. [175] Shannon cited, as important examples of necessary "tag ends", completion of financing and "formalization" of sales contracts. Shannon (D) (May 24, 1979) at 575, EL App. K at 187. [176] Shannon (D) (May 24, 1979) at 575, EL App. K at 187. [177] See 2 ELF at 470. [178] LV did not conduct any passenger operations. It does not contend that its Aldene to Hunter segment, over which CNJ operated passenger service, would have been treated by a public body as a passenger line. Rather, it was included with lines to be acquired by the JT owner, or, in the absence of the JTs, included with LV's freight lines in proffers for private acquisition. Response of the Erie Lackawanna and Lehigh Valley Trustees to the Court's Letter Dated September 24, 1981 Concerning "Predominantly Passenger" Lines in the North Jersey Terminal Area (Oct. 7, 1981) at 4, Sp. Ct. Rep. at N 37257. [179] This figure does not include the value of those portions of CNJ which the JTTs designated as part of the NJ/NYTerm. CNJ claims $100 million for its entire system. [180] An unspecified amount would have to be added to this figure for CNJ's main line, which the JTTs designated as predominantly passenger. PC also alleged an acquisition price of some $667 million for the rest of its Northeast Corridor properties from Massachusetts to Maryland, portions of which the JTTs also assumed would have been acquired by public bodies. [181] This figure in part represents the GPs' estimate of the excess of terminal costs over the terminal revenues that would have been collected from switching charges as set by the JTTs, as well as increased costs accruing from public body acquisition of the predominantly passenger lines and the Northeast Corridor Spine. [182] Many of these arguments apply to AA and CNJ as well. [183] See p. 1334, infra. [184] As precedent for the JTs, the JTTs point to publicly owned terminal railroads in New Orleans, Galveston, Savannah, and Mobile. These, however, have an average length of only 48 miles of main line. By contrast, Philaterm would have comprised 114.2 route miles and the NJ/NYTerm 241.5. Compiled from Storm (S) (JT) (Dec. 15, 1978) App.C. These figures exclude an even greater amount of predominantly passenger lines that would also have been used for terminal operations. [185] In addition to the publicly owned terminal railroads mentioned in note 216, supra, the JTTs find precedent for public involvement in the rail freight business in the 1976 acquisition by the Massachusetts Bay Transportation Authority (MBTA) of certain lines of the Boston and Maine Railroad (B&M), and in the freight operations on the Long Island Railroad (LIRR) conducted by the Metropolitan Transit Authority (MTA). See JTRB at 64. MBTA purchased 270 miles of the B&M's lines in and around Boston and entered into an operating agreement with B&M for the purpose of improving commuter service. See 2 CRA (S) (Dec. 1, 1978) at VI. While most of the trackage acquired consisted of commuter lines used jointly for freight traffic, a few lines which had been used exclusively for freight were also included. Id. at VI 58-65. B&M obtained trackage rights over all the transferred lines, with obligations to share maintenance costs on the jointly used segments and pay the full costs of maintenance on those used exclusively for freight. The acquisition of the freight-only lines apparently was made with the intention of expanding commuter service to them, a right which MBTA secured in the agreement. Although a GPs witness testified that a subsidiary goal of MBTA in effecting the transaction was to remove obstacles to B&M's attempt to reorganize as a freight carrier, see JTRB at 64, nothing about the transaction remotely suggests that MBTA was prepared to undertake freight operations on its own or even to subsidize B&M's freight operations. The MTA's purchase of the LIRR was clearly for the purpose of maintaining passenger service, and its decision to continue freight operations hardly counts as evidence of the strong public commitment to upgrading freight service alleged by the JTTs. [186] This conclusion assumes that the main freight lines would have been sold for continued rail use. If it should turn out that the Ts would have dismantled their freight lines and sold them for scrap, our entire discussion of the acquisition of the JTs properties would become irrelevant. [187] LV's only segment carrying passenger traffic was not proffered for acquisition as a passenger line, see note 210, supra. [188] The GPs again question the permissibility of considering public bodies other than the federal government as potential purchasers. Although we indicated disagreement with their position in the CMV Opinion, 445 F.Supp. at 1037-40, we reserved final decision "to await factual development which will enlighten us as to the true dimensions of the issue", id. at 1040-41. Further consideration of the issue has reinforced our original inclination. It should be noted that public bodies' eminent domain powers do not diminish but rather enhance their bargaining position relative to the Ts, see pp. 1336-1337, infra. Moreover, there are built-in limits to any potential burden upon the federal government's condemnation powers. Consideration of public body purchases is contingent upon a showing that acquisition would have been likely in the absence of the federal government's taking, and the amount of an award is limited to the amount the Ts can demonstrate that public bodies actually would have been willing to pay. In this sense, an award based upon a state's or municipal corporation's willingness to pay is no different from one based upon a private corporation's willingness to pay. Finally, there is nothing unusual about state law influencing the amount paid by the federal government in a condemnation proceeding since state law largely defines property rights. State regulations such as zoning ordinances diminish the value of some parcels of land and increase the value of others. In fact, the burden imposed by such public body action is less limited than that made possible by consideration of public body acquisition since the latter must be backed by the public's willingness to pay whereas the former is not. Cf. p. 1359, infra. [189] The GPs argue in the alternative that if consideration of changes in state law necessary for acquisition is to be permitted, then changes enhancing the power of public bodies in ways that might improve their bargaining position, for example, by authorizing quick-take condemnation, similarly should be considered. With this we agree as well, as intimated in our CMV Opinion, 445 F.Supp. at 1041. [190] The GPs' citation of T witnesses indicating that the absence of the Rail Act was itself unforeseeable as support for the proposition that requisite changes in state law were thus not reasonably foreseeable, 4 GP at Q 23-26, requires no extended discussion. The very premise of the alternative scenario, as in any condemnation proceeding, is that value is to be determined under the hypothetical assumption that the taking had never occurred. See pp. 1214-1215, supra. [191] For example, while EL's contract with New Jersey provided for a return on the net salvage value of facilities that would no longer be needed if passenger operations ceased to exist, it provided no return on facilities required for freight traffic even though used jointly by the passenger service. [192] This extended history of financial support culminated in 1978 with the State's purchase from Conrail of a substantial portion of EL's commuter lines under the Rail Act's 900-day option. The price, however, was much less than the $114.4 million EL seeks here. New Jersey paid only $8.8 million for the 125 miles of EL passenger lines it acquired in 1978. Compiled from Ex. PCT-NJ-100 at 02695, ELP App. 34. [193] 4 GP at P 86. Cf. id. at 70-71 (New York). [194] Id. at 79-80; see id. at 4-5; id. at 66 (section of argument described as reinforcing pricing conclusions). The GPs' reply brief provides a presumably exhaustive list of alleged impacts of their evidence concerning states' financial abilities, 2 GPRB at K 18-19. Frugal pricing policy receives first mention, followed by the claim that public bodies would first have pursued all alternatives to acquisition, discussed, supra, at pp. 1317-1320, contentions as to switching charges, cf. pp. 1342-1346, infra (trackage rights fees), and timing, see pp. 1293-1311, supra, neither directly relevant to this question, and the idea that public bodies would set priorities, buying only lines with high cost-benefit ratios, see also 6 GP at Y 9-21. [195] It is necessary that we include in this figure all public body lines, including those of Ts no longer in the case, unless it could be assumed that New Jersey would have purchased EL's lines in preference to any of the others if funds proved inadequate. Since we find the State's capacity sufficient to purchase all the predominantly passenger lines, we need not decide precisely what the State's priorities among the lines would have been. [196] Even if current subsidies fully reimbursed shortfalls in avoidable costs, the Ts contend that public bodies would have assumed responsibility for the overhead costs on the predominantly passenger lines. The extent of the increased financial burden depends upon the trackage rights fees charged to freight traffic. See pp. 1342-1346, infra. [197] We do not mean to suggest that the GPs' estimates of incremental operating costs were accurate. EL hotly contests these figures, arguing that the operating cost shortfall in 1973 according to the GPs' own data was only $10.6 million, ELRB at 272-73, and that state acquisition would have enabled reductions in cost, ELP at 21-22, 26-27, discussed at p. 1319, supra. In addition, the GPs appear to be implicitly double-counting since their estimates of operating costs include amortization of acquisition costs that the GPs present as a separate component of the financial burden. See ELRB at 273. Once again, because we find that the State could have paid even the higher figure, it is unnecessary to resolve the details of this controversy. [198] First Boston's table, First Boston (S) (Jan. 30, 1980) at 42, unlike the modified version reproduced by the GPs in a later section of their brief, 6 GP at U 22, gives the impression that substantial increases in all three taxes are necessary to finance the operating burden. [199] In addition, First Boston's indication that the operating costs would consume 11.1 percent of the controllable budget, First Boston (S) (Jan. 30, 1980) at 43, translates into approximately 4 percent of the total budget, again including the full $114 million of supposed added costs. [200] EL cites a 2.3 percent margin, ELP at 47 n.41, which, after incorporating the GPs' correction, 2 GPRB at K 33 n.57, yields a margin just under 3 percent. [201] The other two bond issues were not targeted to rail preservation and thus are a less useful guide to the most likely result in the absence of the Rail Act. EL's arguments noted in text apply to these proposals with even greater force. [202] EL's suggestion that cancellation of debt owed to New Jersey by the Ts would be an additional funding source, ELP at 58-59, is without merit. The GPs correctly note that to the extent the State ultimately expected to collect those funds, this source would be the equivalent of financing through general revenues, and to the extent it did not expect ever to collect those monies, cancellation of the obligation would hardly have provided any financial remuneration to the Ts. See Netzer (S) (Nov. 1, 1979) at 32-33; 2 GPRB at K 38-39. [203] The GPs' reference in this context to the "ultimately unsuccessful attempt to repeal the 1962 [Port Authority] bond covenant", id. at 31, is misplaced since the hypothetical sources of the toll revenues would have been in New Jersey, not under the jurisdiction of the Port Authority. The record does not disclose any similar limitations on the use of New Jersey toll revenues to finance the purchase of passenger lines. [204] See p. 1297, supra. [205] CUE Opinion, 439 F.Supp. at 1377-78. [206] Of this amount, $4.2 million represented the training expenses for the approximately 1,000 employees engaged in the passenger service and the remaining $700,000 was accounted for by the costs of such items as the development of maintenance and safety procedures, work rules, and scheduling, pricing, and ticketing procedures. Shannon (S) (EL) (Dec. 31, 1978) at 92. [207] If, however, in the nonrail use phase, X should turn out to be so low as to not square with "notions of what constitutes `just compensation'", CMV Opinion, 445 F.Supp. at 1030, we would reconsider the question of whether a New Jersey court would have awarded a premium over X. [208] EL's evidence to the contrary, ELP at 62, 68-69, is far less persuasive. It is no surprise that the officials of the Port Authority whose testimony they rely on would be more fearful of the effect of the PATH decision than our analysis suggests. Stangl's testimony is a more reliable indication of perceptions of officials from New Jersey since he was an Assistant Commissioner of the New Jersey Department of Transportation and Secretary of New Jersey's Commuter Operating Agency. In addition, it is not clear that EL's evidence supports the award of a premium. For example, the memorandum prepared by a Port Authority official cited in ELP at 69 refers to "some significant compensation", which seems to draw on the PATH court's unwillingness to base an award on a nominal scrap value. And the recommendation of another Port Authority official that the State purchase at a liquidation sale in lieu of condemnation, see ELP at 62 n.57, hardly proves that the sale price would include a premium. [209] "Common basis" refers to appraisal values as determined by CRA. See 9 GP at App. X 12, 123-31. [210] Although we could not decide that X and common basis are indeed the same until the nonrail use phase of the case, we need not defer decision as to the reliability of the CRA Study since we find that the Ts have not proved that consideration exceeded common basis. [211] In four transactions, in which no common basis could be arrived at, comparison was impossible; in one, common basis exceeded consideration; in another, the common basis figure included so great a range that CRA could not determine whether it was exceeded by consideration. [212] For example, in CRA's PATH case study, it failed to discover a valuation study, see J. Boyd (D) (July 31, 1979) at 24-26, 9 GP Supp.App. at 243-45, and did not review court testimony by one of the negotiators indicating the value of the properties, Dalton (D) (July 3, 1979) at 25-26 (PATH), 9 GP Supp.App. at 546-47; Ex. GPS-Historical Transactions-118 at 28-30. Appraisals of properties involved in other transactions also went undiscovered, Wise (D) (June 8, 1979) at 126-27 (Cleveland), 21 GP Supp.App. at 716-17; Nelson (D) (June 28, 1979) at 67-68 (WMATA), 15 GP Supp.App. at 252-53. In yet another instance, CRA never obtained the purchase and sale agreement, presumably the most important document of all, Dalton (D) (June 11, 1981) at 101 (Skokie), 9 GP Supp.App. at 549. [213] For example, they interviewed only one of seven participants in the Long Island transaction, 2 CRA (S) (Dec. 1, 1978) at I 16 (listing participants); Lincoln (D) (June 11, 1979) at 89-94 (LIRR), 13 GP Supp.App. at 598-603, neither of the two most important in the B&M transaction, Burrows (D) (June 12, 1979) at 77-78, 9 GP Supp.App. at 427-28, and none of seven in the PATH acquisition, Dalton (D) (July 3, 1979) at 15-16 (PATH), 9 GP Supp.App. at 550. In one case, CRA did not interview the person it had been told was "the person to talk with," Burrows (D) (July 3, 1979) Ex. 30 (East End), 9 GP Supp.App. at 463; see CRA (S) (Dec. 31, 1978) at 94-108 (list of persons interviewed, which does not include Fred Morgan, the person identified), with the result that CRA apparently overstated noncash consideration by over 200 percent. Compare 2 CRA (S) (Dec. 1, 1978) at V 45-46 (quantifying operating subsidies at $175,000 for January to July 1973 as part of consideration) with F. Morgan (S) (Nov. 1, 1979) at 9-10 (indicating minimal increase in operating subsidy). [214] See, e.g., D. Smith (D) (May 8, 1980) at 62-66, 9 GP at App. X 33-34 (CRA researcher interviewed Donal Smith; at deposition Smith expressed shock at the method of inquiry and that his statements were used in litigation). [215] In one case, CRA reported the view of the chief Penn Central negotiator, Burrows (D) (July 3, 1979) at 264-66 (East End), 9 GP Supp. App. at 449-51, ignoring those of two MBTA officials they had interviewed, see Burrows (D) (July 3, 1979) Ex. 29, 30 (East End), 9 GP Supp.App. at 461-63. [216] See, e.g., J. Boyd (D) (July 31, 1979) at 31, 9 GP Supp.App. at 249 (Boyd insisting that a lease was not a lease). [217] This was the source of 60 percent of the $521.8 million consideration in the five transactions considered in Ronan's testimony. [218] For the acquisition of the Harlem-Hudson lines, CRA used $22.6 million, 2 CRA (S) (Dec. 1, 1978) at IV 42, whereas PC trustees had told the bankruptcy court that the amount was only $13.5 million, J. Boyd (D) (June 20, 1979) Ex. 2 at 8 (Harlem-Hudson), 9 GP Supp.App. at 374. If PC's figures are correct, consideration was overstated by $91 million. [219] See 2 CRA (S) (Dec. 1, 1978) at III 32-38 (SIRT). No CRA evidence indicated that these past payments were in any way perceived by the parties to the transaction as consideration. See Nathan (S) (Nov. 1, 1979) at 26-27. [220] In the East End transaction, CRA included subsidy payments that were totally unrelated to the sale of the properties. F. Morgan (S) (Nov. 1, 1979) at 2-4. [221] See, e.g., Prince (S) (Historical Transactions) (Nov. 1, 1979) at 6, 12 (within two years, passenger service would otherwise have ceased on the West End and Harlem-Hudson lines); Dalton (D) (July, 31, 1979) Ex. 2 at 505 (Skokie) (ICC had expressed doubt that service would continue in the absence of the acquisition), 9 GP Supp.App. at 568. [222] See 9 GP at App. X 54-55. We have not been able to locate any evidence offered in defense of either the change or the ten times multiplier itself. Given the formula used to derive a multiplier, see CRA (S) (Dec. 1, 1978) at 50-51, it would seem that unlike the 4.31 factor, the ten times factor was not derived from a particular assumption concerning the number of years that deficits would continue, since such a calculation would not yield a round number. Rather, one might infer that the multiplier was chosen first, and then estimates of the underlying factual assumption concerning the number of years of future deficits were approximated by working backwards. [223] For some Staten Island Rapid Transit properties, a document indicated a value of $7.8 million as compared with a cash payment of $3.5 million agreed upon by the parties. See Caust-Ellenbagen (D) (July 3, 1979) Ex. 2, 9 GP Supp.App. at 493. See also pp. 1329-1330, infra, (ignored B&M appraisal); note 256, infra, (questionable LIRR appraisal). [224] An appraisal that was available to and relied upon by at least one of the parties valued the acquired properties at almost $87 million. 2 CRA (S) (Dec. 1, 1978) at I 22-29. The common basis actually used by CRA cut over $35 million from the land valuation, Seymour (S) (Dec. 1, 1978) at 3-6, 113, for reasons that are not altogether clear and for which support does not appear in the record. Most importantly, there is no indication that any such adjustment was contemplated by either party to the transaction. [225] Ronan also ignored the dependence of earnings values on the Ts' main freight lines upon the passenger lines remaining in rail use, Ronan (D) (July 10, 1979) at 533, 17 GP Supp. App. at 591; Netzer (S) (Nov. 1, 1979) at 15-16. See pp. 1333-1334, infra. [226] D would be fully lost only if competitive bidding drove the sales price close to capitalized earnings value. It would thus be more accurate to say that the deduction should only be so much of D as would affect the purchase price. However, since we conclude that there would have been competitive bidding for ELLV by Chessie and more than one Western bidder, see pp. 1346-1351, infra, we shall speak simply of a deduction of D. [227] Altshuler (S) (Nov. 1, 1979) at 7-10, 25 (Secretary of Transportation for Massachusetts, 1971-75); Whitehurst (S) (Nov. 1, 1979) at 29-30 (past consultant to various state and interstate transportation authorities; participated in developing and negotiating passenger subsidies for New Jersey); Netzer (S) (Jan. 31, 1979) at 46 (advisor to numerous public agencies); Stangl (S) (Nov. 1, 1979) at 19-20 (Asst. Commissioner for Public Transportation in New Jersey, 1974-78); Tennyson (S) (Nov. 1, 1979) at 6 (Deputy Transportation Secretary, Pennsylvania, 1972-1979); Comments and Recommendations of the State of New York on "Rail Service in the Midwest and Northeast Region", Rossi (S) (Dec. 31, 1978) Ex. 1 at 24-25 (New York's study of branch line contributions to profits as support for inclusion of all the State's branch lines in the FSP); Review of United States Railway Association Preliminary System Plan (Pennsylvania Department of Transportation), Ex. GPS-Public Bodies-142 at 15-16 (arguing that branch lines not contributing sufficient profits to Conrail might contribute sufficient profits to other railroads to justify their continued existence). The GPs' argument that public officials would have had reason to fear charges of bailing out the Ts through using public funds to boost the Ts' profits on the rest of their systems is also persuasive. See Altshuler (S) (Nov. 1, 1979) at 15. [228] We here use the phrase "a passenger line" without determining whether EL's passenger lines are to be viewed separately or as a package — an issue discussed at pp. 1338-1339, infra. [229] Waldner (S) (Jan. 30, 1980) at 3-4 defines "exclusively" as "traffic originating or terminating at stations exclusively served by the public body lines". Traffic merely passing over public body lines or served by other lines must be excluded, see id. at 11 (indicating that the figure for all public body traffic would be 54 percent of revenues), since it could in principle be rerouted if the public body lines were abandoned. This evidence was not excluded by the order and stipulation discussed at note 14, supra. [230] EL's response proves the point: The major failing of the above Government Party argument is that it implicitly assumes that a prospective solvent acquirer of the Transferors' main freight lines would have been willing to proceed with such an acquisition without prior assurances from the State that the State would acquire the Transferors' unprofitable passenger business. The fact is that no profitable carrier would have been willing to acquire the Transferors' main freight lines without such prior assurances from the State, and the State would have provided such assurances in order to induce a private acquisition of the critically important main freight lines. ELP at 66 (citations omitted). In fact, the GPs' argument not only recognizes the interdependence of the transactions, but relies upon it to demonstrate that the Ts would have had no choice but to accept any offer in excess of X minus D. [231] See Netzer (S) (Nov. 1, 1979) at 8-9; Tennyson (D) (Mar. 19, 1980) at 156-59, 20 GP Supp.App. at 757-60; Altshuler (S) (Nov. 1, 1979) at 21-22; Jacobs (S) (Nature and Consequences) at 36-39; Prince (D) (Apr. 25, 1980) at 221-22, 16 GP Supp.App. at 765-66. [232] "Public bodies in the alternative scenario would have recognized [the Ts' aversion to following through on threats of a shutdown] and would have behaved accordingly." Altshuler (S) (Nov. 1, 1979) at 22. See also Ex. GPS-Public Bodies-186 at 8 (statement of R Trustees upon review of SEPTA agreement) ("What the bondholders wish this Appellate Court to do is to require that an order be entered against the Trustees forcing them to play Russian Roulette with the public interest, with very poor odds."). Although Schelling noted that a T would have had "virtually nothing at risk in adopting any tactic that, should it fail, condemns [it] to recovering the nonrail use value of [its] properties", Schelling (S) (Oct. 15, 1980) at 27, this comment is inapplicable to passenger lines for which D is significant since shutdown would risk loss of earnings values on the main freight lines. [233] Of course, when a T came back to the bargaining table to negotiate over what, if anything, of value was left of its system, it might be in a better position, but the cost it would first have to suffer would have been too great, and, by comparison with the public body, the benefit would have been too small. Thus the factual predicate of what Schelling implicitly indicated to be a necessary condition for an effective threat is lacking. Schelling (S) (Oct. 15, 1980) at 24-25. [234] The Ts' bargaining power is further reduced by their relative inability to withstand delay, see Jacobs (S) (Nov. 1, 1979) at 20, which is due to both the dependence of the sales of their main freight lines upon disposition of the passenger lines and the interim operating losses they would suffer. [235] The GPs note that public officials would have been inclined to condemn in any event in order to shift responsibility for paying the Ts high prices to the courts. See, e.g., Altshuler (D) (June 6, 1980) at 362, 8 GP Supp.App. at 662; Netzer (S) (Nov. 1, 1979) at 37 ("Payments by a public body of a premium because of [the] possibility [of a premium being awarded by a condemnation court] would subject public officials to great public criticism."). But it seems likely that a state would have paid a negotiated price if it would have been sufficiently low. Since the Ts do not contend that a state would have condemned passenger lines in lieu of paying a price based upon the GPs' pricing theories — and no historical evidence in the record indicates that such a result would have been likely — we assume that a state would have attempted to negotiate an acquisition at such prices. [236] The doctrine of setting off the value of special benefits conferred on the remaining parcel by the taking is followed in New Jersey. See 3 Nichols, supra, at § 8.6211[31]. [237] The argument that no improvements were contemplated — just continuation of the same use to which the lines always had been put — RRB at 37-39, is without merit. The contention turns upon a formalistic interpretation of "improvement" that excludes from its meaning a state's rescuing the Ts' lines. "Improvement" is a relative term that in the takings context is addressed to the question of what would have happened but for the taking. It should be noted that any interpretation to the contrary would have the necessary result not only of foreclosing set-offs for special benefits but also of foreclosing an award for damages to any remaining property in such situations. This much is implicitly conceded by R's citation of 3 Nichols, supra, at § 8.6201, RRB at 38, since the italicized portion in R's excerpt is immediately followed by a phrase indicating that special benefits "are ascertained according to the same principles as damages". Finally, if a formalistic approach were to be taken, it is not clear that the factual requisite to this attempt at distinction is present since a state would only have reason to perform the act of condemnation after a T's provision for abandonment were complete, at which point the use immediately prior to the taking would not be rail use. [238] The GPs' methodology is described in GPs' Sept. 14, 1981, Response at 17-21, Sp. Ct. Rep. at N 36872-76. Briefly, the GPs added revenues attributable to the Boonton and Suffern lines alone, multiplied the total by the excess of revenue over variable cost using a range of variable cost ratios, and capitalized the resulting annual figure using the methodology EL had used in its case. For the EL-LV Western acquisition, the figures range from $71 million to $213 million; for the EL-LV Chessie acquisition, from $83 million to $190 million, GPs' Sept. 14, 1981 Response at Table E, Sp. Ct. Rep. at N 36887. [239] In the context of the GPs' deduction argument, we find no merit in EL's excuse that the record is silent "[b]ecause of the Government Parties' `all or nothing' approach", EL's Oct. 7, 1981. Response at 5, Sp. Ct. Rep. at N 37258. EL has conceded that the passenger lines contribute to the earnings value of the freight lines, and had ample opportunity to present evidence on the feasibility of rerouting in response to the GPs' claims. [240] It may not appear to go far enough since abandonment of the Boonton line would entail a substantial change in EL's and LV's overall operations in the North Jersey region and thus eliminate some fixed costs. For example, Jennison averaged full cost and variable expenses in estimating the impact of a 50 percent increase in traffic, which historically had moved over LV's system, on FL's line between River Junction and Buffalo. Jennison (S) (EL) (Dec. 31, 1978) at 19. But the revenue measure for the Boonton line refers to the revenue generated over the entire EL-LV system. This figure is large because a substantial portion of that traffic involves long hauls. Thus, in the event of abandonment, fixed costs along with variable costs would be eliminated over the Boonton line itself, but fixed costs over the rest of EL's system would not be substantially affected. Thus the compromise ratio of 71.4 percent appears more reasonable, especially in light of EL's concession that the Boonton line was essential for any sale of its freight lines to a solvent railroad. [241] Purporting to be applying the statement in the FSP that: [w]here freight and passenger operations both use a facility, the dominant user should own the facility and bear all the costs of that facility except those which could be avoided if the minority user were not present. I FSP at 40, which the Rail Service Planning Office approved, 43 F.R. 37693, 37696 (1978), pursuant to authority conferred in § 309 of the Railroad Revitalization and Regulatory Reform Act of 1976, 90 Stat. 59, 45 U.S.C. § 715(d)(5)(A), Shannon arrived at an annual figure of just over $700,000 as the appropriate trackage rights fee for EL movements over the passenger lines. The GPs ridicule this figure as being far too low, indicating, for example, that, under it, passenger service pays 7.9 times the amount charged to freight service on a car mile basis. GPs' Sept. 14, 1981, Response at 23-26 & n.21, Sp. Ct. Rep. at N 36878-81. The reason Shannon arrived at such a low figure was that his formula bore only the vaguest resemblance to that suggested by USRA. On lines where passenger traffic was dominant (over 50 percent on a car mile basis), he assigned none of the costs to freight. And on lines where freight was dominant, he assigned only 50 percent of the costs to freight, and even that was not 50 percent of the total costs, but 50 percent of the costs allocated to freight under the ICC's gross ton mile formula. Shannon (D) (June 21, 1979) Ex. 1 at workpapers ELR 300 XXXXXX-XX. Thus, on predominantly passenger lines, some of which have over 85 percent freight traffic on a car mile basis, he assigned less than 30 percent of the maintenance costs to the acquirer of the freight lines. Shannon conceded that his estimate was equivalent to an avoidable cost freight charge, Shannon (D) (May 22, 1979) at 253, EL App. N at 82; see EL's Oct. 7, 1981 Response at 11-12, Sp. Ct. Rep. at N 37243-44, an allocation formula opposite to that suggested by the GPs' public body witnesses. [242] Technically, the analysis in text both here and later in the section when we address the ability of public bodies to extract additional earnings values through trackage rights fees, see p. 1344, infra, is inaccurate in that the effect of taxes is not directly considered. Neither the GPs nor the Ts raised this issue, and we need not consider it at length since it does not change any of our conclusions. The effect of taxes is to reduce the amount by which trackage rights fees diminish earnings values, which are after-tax figures, but to increase the ability of the state to extract additional earnings values by precisely the same amount. To avoid additional complexity in our discussion, we thus do not account for taxes at each step in our analysis. [243] The GPs' argument in no way indicates any threat to the sale of the freight lines unless trackage rights fees set to recoup D would drive the price solvent acquirers would have been willing to pay below the X for those lines. But this qualification is simply the mirror image to the above-noted reservation that deductions from the sale price of the passenger lines could not be so high as to drive the total acquisition price below X for the entire system. This discussion makes the assumption — consistent with the contentions of both parties and the dependence of sales of the freight lines resolution of the disposition of the passenger lines — that there effectively would have been three-party negotiations in each instance among the public body, the T, and the solvent acquirer, at which purchase prices and trackage rights fees would be determined. See, e.g., Hesse (S) (JT) (Dec. 15, 1978) at 30; Tennyson (S) (Nov. 1, 1979) at 6-7. [244] It is arguably more plausible that the trackage rights fee method of recovering D would have been chosen since otherwise solvent acquirers of the main freight lines, having already paid the Ts for the full earnings values, would be subject to additional extractions from the public in the future. By paying less for the main freight lines with the expectation that the state would charge trackage rights fees that would recover D, the solvent acquirers of the Ts' freight lines guarantee that they never will be subject to double payment. The GPs in oral argument note the seeming inconsistency between the Ts' insistence that no solvent acquirer of the freight lines would have considered acquiring even fully subsidized passenger operations on the ground that the public could not be trusted to continue the subsidies at the necessary levels in the future and the Ts' implicit assertion that the public could be trusted to keep the level of trackage rights charges at low initial levels. Tr. at 320-21. But there is a difference between a subsidy that requires renewed appropriations and thus appears on the debit side of a state's budget and a trackage rights fee based upon a predetermined formula for services rendered. In any event, the GPs have offered no basis for quantifying such a further deduction. [245] EL's $2,788,000 figure can profitably be compared to 56.2 percent of the unadjusted maintenance of way total of $4,775,000 — $2,683,550. It is not surprising that the two figures are of similar magnitude since the former did represent a measure of freight's share of maintenance costs. [246] EL's assertion that such costs are already included in the the trackage rights charge incorporated into their pro formas, EL's Oct. 7, 1981, Response at 16, Sp. Ct. Rep. at N 37248, is irrelevant since we are recomputing that charge and simply wrong since Shannon's $733,000 figure does not purport to include anything besides maintenance of way costs. See Shannon (D) (May 21, 1979) Ex. 1 at ELR 300 206134, 18 GP Supp.App. at 745. EL concedes as much in explaining the derivation of Shannon's figure earlier in its own statement. EL's Oct. 7, 1981, Response at 10-12, Sp. Ct. Rep. at N 37263-65. [247] The GPs object to our consideration of these proffers on the ground that LV's estimate of its own salvage value was so large that EL would never have accepted the balance of the total as adequate. We think it unlikely that if a sale of the combination would have produced the best result for both EL and LV, they would have foregone this sale because of inability to agree on a fair division of the proceeds. At oral argument counsel stated the two roads would have divided the proceeds on the basis of the relative net salvage values of the properties. Tr. at 67, 107. [248] EL includes, properly we think, Union Pacific as a potential bidder. Although in 1973 that road's eastern terminals were Omaha and Kansas City, it reached Chicago via a friendly connection with Chicago & Northwestern. Shannon (S) (EL) (Dec. 31, 1978) at 21. [249] In both instances, Chessie's disadvantage could have been alleviated by acquisition of R's class (a) lines. [250] Darling's attention had been attracted when he noted that EL had an unusually large car hire deficit, a situation which could be corrected by a road, such as Santa Fe, having adequate capital. This led him to go further. Darling (D) (Jan. 31, 1980) at 16-19, EL App. A at 93-96. [251] Barriger, as a GP witness, testified that Santa Fe had concluded not to pursue the acquisition for fear of possible retaliation. However, he conceded on cross-examination that this conclusion did not describe what Santa Fe's attitude would have been in the absence of the Rail Act. Barriger (D) (Jan. 31, 1980) at 6, EL App. A at 105. [252] Noting this, Shannon proposed that the EL earnings predicted in the study should be adjusted upward in various respects. One was that the Chessie study used 1976 cost levels but did not take into account rate increases beyond October 11, 1975. A conservative assumption that a 5 percent increase would be in effect for the last three quarters of 1976 would increase annual pre-tax income by nearly $10 million. Shannon further adjusted for what he considered to be Chessie's overestimate of normalized maintenance of $10 million. Shannon also pointed out a mathematical error in the analysis of joint facility contracts which would have increased pre-tax net income by $4.2 million. The last adjustment is apparently noncontroversial and the others were not controverted. As EL points out in its reply brief, ELRB at 20 n.12, the net result of these three adjustments is an increase of Chessie's figure for pre-tax income before fixed charge from $30.7 million to $54.8 million — as compared with Shannon's own figure of $53.7 million for the fifth year. Shannon believed that additional adjustments for deficiencies in the rerouting analysis and underestimates of traffic increases should have been made, but these were not quantified. The GPs seeks to discredit the importance of the Chessie study by suggesting, 1 GPRB at F 13-14, that at a 20 percent discount rate, one of the alternatives posed in the Chessie study, the capitalized earnings of EL would have been only $34.5 million as against $52,936,444 proposed in the March 1, 1976, Green Book. However, the GPs have never proposed a discount rate as high as 20 percent; their witness Smith's figure was 10.45 percent. The Chessie study showed the results at discount rates of 0, 10, and 20 percent to indicate a range, without endorsing any specific figure. The GPs' argument also fails to take account of Shannon's adjustments, some of which, as indicated, do not appear to be disputed. [253] Barnett admitted that he had not studied EL's rehabilitation figures and, indeed, that he was generally unfamiliar with EL's case. Barnett (D) (May 30, 1980) at 56-57, EL App. H at 8-9. [254] Exhibit 8 to Shannon's testimony for EL is a letter dated January 27, 1975, from the President of the Missouri Pacific objecting to designation of PC east-west lines to Western carriers and stating that if this should be given serious consideration, "MoPac hereby suggests and proposes that PC and EL segments be granted to MP either exclusively or as a participating owner and user as shown on the attachment." Unfortunately the "attachment" is not attached. [255] EL Map 3 shows how well the routes of EL-LV between Sterling and the Northern New Jersey terminal would have fit the Chessie system. In addition to providing a direct line between Chicago and New York in place of the multi-carrier route through Shippensburg, Pennsylvania, the acquisition would have given Chessie an effective route south of Lake Erie between Buffalo and Chicago, thus enabling it to abandon its operation across Canada. EL's branch between Warren and Cleveland, Ohio, would have shortened Chessie's route between Cleveland and the East. [256] The latter figure is for a Chessie acquisition of EL and LV. [257] Jennison is a Vice President of Wyer, Dick & Co. and has been associated with that company since 1961. Prior to that, he held several management positions with the Bangor & Aroostook Railway. [258] We have been particularly unimpressed by the adjustments to their study proposed by GP witness Scheer. Scheer (S) (Jan. 30, 1980). These were torn in shreds in EL's opening freight brief, 2 ELF at 319-37, and the GPs' reply brief made almost no effort to rehabilitate them. [259] Indeed, the GPs' criticism of Laurie for not having introduced more variables with respect to the changed ownership of other roads contrasts strangely with their objection to EL's having submitted four separate disposition proposals. 1 GP at A 69. [260] The figures for 1969-75 are as follows: Year Revenue Ton Miles 1969 15,505,454 1970 15,039,019 1971 13,786,020 1972 13,293,395 1973 14,206,262 1974 12,882,200 1975 9,562,398 Corresponding figures for LV are: Year Revenue Ton Miles 1969 3,140,579 1970 2,915,181 1971 2,474,872 1972 2,653,256 1973 3,230,572 1974 3,602,720 1975 2,911,924 R. Murphy (S) (Jan. 30, 1980) Apps. 6, 8. [261] The GPs also attack Shannon's estimate that traffic diverted from trucks in a Western acquisition, an addition to revenue that the GPs sharply question, could have been handled at a revenue ratio of only 60 percent. They assert that at the very least Shannon should have used Jennison's incremental ratio of 71.4 percent, which would have reduced annual net railway operating income by approximately $2 million. EL responds that the 60 percent figure is justified since small amounts of traffic are involved, to wit, a 1½ percent annual increase, the average length of haul is great, and there would be no interchange with any other railroads. It also cites ICC rulings that losses or gains of less than 10 percent of traffic would be accompanied by a 50 percent reduction or increase in related costs. See Pennsylvania R. R. — Merger — New York Central R. R., 330 I.C.C. 328, 347-50 (1967); Norfolk & Western R. R. — Chicago & St. Louis R. R. — Merger, Etc., 330 I.C.C. 780, 817 n.21 (1967); Burlington Northern, Inc. — Control & Merger — St. Louis, San Francisco R. R., 360 I.C.C. 783, 1110, 1123 (1980) (47 percent of lost revenues would be saved by lowered costs). [262] EL also makes an extended argument to show that its application of system average switching costs in the EL-LV combination overstated rather than understated such costs. This is because the average, reflecting supposedly higher New Jersey costs, was applied outside New Jersey, whereas New Jersey switching costs were based on the JTs studies. In light of our conclusion with respect to the JTs, it is unnecessary to consider this argument. ELRB at 124-26. [263] Because of differences in the amount of property to be conveyed, these amounts differed as between the different acquisitions. [264] This yard was not included in the Chessie acquisition since Marion is west of Sterling, Ohio, where EL and Chessie connect. EL also points to testimony by Shannon, Shannon (D) (July 10, 1979) at 1096-1103, EL App. O at 487-94, that a Western acquirer could have achieved savings comparable to these contemplated for the reconstructed Marion Yard by using its own yards at or west of Chicago, without having to entail the capital costs estimated for the Marion Yard. However, Shannon conservatively took no account of this. [265] This yard was not included in the EL-LV combinations, since most of the service previously performed at Croxton would be performed by LV's Oak Island Yard. [266] In making this seemingly high estimate Shannon compared the location of EL traffic officers and those of Chessie, Norfolk & Western, and Santa Fe. Using a color coding, he found that virtually all EL's traffic offices were duplicative and that substantially all their work could have been done by employees of the purchaser. Shannon noted that his estimates of savings in traffic department costs were supported by the Chessie study. Shannon (S) (EL) (Dec. 31, 1979) at 84, id. at Ex. 13. [267] Moon is President of Albert E. Moon Analytical Services, Inc. He is the former director of Transportation Program Development in the Transportation and Industrial Systems Center of Stanford Research Institute. [268] Shannon's study was property-specific in the case of the traffic offices. We frankly doubt "the benefit of looking at the employees, offices or properties hypothetically merged" in the case of many of the overhead accounts. Looking at the chief executive of EL or at his office is hardly necessary to support the conclusion that in a merger he would have disappeared as such; the question is whether he might not have continued in some other executive capacity. Experience is a better guide on this question than visual inspection. [269] Wyer Dick also studied the Union Pacific — Rock Island merger, which never came to fruition, and the Louisville & Nashville — Chicago & Eastern Illinois merger, which went through on a radically reconfigured basis. [270] The GPs also quoted a 1972 report prepared by the Staff of the Senate Commerce Committee on the Pennsylvania — New York Central merger, which minimized the savings expectable in railroad mergers. EL, however, cited a much earlier report by the Senate Commerce Committee, 86th Cong., 1st Sess., Report No. 445 at 244 (June 1961), saying that "[t]he savings in organizational expenses do not constitute the major areas of savings, but are perhaps the least controversial." [271] This was what Chessie did. As indicated, Shannon testified that Chessie's results were comparable with his own, and we have not discovered any contradiction of this. See Statement of the EL Trustee in Response to Court's May 21, 1981, Memorandum (June 1, 1981) at Part I, Sp. Ct. Rep. at N 35802-06. [*] Since none of the proposals for continued freight use advanced by EL involved the reconstruction of the Poughkeepsie Bridge, which would be essential to revival of the Maybrook gateway to New England or operation by the acquirer down the east bank of the Hudson, the much bandied about $100 million figure is illusory, except insofar as Rossi stated that "[i]nternally our position was that if Chessie insisted upon up to $100 million to make the acquisition as proposed in the Final System Plan, that money would have been committed", Rossi (S) (Dec. 31, 1978) at 8. [**] The EL-LV proposals involve converting most of the EL line east of Waverly from double to single track and using the LV line from Waverly to Newark to carry the bulk of the traffic between points west of Waverly and the Northern New Jersey — New York City areas. While the line through Port Jervis would be equipped with a modern signalling system and would remain capable of carrying substantial freight traffic, it may be doubted whether this is what Commissioner Schuler had in mind when he spoke of maintaining the Port Jervis line "as a main east-west freight line". [*] See Memorandum of the EL Trustees in Response to the Court's Letter of June 19, 1981, (July 10, 1981) at 6-7, Sp. Ct. Rep. at N 36159-60. [**] The GPs note the "brief time" allowed them to comment on the EL's July 10 submission. GPs' Response to July 10, 1981, Filings of Erie (July 24, 1981) at 1, Sp. Ct. Rep. at N 36345. We do not believe that the two weeks allowed the GPs were insufficient. Given the essentially simple approach to the problem taken by EL, the paucity of criticism contained in the GPs' Response indicates to us that more time would not have profited the GPs significantly. More important, if the GPs believed they needed more time, they could have requested an extension as they have done on other occasions. [**] There is no inconsistency between this holding and our order with respect to evidence tendered by North Penn and Delaware & Bound Brook on October 15, 1980. See Memorandum and Order on Motion of GPs to Strike Direct Evidence Tendered on October 15, 1980, by North Penn and Delaware & Bound Brook (November 13, 1980), Sp. Ct. Rep. N 22612. In their initial evidentiary submissions, North Penn and Delaware & Bound Brook sought to show that their lines would have been acquired by a public body for a price based on their net liquidation value. Two years after the close of direct evidence from the Ts, North Penn and Delaware & Bound Brook sought to introduce evidence showing that their lines would have been acquired by a private railroad on the basis of earnings value. The evidence submitted covered a wide range of topics wholly unrelated to and heretofore unmentioned by the parties' initial witnesses. But for an ambiguous letter sent by the court in June of 1980, see Memorandum and Order of Nov. 13, 1980, supra, at 5-8, Sp. Ct. Rep. at N 22616-19, we would have granted the GPs' motion to strike this evidence. The differences between EL's position and that of North Penn and Delaware & Bound Brook are glaring. Rather than seeking to present a wholly new and inconsistent case relying on evidence unrelated to prior submissions, EL's July 10, 1981, memorandum discusses a contention previously endorsed by two witnesses and relies completely on materials already in the record. [272] For example, the cost of originating or terminating a TOFC in the JTs study was projected to be $124.99 per car. Under Jennison's study the historic cost was $89.69 per car for EL and $75.59 per car for LV. As a result of this and other adjustments, the total cost of operating EL's retained traffic fell from $234,983,000 in the EL-LV without JT Western acquisition to $229,371,000 when the JT assumption was abandoned. The total cost excluding overhead of operating LV traffic fell from $57,421,000 to $53,683,000. Workpapers to Memorandum of the Erie Lackawanna Trustees in Response to Court's Letter of June 19, 1981, (July 10, 1981) at 3-4. Similarly the total cost for the EL-LV Chessie acquisition fell from $165,716,000 to $160,130,000 for EL and from $50,083,000 to $46,343,000 for LV when the JT assumption was abandoned. Id. at 5-6. [*] EL and the GPs agreed to a great extent over how this adjustment should be made. The one point they could not resolve among themselves is whether revenues and costs associated with six particular traffic movements over LV's lines and certain other LV/CNJ revenues should be adjusted. We believe that EL's treatment is the correct one since it is not proper to take out revenues and costs for traffic that were not originally included. See Reply of EL to GPs' Response to Court's September 24, 1981, Letter (Sept. 24, 1981) at 4-5, Sp. Ct. Rep. at N 37094-95. The GPs also contended that additional New Jersey property taxes for the LV of $403,000 should be included as a cost of operating the EL-LV lines and that maintenance of way expenses for the Phillipsburg to Port Morris line should reflect the fact that the line is projected to be upgraded to a heavy density main line. We agree. The GPs finally argue, GPs' Response to July 10, 1981, Filings of EL at 8-9, Sp. Ct. Rep. at N 36352-53, that the points at which CNJ and EL-LV interchanged traffic would have been the subject of negotiation and that without determining those interchange points, the division of revenues between EL-LV and CNJ can not be determined. The implication is that CNJ would have been able to lengthen its average haul at the expense of EL-LV. However, it is equally if not more likely that CNJ would have lost rather than gained in these negotiations since the combined EL-LV would have been a stronger competitor than either railroad had been individually. We assume, therefore, that the historic traffic patterns would have remained intact. [273] Applications to abandon operation of these ferries were pending before the ICC. [274] This segment included a 6.5-mile branch from Pittsfield to Saline, Michigan, to service a Ford plant. All references to the segment include the Saline branch. [275] The GPs also argue that the ICC would not have permitted any segmentation of AA. It may well be that AA would not have been entitled to abandon the losing segments if the Toledo-Ann Arbor segment produced a profit sufficient to cover the deficits on the losers and provide a fair return on the entire investment. However, no one contends that this was the case. [276] In light of our earlier conclusions, see pp. 1366-1369, supra, there would be no need for us to consider whether Michigan had the statutory capacity to acquire and operate rail lines. In any event an affirmative answer is compelled by the Advisory Opinion of the Supreme Court of Michigan, cited infra. [277] Although AA does not suggest this, some of the profits might remain available because of Michigan's ability to finance the purchase at lower interest rates commanded by tax exempt securities as compared to those a private acquirer would have to pay. [278] It is significant that, in contrast to the parade of officials from the Northeastern states whom the GPs called to testify that these states would not have been willing to purchase Ts' freight lines, they called no official to give similar testimony with respect to Michigan. While AA likewise called no Michigan official, this was unnecessary in light of its documentary case. [279] The GPs do emphasize that with the Rail Act, Michigan pleaded to the federal government for funding, 4 GP at R 176-78, but for the reasons outlined previously, see p. 1321, supra, such statements provide little indication of what a public body would have done if federal funds had been unavailable. [280] An example much discussed in the briefs concerns the shipment of cement from Dundee, Michigan, (on AA's Toledo-Ann Arbor segment) to Florida. While AA was under D,T&I's control, although AA could have carried this as far as Toledo, this traffic was carried by AA only to Diann, Michigan, where AA connected with D,T&I, thence over the entire length of D,T&I to points in Southern Ohio, and thence over the lines of other carriers to points in Florida. If an acquirer of one or more of the IFLs purchased AA, it would follow the same policy of obtaining the long haul. If all the IFLs were acquired by one purchaser as proposed in AA's initial case, it would have gotten the longest haul by carrying the traffic to Toledo and thence to Potomac Yard near Washington, D.C., where it would have turned the traffic over to the Richmond, Fredericksburg & Potomac R. R. From the standpoint of D,T&I it would be inconsequential whether a single purchaser acquired all the IFLs or a different purchaser acquired the Detroit-Cincinnati IFL. Either would have been in a position to divert much of the substantial traffic that AA had exchanged with D,T&I at Diann. See AA at 68-70. [281] We discuss below the GPs' argument that it is impermissible for us to consider an acquirer of IFLs. [282] Indeed, one of the GPs own witnesses with respect to AA, Evan Allen, had prepared a valuation study of a possible acquisition of D,T&I when he was a senior financial analyst for Southern, in which he included the value of revenues from D,T&I traffic that could be routed onto Southern's line. Allen & Kozu (D) (July 10, 1980) at 754, 1 AA Supp.App. at 19-21. [283] The GPs also contended that the diversion featured in Laurie's initial testimony, see note 318, supra, would not have been feared since the service via Potomac Yard would have been slower than that historically provided. Laurie effectively refuted this in his rebuttal testimony, Laurie (S) (Oct. 15, 1980) at 4-6, by showing, on the basis of timetables, that the time would have been the same. [284] See note 57, supra. [285] Laurie added, Laurie (S) (Oct. 15, 1980) Ex. 5 at 6, in line with Shannon's testimony, that the Seaboard Coast Line through its subsidiary, the Louisville & Nashville, was another possible acquirer of the Detroit-Toledo-Cincinnati segment and that if it succeeded in making that acquisition, it would have had an interest in the Ann Arbor-Toledo segment similar to Southern's and would have had roughly similar off-segment earnings. While this suggestion produces a second bidder for the Detroit-Cincinnati IFL, it does not add to the number of bidders for the Ann Arbor-Toledo segment. If the diversion of southbound traffic on the segment to Potomac Yard was as important as Laurie's initial study indicated, we would suppose that a purchaser of the Chicago-Toledo-Cleveland-Buffalo-New York IFL might have had an interest in the segment, but AA has not suggested this possibility and we shall disregard it. [286] The situation would be different if the Trustees of PC and AA had offered the Detroit-Cincinnati line and the Ann Arbor-Toledo segment as a package, but there is no evidence that this was ever considered. [287] The GPs attempt to weaken this, 1 GPRB at 3 n.4, by arguing that Wilmot uniformly disregarded potential diversions of traffic to PC lines because of his belief that these would be offset by potential losses and therefore "implicitly" credited the IFL with off-segment revenue by not deducting revenues that hypothetically would be lost. Whatever the merits of the argument might be in other contexts, e.g., acquisition of an east-west line by a Western carrier where diversion of traffic to the acquirer might provoke retaliation by other Western lines, the GPs have not pointed out what retaliation D,T&I would have been in position to administer if an acquirer of the segment routed traffic to Toledo and thence to Cincinnati. [288] The force of the answer is diminished by the fact that, as owner of 99 + percent of AA stock, D,T&I was not truly concerned whether the plant was served solely by AA or by itself as well. Much the same comment applies to PC, which owned 99 + percent of D,T&I. [289] We note in passing the inconsistency of the GPs' argument about the PC on this phase of the case and their objection, discussed above, to our considering the acquirer of a PC line as a potential bidder. [290] See note 289, supra. [291] Jennison states that his purpose was to permit Laurie "to estimate the impacts upon the profit levels of Penn Central and the Detroit, Toledo & Ironton Railroad that would have been associated with the abandonment of the Ann Arbor". Jennison (S) (AA) (Dec. 31, 1978) at 2. While this was a meaningful statement for D,T&I, except that Laurie contemplated not an abandonment of the segment but its acquisition by a competitor, it was meaningless with respect to PC, where the problem was to measure the cost of handling added traffic. Implicitly, Jennison assumed that the cost-revenue ratio for moderate amounts of lost traffic is the same as the one for moderate gains in traffic. Since the GPs do not challenge this assumption, we see no reason to question it. [292] The difference in the PC and D,T&I ratios is not due to the choice of different percentages or variability but to the relative importance of different accounts. [293] This was preserved by an order of Judge Thomsen, Laurie (D) (Feb. 3, 1981) 275, 278, 1 AA Supp.App. at 312, 313. [294] The figures come from AA's Quarterly Reports for Freight Commodity Statistics. They show a marked decrease in cement and auto parts origination in 1974 as against 1973 but a slight increase in American Motors jeeps. [295] Although the GPs had no opportunity to answer this argument in briefs, excerpts from the Quarterly Reports introduced in rebuttal would seem to be self-authenticating under Rule 902(5). Alternatively, all the reports may be subject to judicial notice under Rule 201. [296] The GPs also rely on a study made by Peat, Marwick, Mitchell & Co. for the AA reorganization trustee to determine whether AA could attain viability under any one of ten different configurations. One of these, an operation from Toledo to Whitmore Lake, only 9.5 miles north of Ann Arbor, is alleged to correspond closely to the Ann Arbor-Toledo segment. Peat Marwick's conclusion with respect to the viability of this operation, as of all the others, was negative. Laughlin (S) (Dec. 31, 1978) at V 2-3. The GPs also cite a statement of an AA witness that Peat Marwick gave preliminary consideration to the Ann Arbor-Toledo segment itself but rejected this as less attractive than inclusion of the 9.5 miles north of Ann Arbor, Laughlin (D) (June 6, 1979) at 6, 13 GP Supp. App. at 431. AA's reply brief, AARB at 73-76, answers this in detail, pointing out significant differences in the assumptions underlying the Peat Marwick and the Laurie studies. The most important is that the Peat Marwick study assumed the operation of two daily sets of trains and crews, one between Whitmore Lake and Dundee and the other between Dundee and Toledo, while Laurie's study contemplated only one. We find AA's refutation persuasive. [297] L&NE points to a proposal made in 1972 by the LV Trustees, with the approval of the Attorney General of the United States, wherein the LV Trustees offered to pay $1,000,000 for L&NE's stock being held by the Attorney General and to guarantee $2.5 million in L&NE outstanding mortgage bonds, and the Attorney General also approved a credit of $5 million to be applied against the debt of L&NE's parent, CNJ, to the United States. This offer did not take account of the prospect of isolation of the Tamaqua branch. Moreover, it was subject to the approval of the LV and CNJ reorganization courts. L&NE also submitted three offers made by private persons to the Department of Justice in 1975. The value of these offers, which assumed continuance of the Tamaqua branch and contained numerous conditions, is difficult to appraise, and L&NE offered no evidence (other than that contained in some of the offers themselves) as to the financial responsibility of the offerors. Moreover, the federal rule appears to be that a condemnee may not rely on offers to show market value. Sharp v. United States, 191 U.S. 341, 24 S.Ct. 114, 48 L.Ed. 211 (1903); see 4 Nichols, Eminent Domain, § 12.311[2] (1979). Although we would question some of Justice Peckham's reasoning in a case where the condemnee demonstrated the financial worth and bona fides of the offeror, these conditions were not met here. [298] According to Timpany, Sagner told him on July 24 that he had concluded not to proceed with a tender offer. Timpany (S) (Dec. 31, 1978) at 29. But on July 26, when Sagner requested the Peabody study, one of the questions he wanted answered was whether to make a tender offer. Stangl (D) (Mar. 25, 1980) Ex. 18, I CNJ App. at 424. Similarly, Goodstein, who, like Timpany, testified on CNJ's behalf, testified that Sagner was still considering that possibility on July 30. Goodstein advised Sagner against it. Goodstein (S) (Dec. 31, 1978) at 7-9. [299] As we have noted elsewhere, the term "net liquidation value" has taken on something of a specialized meaning in this litigation, one that includes sales for rail use as well as scrap sales. See In the Matter of the Valuation Proceedings, 425 F.Supp. 266, 282; CMV Opinion, 445 F.Supp. 994, 1004. As the term is used here and elsewhere in CNJ's evidence, however, it refers more conventionally to scrap sale value of physical rail assets. However, even that term means different things to different parties, see pp. 1211-1213, supra. [300] These later events also suggest that Sagner's recollection of the July 18 meeting is far more likely to be accurate than Crane's since Sagner's actions are inconsistent with his previously having made a detailed offer for CNJ's property. [301] They are: a. $119,800,000: Goodstein's proposal of July 18 ($25 million cash plus assumption of obligations estimated at $94,800,000). Goodstein (S) (Dec. 31, 1978) at 7. b. $117,907,000: New Jersey Department of Transportation's valuation of CNJ's net liquidation value as of 1974 adjusted to omit PC's half-interest in New York & Long Branch Railroad. Jacoby (S) (Dec. 31, 1978) Ex. 64 at 1. c. $112,981,000: Peabody's valuation of net liquidation value as of December 31, 1973. Jacoby (S) (Dec. 31, 1978) Ex. 62 at ii. d. $105 million: Timpany's proposal of July 26. Timpany (S) (Dec. 31, 1978) Ex. 18. e. $100 million: Timpany's estimate in testimony of the probable sale price. Timpany (S) (Dec. 31, 1978) at 52. f. $100 million: The bondholders' first suggested price at the March 21 meeting. Goodstein (S) (Dec. 31, 1978) at 5. g. $80 million: Sagner's reported statement at the July 24 meeting that he "understood" the State would have to assume obligations. Timpany (S) (Dec. 31, 1978) at 29. h. $79,654,000: Wyer Dick valuation of net liquidation value as of June 30, 1968. Timpany (S) (Dec. 31, 1978) Ex. 19, I CNJ App. at 29-34. Also adopted as a finding of fact by the United States Secretary of Transportation. Timpany (S) (Dec. 31, 1978) at 52. i. $71.5 million: Book value of the conveyed assets. Timpany (S) (Dec. 31, 1978) at 59. j. $70-80 million: Sagner's estimate of net liquidation value at the August 23 meeting. Jacoby (S) (Dec. 31, 1978) Ex. 60 at 2. k. $25 million over all prior indebtedness: The bondholders' second suggested price at the March 21 meeting. Timpany (S) (Dec. 31, 1978) at 5. l. Fifty percent of the face value of the bonds: Sagner's suggested price at the March 21 meeting; estimated to result in $21 million over debt. Timpany (S) (Dec. 31, 1978) at 5. [302] Although CNJ suggests in its brief that Chessie had long been interested in CNJ, e.g., CNJ at 155-57, there is no contention that anyone other than New Jersey would have brought CNJ for rail use. [303] We note, however, that most of the discrepancies between the JTTs' case and CNJ's proffer in the North Jersey terminal area concern CNJ lines not conveyed to USRA and, therefore, not part of this suit. Other discrepancies disappear by virtue of our rejection of the JT proposal, see Part XI. [304] Long before trial we warned the parties that: As at present advised, the Court is not disposed to entertain argument that any transferor would have obtained a strategic advantage over other transferors (as distinguished from such advantages as ending of deterioration, entitlement to interest, etc.) by negotiating or consummating a disposition at an earlier date than other transferors. Second Pre-Trial Order at 4, Sp. Ct. Rep. at N 11801-02. We have since remarked that: While, as indicated by the language, that view is not frozen in stone, we could not properly depart from it without being fully advised of the effect of such a departure on the Government Parties and other transferors. Memorandum and Order of April 23, 1980, at 4, Sp. Ct. Rep. at N 20242. There is room for debate as to whether CNJ's proffer violates the letter or spirit of these admonitions. Nevertheless, the point is moot. The only "strategic advantage" CNJ might have garnered from its timing contentions would be a favored position for certain CNJ properties as compared with other Ts' properties — for example, a conclusion that, because of CNJ's early disposition, New Jersey would have acquired CNJ's Phillipsburgto-Newark line rather than LV's parallel line. Since we have already concluded (without considering the effect of timing) that all of CNJ's properties would have been purchased for rail use, it is unnecessary to decide whether CNJ may invoke its timing allegations to establish that point. [305] CNJ devotes several pages of its brief, with quotations from and citations to the evidence, to its contentions concerning LV, R, and EL. CNJ at 121-24. In contrast, its sole statement concerning PC is this: "Whether PC was truly ready to seek to liquidate (in view of the risks of traffic diversions by the other Transferors, the views previously noted of the Federal Railroad Administration, its own views of a profitable core, and the downside risks of a liquidation elaborated in the CUE Opinion) is an interesting question." CNJ at 124-25 (emphasis added). [306] CNJ asserts that the State had encouraged the bondholders to seek a total shutdown in December 1972. CNJ at 131. The evidence cited for the assertion, however, is only the conclusory impression of Roger Ward, a bondholders' representative, Ward (S) (Dec. 31, 1978) at 24-25, II CNJ App. at 167-68; it is not backed up by any alleged oral or written statements by any State official. Moreover, even if the State did give such encouragement, it did so only in contemplation of a purchase of CNJ by LV and R. There was no prospect of such a purchase in the fall of 1973. [307] Indeed, CNJ goes so far as to request us to find that even without any sale for rail use CNJ would have realized $100 million in actual liquidation for scrap. CNJ at 163. [308] In light of the pending settlement between the GPs and EL, we, of course, do not expect EL to petition for reconsideration nor do we expect the GPs to raise issues relating solely to EL. If the settlement should not be approved, we will entertain applications for reconsideration from EL and from the GPs in respect of EL. [309] A petition for clarification or reconsideration filed by the Trustee of the Ann Arbor Railroad on December 23, 1981, Sp. Ct. Rep. N 38766, has been determined by an order dated January 11, 1982, Sp. Ct. Rep. N 38958. [310] LV adopted all of EL's evidence dealing with the EL-LV combination. See LV Summary of the Testimony (Dec. 31, 1978) at 2-3, 25; LV Opening Brief at 8. [311] Beyond this there would have been greater ease of transfer for EL employees. Although the surplussed LV employees were based primarily in Sayre, Pa., and Buffalo, N.Y., there would have been some problem with respect to the Buffalo employees since no former LV traffic would move over EL's Buffalo branch. Shannon appears to have conceded in his deposition that LV's Buffalo employees would require special treatment, Shannon (D) (Jan. 15, 1981) at 707; when he was pressed on the issue, it became apparent that he had made no detailed study of LV's operations in the Sayre-Buffalo area, id. at 704-07, 720-22. [312] The Chessie Study did not include the rehabilitation and maintenance costs for EL's passenger lines. See Shannon (D) (July 10, 1979) Ex. 36 at 1, EL App. B at 201. [313] LV points out that on this theory the correct figure should have been $413,000 rather than $403,000. [314] Indicates earnings values for the EL-LV combination, not EL alone. The two last figures adjust the earnings values of the Western and Chessie EL-LV acquisitions to reflect inclusion of EL and LV's JT freight lines in the acquired properties. Sources: First four figures from 3 GP at H 12. The last two figures are calculated from the EL's July 10, 1981, Statement, Part II, Sp. Ct. Rep. N 36153 using EL's witness' valuation methodology. [315] The figures for the EL-LV combinations do not include the salvage value of $43.4 million for EL's freight lines and of $34.9 million for LV's lines in the JT area and of $39.6 million for LV's non-terminal lines. Source: 3 GP at H 12. [316] Numbers in parentheses indicate deficits. [317] Numbers in parentheses indicate need for new hires in excess of surplussed employees pool. [318] Source for number of employees receiving payments: Table 7. [319] Rehire rates were taken into account in determining the number of employees assumed surplussed each year. See p. 1276, supra. Consequently, we make no allowance for rehires here, to avoid double-counting with Tables 5 & 6. We assume, as we think an acquirer would have, that the excess available positions in year 1, see Table 6, would not be available for surplussed employees in later years, because of timing difficulties and efficiency requirements. [320] 314 rehires, as reflected in Table 6. All rehires assumed to come from earliest group possible. Thus, 314 - 63 - 118 - 55 = 78, is greater than 0, as reflected in year 5. [321] Additional 75 positions open to reflect effect of pre-Day-1 attrition. [322] Rehires, from Tables 24 & 25, deducted from earliest possible group. No rehires deducted from earlier years to avoid double-counting with Table 25. [323] Benefits available for only five years. [324] Only WJPA benefits available after expiration of four year period following acquisition. [325] Calculations reflect 12 months attrition to account for six months of pre-Day 1 attrition, assumed available to reduce Day 1 dismissals. [326] Dismissal allowances limited to five years, see Table 41. [327] 79 rehires in periods 10/80 and 10/81, from Table 36, assumed to come from earliest groups dismissed that are not receiving WJPA allowances. [328] Source: Tables 36 & 40. [329] From GPs' Sept. 14, 1981, Response, Table C, Sp. Ct. Rep. at N 36885 and GPs' Sept. 21, 1981, Response to EL at 1 n.1, Sp. Ct. Rep. at N 37071. [330] Refers to originating or terminating traffic. There is some traffic that originates and terminates and terminates on EL's passenger lines. Any overstatement that results from this double-counting is insignificant since EL's intrastate freight business "was by any standard negligible". ELRB at 275 & n.273. [331] From GPs' Sept. 14, 1981 Response, Table C, Sp. Ct. Rep. at N 36885. [332] These figures are derived by multiplying the revenue figures for these lines, GPs' Sept. 21, 1981, Response at n.1, Sp. Ct. Rep. at N 37071 by .9369, which is the ratio of Western acquisition revenues, GPs' September 14, 1981, Response, Table C, Sp. Ct. Rep. at N 36885, to historical revenues, GPs' Sept. 14, 1981, Response, Table B, Sp. Ct. Rep. at N 36884, for all but the Hoboken to Port Morris line, see GPs' Sept. 14, 1981, Response at 13-14, Sp. Ct. Rep. at N 36868-69. [333] If D exceeds X for any segment, the price for that segment is simply zero.