The bulk of Soo Line's estimated revenue loss is attributed to diversion of cars from Chicago-Twin Cities to Noyes and Portal. The largest portion of the gross revenue loss estimated by Soo Line was comprised of overhead movements ($1 million). Soo Line states that only $281,196 in revenue would be diverted on overhead traffic interchanged with CP. The largest single amount of overhead traffic losses ($483,707) was on movements between the Twin Ports and Chicago. Soo Line states that most of this diverted traffic originates or terminates on CN or its short-line connections, and is interchanged with Soo Line by the DW&P (a CN subsidiary) at the Twin Ports. Soo Line's connections at Chicago for the traffic are preponderantly those which serve "southern and southeastern points."

Soo Line claims that this traffic represents a market for which both Soo Line and BN vigorously compete. Applicants respond that every movement between the Twin Ports and the Twin Cities, which was considered divertible by Soo Line could have been handled in 1976 by BN, not only between the named points, but beyond Chicago to and from gateways such as Kansas City, St. Louis, Centralia, and Woodlawn. In short, BN already had the capability in 1976 to handle the same traffic further than Soo Line. Applicants contend that since Soo Line was in the routing rather than BN, factors other than longer, single-line service were controlling, and that the merger would not change those factors.

Some $302,227 of the Soo Line overhead traffic revenue considered divertible by Soo Line originated or terminated on CP. Applicants contend that it is unrealistic to expect Soo Line's parent to short-haul or eliminate Soo Line on this traffic. Soo Line replies that the diversion would take place because shippers, not CP, would divert the traffic and that similar traffic today is routed via BN instead of Soo Line. Further, Soo Line points out that the $302,227 diversion represents less than 2.5 percent of the CPSoo Line traffic which is competitive with a CP-BN routing.

On balance, we find applicants' arguments well taken. Although Soo Line might lose a very small part of this traffic, the traffic is actually as susceptible today to diversion by BN as it would be after the merger. The merger in and of itself would not cause this traffic to be diverted. Further, if the merged company does in fact provide an improved level of service for this traffic, it would be incumbent on Soo Line and its Chicago connections to improve their service also.

Similar arguments were made by applicants and Soo Line with respect to the latter's estimate of a $454,241 diversion of joint CN-DW&P-Soo Line traffic moving south of the Twin Ports gateway, where it was interchanged between DW&P and Soo Line. Our findings are the same: If the merged company were to provide improved service, Soo Line and its connections would be encouraged to improve their services. We recognize that such actions would require initiative and innovations by Soo Line and its connections. That, however, is free enterprise.

We have consistently made findings similar to those above with respect to Soo Line's overhead traffic in prior merger and control cases.'

In summary, we find that as to traffic handled jointly between Soo Line and applicants, the latter's estimate of $243,000, higher than Soo Line's own estimate, is a reasonable minimum of expected diversion. As to the nonjoint traffic, we find that Soo Line's estimate of slightly over $1 million should be reduced at least by the approximately $750,000 attributable to the Canadian traffic discussed above. Thus, the diversion would be no more than $490,000, and therefore insignificant.

'See Chicago & N. W. Ry. Co.-Merger-Chicago G. W. Ry. Co., 333 I.C.C. 236, 249 (1968); and Chicago & N. W. Ry. Co.-Control. 347 1.C.C. 556 (1974).

Finally, it should be noted that even if we were to give full weight to Soo Line's estimate of $1.3 million in diversion, the amount represents less than 0.7 percent of Soo Line's 1976 railway operating revenue and does not constitute an amount which would adversely affect Soo Line's ability to provide adequate and essential rail service to the public.

To measure the impact of the western gateway condition, Soo Line conducted a traffic study based on the traffic interchanged between itself and BN at the Twin Cities and the Twin Ports in 1976. The frame of traffic studied consisted of 16,315 cars interchanged at the Twin Cities and 466 cars interchanged at the Twin Ports. Interline forwarded traffic and interline received traffic were stratified to separately identify cars destined (1) to BN stations in Minnesota and (2) all other traffic. Overhead traffic was stratified by direction and on-coming or off-going gateways.

The total of 16,781 carloads were placed in 10 strata. A total of 205 carloads was selected to comprise the random sample. Evaluation of the sample cars was based on criteria and underlying data substantially the same as employed in Soo Line's traffic diversion study. However, the traffic evaluator obtained complete waybill or waybill abstract data only on sample movements for which he had already determined that a gain would accrue to Soo Line. Of the 205 sample carloads, 57 carloads were determined to be gained (Tr. 3929-3930). Similar methods were also employed in estimating the likelihood of a movement being gained, i.e., percentages were applied to each movement judged as a gain.

Revenue estimates were calculated on the basis of division formulae set forth in existing Western Trunk Line division sheets for transcontinental traffic, using mileage prorates west of the Twin Cities gateways, subject to a 25-percent minimum.

All sample carloads were interchanged at the Twin Cities. Soo Line estimates that imposition of its western gateway condition would result in a gain of 1,370 carloads, of which 1,009 would be rerouted via Minot, 194 via Detroit Lakes, 69 via Erskine, 90 via Thief River Falls, and 8 via Bismarck. About 83 percent of the traffic considered by Soo Line to be gained is comprised of interline forwarded movement.

ICG TRAFFIC DIVERSION STUDY

Applicants estimate that ICG would lose $6,592,500 in revenue to the merged company. Applicants' estimates, as noted earlier, are based only on their own traffic. ICG conducted its own traffic diversion study, designed to ascertain the amount of traffic which would be diverted by the merged company; and to estimate the gains in traffic which ICG could expect if its trackage rights condition were imposed. To accomplish this, a sample of ICG movements were drawn and evaluated under each assumption.

The data base is all waybill information and settlement data, as well as car movement data and the estimated cost of transportation. The data utilized from this system consists of 1,662,252 carloads of traffic settled in 1976, which closely corresponds to the 1,624,836 carloads actually handled in 1976.

In developing the sample the 1,662,255 carloads were first sequentially grouped into 1,000 clusters, each containing approximately 1,660 records. Fifty of these clusters were randomly selected for possible use in the sample. The first 15 of the 50 randomly chosen clusters were used as ICG's sample. The 15 clusters, or subsamples, contained 54,938 records and were deemed a sufficient sample for further analysis. From each of the 15 subsamples, the following traffic was eliminated as being nonrelevant, first by computer and then, as necessary, manually:

a. pulpwood and woodchips;

b. sand, gravel and stone;

c. intrastate traffic, except Missouri intrastate traffic; and

d. movements having both origin and destination within the following areas: Connecticut, Delaware, Illinois, Indiana, Iowa, Maine, Manitoba, Massachusetts, Michigan, Minnesota, Nebraska, New Brunswick, New Hampshire, New Jersey, New York, Newfoundland, North Dakota, Nova Scotia, Ohio, Ontario, Pennsylvania, Prince Edward Island, Quebec, Rhode Island, South Dakota, Vermont, and Wisconsin.

Elimination of the foregoing movements left 3,021 movements relevant to the traffic study covering possible losses to a merged BN-Frisco. Written instructions for evaluation of the relevant movements were prepared, and initial evaluations were made by field traffic personnel. These initial evaluations were then reviewed by area sales managers, and the final evaluation was made by the assistant vice president-sales. ICG evaluated each movement to determine the degree, if any, to which diversion by the merged company was possible. If a movement was judged divertible, the evaluator was instructed to assign the percentage of the movement (in increments of 5 percent up to and including 100 percent) that was likely to move over a new route. For percentages of less than 100 percent, the criteria leading to the judgment was specified by the evaluator.

The criteria for determining divertibility are substantially the same as those used by applicants and other carrier parties in these proceedings and need not be repeated. ICG estimates that its gross revenue loss to the merged company would be $8,672,901. (Actually, ICG points out, the loss would lie between $7,773,627 and $9,572,174. This range was determined through calculation of the standard error at the 95 percent confidence level, i.e., the probability of the actual loss being within the described range is 95 out of 100.)

ICG studied both traffic handled jointly with applicants and traffic not handled jointly with BN and Frisco. According to applicants, $1,085,540 out of the total estimate represents ICG's estimate of losses on nonjoint traffic, with the balance ($7,587,361) attributable to joint traffic. ICG does not dispute applicants' breakdown of the joint and nonjoint traffic estimates.

Applicants argue that none of the ICG estimates are valid on nonjoint traffic, and that ICG's estimates on joint traffic are overstated by nearly $1 million. Applicants contend that the total ICG estimate therefore is overstated by about $2 million. In support of these contentions, applicants specifically criticize ICG's diversion evaluation on all nonjoint traffic sample cars and on several joint traffic sample cars. ICG's response is that whether the Commission accepts applicants' estimate or that of ICG (or an amount in between), the financial impact on ICG would be too great to be absorbed.

JOINT TRAFFIC

ICG origin. With respect to joint traffic, applicants dispute ICG's diversion judgments on 34 sample movements. Fourteen of these movements originated at points served exclusively by ICG. This Commission has consistently recognized that an originating carrier serving consignors exclusively has the greatest influence on the routing of the traffic compared to all other carriers participating in the movements.

'If unwritten instructions were issued for a particular movement, they were recorded by the evaluator.

ICG would not lose the 14 movements, because, as applicants point out, the shippers must rely on ICG for car supply, switching, rate adjustment, and car tracing and expediting services.

ICG common origins.-Another five ICG sample disputed movements were originated by ICG at points served in common with other railroads. Two of these movements originated at Memphis and terminated on BN at Chariton, IA. Applicants' argument that ICG would continue to hold the traffic, as the originator and supplier of the cars, ignores the fact that the merged company would be able to provide singleline service for the entire movement. The merged company would be in an extremely good position to solicit this business away from ICG at the origin. The remaining ⚫ three movements originated on ICG in Louisiana at points served in common with railroads other than applicants. ICG claims that the merged company would be able to short-haul ICG by requiring interchange at Memphis rather than at Centralia (the existing interchange point). Two of the movements terminated on BN, while the third represented a bridge movement over BN with termination on UP. In each of these situations, we conclude that ICG, as the origin carrier, will retain this traffic, especially since the origin points are not served in common with applicants.

We conclude that ICG's estimated revenue loss should be reduced by $460,187 for the traffic represented by the 17 sample movements found nondivertible in the above discussion.

ICG special equipment.-Applicants also dispute ICG estimates of diversion on four sample movements, each of which utilized ICG special equipment. Two of the cars originated on ICG, and the other two moved as overhead traffic on ICG. On the two originated movements, the ICG evaluator claimed that applicants would short haul his railroad. This is extremely unlikely since ICG served the consignor and provided special equipment. On the two overhead movements, it is clear that ICG was inserted in the rouging due to the use of its special equipment. The ICG revenue represented by these four movements is $68,780, and the total diversion estimate made by ICG will be reduced by this amount.

Transit time.-"Faster transit time" was used as a reason for diversion on three sample movements (all ICG bridge movements) disputed by applicants. In each instance, as applicants point out, either time was not a factor, faster routes were already available, or routes involving fewer carriers were available. However, on one movement, transit privileges applied and were effected on BN's portion of the movement. BN was the origin carrier, and the destination was on Santa Fe, which was the owner of the specially equipped boxcar utilized. Thus, ICG was justified in projecting a loss on this movement. As to the other two movements, applicants' argument is well taken. Further, one car originated on ICG at a station served also by L&A, the movement involved privately owned or leased special equipment, and was routed ICG-Chicago-Soo Line-Minneapolis-BN. ICG could have been eliminated from the routing without the merger. The other car also was special equipment, and the routing involved ICG as a bridge carrier over its route between Birmingham and Memphis via Fulton, KY, and the bridge movement by Mopac between Memphis and Kansas City. As applicants point out, the movement could have gone from its Florida origin all the way to Woodlawn on Family Lines, then, BN to destination. The expanded revenue on the latter two cars discussed here totals $63,346, and ICG's estimate of diversion will be reduced further by this amount.

Miscellaneous losses.-Applicants also dispute ICG estimates of losses taken on eight "miscellaneous" kinds of movements. For three of these, applicants contend that the route predicted by ICG was not applicable under existing tariffs. Nevertheless,

there were other routes to which the movements could be diverted. The record does not reveal whether those other routes would change ICG's estimated revenue loss on the three movements, but giving ICG the benefit of the doubt, we conclude that ICG's diversion estimates on the three movements are reasonably accurate.

On two of the miscellaneous sample movements, applicants point out that the ICG evaluator did not know the identity of the consignor and the consignee. The two sample cars moved as bridge traffic over ICG. BN was the originating carrier for one car, and the terminating carrier for the other. We have consistently recognized overhead traffic as being the most susceptible to diversion. However, the absence of consignor and consignee identities, makes an informed diversion judgment nearly impossible. We therefore must find that ICG's diversion estimate should be reduced by $21,795, the expanded gross revenue represented by the two sample cars.

Of the three remaining sample cars in the miscellaneous group, we agree with applicants' criticism of ICG's evaluations. Two of the movements were of lumber, and both of these were reconsigned in transit after leaving BN. The third car would be retained by ICG because transit privileges applied on ICG's own lines. The expanded revenue represented by the three sample movements total $112,882.

Special privileges.-Finally, three disputed movements involve transit and stopoff privileges. The privilege was exercised on the ICG, and the proposed merger would not affect the movement. The transit movement was an outbound move wholly on ICG from Chicago to Memphis. The inbound move to Chicago had originated on BN nearly a year earlier. The outbound movement (i.e., the sample movement here) was switched to the ICG by a terminal carrier in Chicago. Contrary to ICG's evaluation that transit time would be reduced by a merged BN-Frisco, the switching movements in Chicago could not be eliminated or materially reduced, and the movement would not be affected by the merger. The revenue on these three sample movements on an expanded basis is $82,744, and ICG's estimate should be reduced by that amount.

NONJOINT TRAFFIC

Shipper allocation.-Applicants dispute all of ICG's estimates of diversion on nonjoint traffic. Specifically these estimates are based on 31 sample movements. Seven of these movements involved the same shipper and moved from a point in Michigan to Memphis. The shipper has a traffic allocation policy and transit privileges apply at Memphis. Applicants concede that if Frisco had been included in the routing as the terminating carrier the movements would indeed be divertible, but in these seven movements, ICG would retain its allocated share of the business. We agree with applicants.

Transit.-Two movements originating in Indiana moved over ICG to Kansas City, where Katy received them for termination on its lines. Transit privileges were exercised on ICG. Therefore, we find the two movements not divertible.

Other carriers' equipment.-Four sample movements involved shipments of lumber and plywood. In each instance, the origin point was served by BN or a BN subsidiary in common with the originating carrier. The originating carriers (UP, SP, MILW, and Soo Line) each supplied the equipment, resulting in the elimination of BN or its subsidiary as the originating carrier. In each of the four movements ICG was either a bridge carrier or the terminating carrier. The provision of equipment for lumber shippers is of great importance, and we find that the four sample movements would not be affected by the proposed merger.