This research paper was commissioned by the Canada Transportation Act Review. It contains the findings and opinions of the author(s) and does not necessarily represent the views of the Review Panel or its members.

North American Railway Restructuring and Implications for Merger Policy

Research conducted for the Canada Transportation Act Review

Report prepared by
R.L. Banks & Associates, Inc.

February 2001

North American Railway Restructuring
and Implications for Merger Policy

A Paper Prepared for the

Canada Transportation Act Review Rail Symposium

R.L. Banks & Associates, Inc.

Transportation Economists and Engineers
1717 K St. N.W.
Washington, D.C. 20036
Winnipeg, February 27, 2001

North American Railway Restructuring
and Implications for Merger Policy

A Paper Prepared for the
Canada Transportation Act Review Rail Symposium
by R.L. Banks & Associates, Inc.
Winnipeg, February 27, 2001

Table of Contents

Executive Summary

Introduction

Industry Restructuring in Recent Years

Reasons for Rail Consolidation

Crossborder Issues

Future Restructuring: Duopolies and the Public Interest

Canadian and U.S. Regulatory Frameworks

Issues in Rail Merger Evaluation

Summary of Conclusions and Recommendations

Appendix: U.S. Railway Restructuring as Reflected in the

U.S. Department of Justice's Market Concentration Methodology

Executive Summary

The Canada Transportation Act Review Panel's (CTAR) charge requires it to examine, among other things, the extent to which the current legislative and regulatory framework affecting transportation is appropriate to the public policy requirements that may arise from emerging industry structures. This discussion paper has been commissioned in support of that assignment. It reviews the dimensions, causes and consequences of the past two decades' industrial restructuring of North American railways, and identifies principal policy concerns respecting the government's future role in the process.

The principal conclusion is that further coalescence of the North American railway industry should be permitted to proceed only under the most stringent guarantees that competition will not be eroded. New arrangements may well produce opportunities to tap new markets and exploit changes in trade flows resulting from policies promoting integration of regional economies. However, any reduction in the number of Class I carriers must be accompanied by effective compensatory mechanisms to ensure that railways remain the servants of regionalization, and not the masters.

Restructuring of the North American railway industry over the past two decades has resulted in its becoming substantially smaller, more concentrated and more efficient. These trends antedate enactment of deregulatory legislation, but they have been accelerated by the Staggers Act in the U.S. and NTA 87 and CTA in Canada.

In Canada, the principal means of restructuring has been through line abandonment and sales to short line and regional carriers. The two major Canadian carriers have shed thousands of miles of underperforming lines, in the process spawning dozens of short lines, and aggressively expanded reoriented traffic flows. To the extent that plant rationalisation and labour force reduction has occurred and attendant economies have been achieved in the U.S., a similar process - and not mergers - has been the prime catalyst.

The efficacy of mergers in producing public benefits has also been sharply questioned following the flawed execution of three of the past five major North American railway mergers.

Deregulatory legislation and market concentration has led to lower overall rail rates, but also to increased opportunity for, and application of, price discrimination to the detriment of a large coterie of captive shippers.

As a consequence of these developments, severe shipper backlash has developed in both Canada and the United States; threatening both the deregulatory process and the continuing ability of railways to reconfigure networks in a manner most conducive to responding to the changes in tradeflows and changes in shippers' logistical needs.

Negative outcomes of the restructuring process have compelled the strongly pro-railway U.S. regulatory agency, the Surface Transportation Board, to 1) examine competitive access proposals in a formal proceeding, 2) impose a 15-month moratorium on major railway mergers, and 3) re-examine and propose significant added restrictiveness to its merger evaluation guidelines. Absent such revisions to merger rules, generally expected outcomes include industry reregulation and creation of a continent-wide duopoly.

In an era where mergers promise few benefits from improved operational efficiency, the extent of benefits accruing to carriers from foreclosure of competitive services must be subject to close inspection. Reinstitution of dual merger oversight authority in Canada would properly raise the bar for prospective Class I railway merger applicants. Given the gravity of the decisions to be made - likely determining the structure of the North American rail network for decades to come - the cost of added circumspection is certainly justified.

Governmental oversight of the restructuring process will require the application of two different skill sets: competitive analysis, resident in the Competition Bureau, and assessment and ongoing monitoring of the mechanisms deployed to protect competition, requiring industry-specific skills resident in the Canada Transportation Agency (CTA).

Further, as multinational rail consolidations may be conditioned based on the divergent and quite possibly conflicting views of Canadian and U.S. authorities, the institution of a Joint Canada - U.S. Commission, with the authority to resolve inconsistencies in merger decisions, should be invoked when necessary.

Canada should modify its merger rules in a manner which is neither obstructionist nor liable to codify an industry structure which turns out to be detrimental to long-run national interests. This requires formulation of rules which, in general, are directed towards promoting incrementalism and reversibility in restructuring initiatives.

Introduction

The Canada Transportation Act Review Panel's (CTAR) charge requires it to examine the extent to which the current legislative and regulatory framework affecting transportation is appropriate to the public policy requirements that may arise from emerging industry structures. This discussion paper has been commissioned in support of that assignment. It reviews the dimensions, causes and consequences of the past two decades' industrial restructuring of North American railways, and identifies principal policy concerns respecting the government's future role in the process. The single most salient and obvious observation is that further coalescence of the North American railway industry should be permitted to proceed only under the most stringent guarantees that competition will not be eroded. New arrangements may well produce opportunities to tap new markets and exploit changes in trade flows resulting from policies promoting integration of regional economies. However, any reduction in the number of Class I carriers must be accompanied by effective compensatory mechanisms to ensure that railways remain the servants of regionalization, and not the masters. Combining railways should be free to propose the design of such compensatory mechanisms, but their ultimate acceptance should only follow rigorous examination under standards that give due consideration to the public interest.

Bifurcated Merger Oversight and its Demise

In 1993, the National Transportation Act Review Commission (NTARC) recommended that oversight of railway mergers be entrusted solely to the Competition Bureau, and the National Transportation Agency's (NTA) authority be terminated.1 Citing "[the] unnecessary overlap...between the Competition Act and National Transportation Act, 1987 (NTA 87) provisions." NTARC came down in favour of "the better developed provisions of the Competition Act" as against the "too sweeping and unnecessarily costly" scheme incorporated in NTA 87.2 Observing that industry-specific rules tend to be sculpted in light of transitory concerns, NTARC concluded that:

In the long term... Competition Act provisions [such] as... review of mergers may become more appropriate for ensuring competition in commercial transportation arrangements not yet foreseen. 3

And laid down the challenge:

[T]hose advocating specialised procedures to encourage competition in any one mode should have the responsibility to demonstrate, first, why the Competition Act procedures are insufficient for this purpose and, second, why any proposed transport-specific rule would work better.4

Underlying NTARC's recommendation, which subsequently was incorporated in the Canada Transportation Act (CTA 96)5 was a strong concern as to the long-term viability of the Canadian rail network. From 1986 to 1991, Canadian Railways' tonnage declined two percent, and traffic composition continued to shift from higher to lower-valued goods. The industry's operating ratio deteriorated from .90 to .96 between 1986 and 1991, reflecting a five percent decline in operating revenues accompanied by a one percent increase in operating expenses. In constant dollar terms, the respective changes were negative 21 percent and negative 15 percent. In 1991 both CP Rail System (CP) and CN experienced losses - $6.9 million for the former and $34.3 million for the Crown Corporation.6

Improvements in the financial viability of Canadian railways clearly depended then, as now, on the reduction of the costs of operation. The principal means to accomplish this has been through the rationalisation of plant, a process strongly encouraged by the abandonment provisions of NTA 87 and CTA 96. The dual-oversight regime instituted with the enactment of NTA 87 threatened to impede the ability of mergers to contribute to the rationalisation process, and was therefore an apt target for change. In particular, the CN and CP railway network in Eastern Canada was the subject of reconfiguration proposals in the early 1990s.7 (Notably, one result clearly not intended by NTARC as a result of the termination of NTA merger authority was consolidation of the entire CN and CP systems.)8 Any Eastern Canada solution could be enfeebled by an NTA whose sole statutory guidance was to disallow consolidations "against the public interest." As in any delegation of authority to an administrative agency, the lack of express standards increases the vulnerability of the agency's decision making process to political influence.

The removal of NTA's merger oversight authority, one of several reductions in regulatory authority incorporated in CTA 96, also represented a victory for the viewpoint, shared by the Competition Bureau, that regulatory agencies are regularly co-opted by the industry that they regulate.9 This phenomenon is well established. Even those agencies, which are expressly created to protect consumers, tend to drift in perspective; this is natural, since it is representatives of the regulated industries, not their customers, who interact with agency staff on a regular basis. However, NTA's authority was essentially a negative one - it was empowered only to veto a merger approved or approvable by the normally merger-averse Competition Bureau. The apparent concern, therefore, was that NTA could, through denial or forced inclusion, act to preserve uneconomic incumbents threatened by an efficiency-enhancing merger.

Addressing NTARC's Challenge: Are Industry-Specific Rules Required?

Recent history has put a spotlight squarely on NTARC's challenge: is the Competition Act an adequate vehicle to preserve or encourage competition in the railway industry, or must additional, industry-specific mechanisms be provided? Have things changed so much - has the railway industry's structural realignment taken it to the point where fundamentally different rules are now appropriate? What are the concerns with maintaining the current system of oversight?

In December 1999, a proposed consolidation of CN and BNSF sparked a reconsideration of merger guidelines applicable to transactions in the United States. The guidelines proposed by the STB involve an extensive expansion of filing requirements and an unprecedented depth of inquiry, all with the intended effect of slowing or halting industry restructuring.10 Virtually all of the proposed rules are highly industry-specific: there are few analogs with the criteria, generally applicable to the rest of the economy, that the U.S. Department of Justice and the Federal Trade Commission apply in their examination of mergers. Further, the rapid integration of regional economies and railways' responses has led the STB to an unprecedented examination of cross-border issues, including the possibility of co-ordinating merger oversight amongst affected nations.

Are industry specific oversight procedures appropriate for Canada? No change is required, suggests CN, which with BNSF could make a strong argument that the STB had targeted them unfairly when it imposed a merger moratorium and proposed new rules. CN has submitted to the CTAR that:

CN sees no compelling reason why such mergers or acquisitions should be treated any differently than those in other industry sectors. Railroad transactions have been subject to general competition laws for many decades. Parliament has, in the past, assigned a role in merger review to relevant Ministers of the Crown only in instances where bona fide public interest concerns have been established. Canadian National would not object to an amendment to the Competition Act to give the Minister of Transportation standing to make representation to the Commissioner of the Competition Bureau where public opposition to a proposed rail sector transaction exists. Canadian National would not object to a further amendment to the Competition Act that would require the Commissioner to take into consideration the views of the Minister.11

Whatever the other merits of CN's position, its concession that consideration of the Minister of Transportation's views might be required by Statute cannot be viewed as substantive. CN is certainly aware of the utter inefficacy of similar requirements in the U.S. where agencies disagree.12

Shipper and Provincial Concerns

Canadian shippers have been quick to point to the CN - BNSF episode as reason to re-examine the current regulatory scheme. The Canadian Shippers' Summit complained that:

It is ironic that the representations of shippers with respect to this proposed combination [CN-BNSF] were directed to the STB in the United States because the Agency had no jurisdiction to deal with the matter...

And recommended:

[T]he provisions contained in Part VII of the NTA, 1987 should be re-instated and expanded as necessary, to include combinations of the type proposed by the CN/BNSF transaction and any other transactions which result in a merger or consolidation howsoever structured.13

A number of other provincial and shipper parties have indicated that changes need be made; British Columbia Ministry of Transportation and Highways' (MoTH) recommended dual CTA/Competition Bureau oversight14 as did the Canadian Chemical Producers' Association (CCPA).15 Saskatchewan advised, in addition to "a new section in the revised CTA that would invoke a review of rail competition issues" responsive to proposed Class I railway mergers, the "[c]reation of a Joint Commission between the U.S. Surface Transportation Board and the CTA Agency and/or Competition Bureau to discuss procedural issues" - in the event such mergers involved a U.S. railroad.16

Clearly, these non-railroad parties have sensed "a bona fide public interest concern," one which would justify CTA or Ministerial participation in prospective Class I railway mergers. To some, highly indeterminate, extent the desire to rein in merger activity reflects shipper frustration with poor rail service and with differential pricing, both of which may be only tenuously associated with actual rail consolidation. Why tenuous? Because these problems antedate the recent series of mergers; service quality has been the bane of rail shippers for over a century - the very first case before the ICC, in 1887, involved equipment supply - and the problem has been studied and restudied by railways themselves with limited, if any progress being evident.17 Further, the severe service disruptions associated with the Union Pacific-Southern Pacific (UP-SP) and Conrail transactions, which injected new energy into shippers' movements, are in all probability temporary problems associated with merger execution, and not manifestations of diseconomies of scale inherent in such large systems.

But the association in shippers' minds of poor service with monopoly power is natural. In a congressionally-ordered study of rail service quality in the period 1990 through 1996, the U.S. General Accounting Office (GAO), an independent research arm of the U.S. Congress, reported that railways insisted that service levels were "adequate" in the face of profound shipper dissatisfaction.18 Normally, it is the customers' viewpoint that is determinative of whether service is "adequate." Normally, if a customer finds that a service provided to it is inadequate, it will find another service provider. It is the utter frustration of being unable to engage competitive rail service that spurs demands for competitive access, to provide another carrier with the opportunity to break a local rail monopoly. But other than for a combination CN and CP, there is little risk that mergers will create more Canadian captive shippers.19

The Surface Transportation Board Reaches an Epiphany

Judging the extent of the STB's reversal of policy can be seen in light of the restructuring activity that it and its predecessor agency endorsed in the last five years. The final major act of the Interstate Commerce Commission, before it was "sunset" and its responsibilities transferred to the STB, was to approve the merger of the Burlington Northern Railroad and The Atchison Topeka & Santa Fe Railway Company, which became effective in October of 1995. The STB, which came into being on December 29, 1995, was faced with the already-filed Union Pacific - Southern Pacific merger application. This was unanimously approved in August 1996 and became effective the following month. Virtually all opposition to the merger was swept aside; to many, the STB appeared to perform more as an advocate than an impartial arbiter. NS and CSX's take-over of Conrail was approved in July 1998, and was implemented on June 1, 1999. CN's merger with IC, approved in May 1999, was implemented in stages almost immediately thereafter.

Between CN and BN's merger announcement in December 1999 and the following March, the STB came to the realisation that:

Our current merger regulations were adopted soon after the passage of the Staggers Act of 1980.... The merger regulations - aimed at encouraging railroads to formulate proposals that would help rationalise excess capacity so long as competition, access to essential service, and other public interest goals were not degraded - were a proper and reasoned response to the serious problems affecting railroads and their customers at that time...

The goals of the merger policy have largely been achieved. It does not appear that there are significant public interest benefits to be realised from further downsizing or rationalising of rail route systems, as there is little of that activity left to do. Looking forward, the key problem faced by railroads - how to improve profitability through enhancing the service provided to their customers - is linked to adding to insufficient infrastructure, not to eliminating excess capacity.20

Industry Restructuring in Recent Years

The North American railway industry has undergone a significant restructuring over the past 20 years, although it has been with a distinctly different cast in each of the three countries signatory to the North American Free Trade Agreement (NAFTA). Elements of these changes are a continuation of long, historic trends, although in some cases, those trends were accelerated since deregulation of railroads in the United States in 1980, and changes in Canadian transport legislation in 1987.

Canada

The two major Canadian carriers have shed thousands of miles of underperforming lines, in the process spawning dozens of short lines, and aggressively reorienting traffic flows to take advantage of increased hemispheric trade.

Among the prime movers of this restructuring were a number of legislative enactments and ratification of international treaties. NTA 87 provided for the abandonment of up to four percent of CN and CP's route systems annually, setting in motion the rationalisation of excess plant. The 1989 Canada U.S. Free Trade Agreement (CUSTA), despite its specific exemption of rail services, and the 1994 enactment of NAFTA, have dramatically lowered barriers to the cross-border flow of capital, goods and services.21 Truly catalytic roles were played by the February 1995 budget which eliminated railway freight transportation subsidies, the 1995 privatisation of Canadian National and the 1996 CTA, which eliminated quotas on branch line divestitures and promoted the growth of a short line industry.

As reported by Statistics Canada:

Canadian railways increased tonnage handled by 21.1% between 1993 and 1998... The strategic role that railways once played in the development of the Canadian economy has dwindled over time as other modes of transport have developed and governments have withdrawn or reduced their support for uneconomical railway lines.... Cash strapped railways, also faced with the problem of excess capacity, have endeavoured to rationalise their lines. The National Transportation Act, 1987 provided for rationalisation but the process was slow because of certain restrictions on abandonment and/or transfer of lines. The new Canada Transportation Act, enacted in July 1996, has facilitated the process of abandonment and acquisition of surplus lines by short line railways... The length of track operated has declined steadily during the period under observation. The total including owned, leased and shared track has decreased 13.3% from 85 000 km in 1993 to 73 000 km in 1998. CN and CP together operated 20.3% less track while trackage operated by short-haul carriers increased 32.7%. This increase reflects the growth of short line railways in Canada. 22

The growth of the short line industry, which came belatedly to Canada, provides the opportunity to take advantage of two very different models of railroad economics. Unencumbered by restrictive work rules and with lower salary structures, short lines can operate with much greater flexibility than their Class I counterparts, and provide service to on-line customers far beyond the capability or typical practices of larger carriers. Class Is, on the other hand, have been freed to concentrate on high-volume, long haul business, relying heavily on short lines for feeder service. Approximately 35 short lines have been formed since the enactment of CTA 96, more than triple the number formed in the period NTA 87 was in effect. They operate approximately 20 percent of Canadian trackage and terminate or originate a comparable percentage of Canadian rail tonnage.

United States

In response to the collapse of Penn Central and many other railroads in the northeast, Congress passed the Staggers Act in 1980 to restore the financial health of the rail industry. It left the ICC in charge of deciding on merger applications. The Staggers Act also set out less stringent standards for unification involving smaller railroads. In 1981, the ICC interpreted public interest to mean "balancing the potential benefits to applicants and the public against the potential harms to the public." In 1982, the ICC stopped imposing conditions that protect traffic for railroads adversely affected by the merger. It discouraged applications for parallel mergers that reduced shipper alternatives and frowned upon financial transactions that did not integrate management and operations.

The restructuring of the railroad industry has included the following major components:

Other elements of the restructuring process include the acceleration of cross-border ownership and common operations and an increase in shared facilities as conditions of mergers and acquisitions.

The restructuring of railroads has created something of a paradox. On one hand, the railroad structure is characterised by a few, highly efficient, "super" or "mega" railroads, with a significant concentration of power. On the other hand, track ownership, rolling-stock investment, and operations, has been disbursed among hundreds of new local and regional railroads, shippers and consignees, transportation brokers, and other car suppliers - sometimes with cross-border implications. The result is that over the past 20 years, more and more business entities, throughout Canada and the U.S., have come to depend on a smaller number of railroads. In essence, the list of railroad customers has been expanded to not only include those that ship and receive products, but also those entities which invest in and/or operate a segment of what was previously the railroad business - including several hundred local and regional railroads, the majority of which interchange with but a single Class I railway.

Historic railroad development was aided by significant public support, including rights of eminent domain, massive land grants, and outright financial assistance. Spurred by regional competitiveness, by the early 1900s the industry ballooned into an over-built, inefficient, and labyrinthine network of somewhat uncoordinated operations, which reached a peak road-mileage in June 1916. The U.S. federal government's experience in nationalising the railroad system during World War I led to the enactment of the Transportation Act of 1920, which emphasised economic efficiency and the need for an integrated system of railroad operations. This legislation gave legitimacy to what sound economics demanded - the rationalisation of the industry.

The downsizing of the railroad industry (Class I railroads) between the early peak years and 1999 has been substantial. As competition from motor carriage increased, so did the pressure on railroads to merge, abandon light-density lines, and in some cases, declare bankruptcy. As is shown in Table 1, while the pre-1980 downsizing was steady and substantial, the post-1980 downsizing was accelerated. Also, the pre-1980 downsizing did not radically alter the distribution of railroad services. As discussed below, the shifting of railroad services to non-railroad entities was to accelerate in the post-1980 period.

Table 1

Consolidation of U.S. Class I Railroads

Source: Association of American Railroads

While a multitude of railroad functions have shifted to other non-traditional parties - such as increased loading and unloading by shippers, marketing by transportation brokers, and services performed by water carriers - the major shift in railroad functions centres on two changing events: 1) the huge increase in intermodal traffic involving motor carriers, and 2) the shifting of freight-car ownership from Class I railroads to other parties. Intermodal traffic (trailers and containers) has grown from three million units in 1980 to nine million units in 1999. Much of this traffic has been marketed by trucking companies and transportation brokers. Furthermore, it eliminates railroad loading and unloading, pickup and delivery, and a number of administrative functions.

The second major shift in providing railroad service is illustrated by the diminishing percent of the freight-car fleet owned by Class I railroads. As shown in Table 2, Class I railroads have reduced their relative car ownership from 68 percent of the total car fleet in 1980 to 42 percent in 1999. While some of this change is attributable to the downsizing of Class I railroads and the increase in the number of local and regional railroads, the trend is also explained by the growth in car ownership on the part of car leasing and owning companies and shippers. For example, the car companies and shippers owned 441 thousand cars in 1980; that figure increased to 663 thousand cars in 1999. The additional 222 thousand cars is some ten times greater than the increase in the fleets of the local and regional railroads between 1980 and 1999.

Table 2

One of the most noted changes in the structure of the railroad industry since 1980 has been the increased concentration of business, and power, among a relatively few railroads. The concentration of traffic among the four largest railroads had held at a relatively steady rate between 1929 and 1969 (about 25%), and then increased gradually to 43% in 1980. But in the post-1980 period, the concentration percentage more than doubled. As shown in Table 3, in 1999 the four largest railroads handled 95% of the traffic. This substantial increase was due to the acquisition of large railroads by other large railroads - most recently and notably: 1) the Burlington Northern's purchase of the Santa Fe in 1995, 2) the Union Pacific's acquisition of the Southern Pacific in 1996, and 3) the CSX and Norfolk Southern take-over of Conrail in 1999. Other significant railroad mergers took place in the early 1980s.

Table 3

U.S. Class I Railroad Concentration, 1929 - 1999

Source: Association of American Railroads

While railroad track continued its long-term trend of abandonment over the past two decades, a significant portion of what might have otherwise been abandoned was sold to local and regional railroad operators. As shown in Table 4, since 1980 there have been in the U.S. 433 new local and regional railroads created, accounting for 32,810 miles of owned road. While some of these railroads have since terminated operations, in 1999 there was a total of 550 new and pre-existing local and regional railroads, accounting for 26 percent of railroad miles of road owned and 29 percent operated. Yet these smaller carriers comprised only 18 percent of the industry's carloads originated (traffic) and just nine percent of total freight revenue.

Table 4

Source: Association of American Railroads

Mexico

As in the U.S. (Conrail) and Canada (CN), Mexican rail restructuring has involved the consolidation of multiple failing lines into a single government-supported entity followed by the privatisation of a reinvigorated system. Ferrocariles Nacionales de Mexico (FNM or Ferronales), which opened its first trunk line in 1873 (from Veracruz to Mexico City), absorbed all four other Mexican common carrier rail systems over the period 1982 to 1987.

In the late 1980s, controlling FNM's large and mounting deficits, at times exceeding US $600 million per year, had been a major concern to the government; it became clear that any hope of Ferronales becoming self-sustaining must lie in major changes which would grant the railway the freedom to operate on a more commercial basis. Among the constraints faced were restrictions in railway rate policy - any tariff revisions had to be ratified by the Government, severely limiting the railway's ability to compete with trucks. On the cost side, labour was a major impediment; with some 80,000 employees, Ferronales achieved output per employee of only ten percent of what is achieved in the U.S.; it was severely restricted in its ability to terminate employees, with the only possible grounds being desertion, alcohol or drug consumption, fighting and lack of decency.

The privatisation of FNM was accomplished in several steps. A series of "convenios" - agreements between the railway and the government - were concluded beginning in 1985, aiming toward FNM's financial self-sufficiency by 1988 and then progressively later dates. However, as FNM's operating ratio ballooned from 95 in 1984 to 129 in 1988, other measures clearly became necessary. Contracting out of support services, in particular maintenance of equipment, was initiated in 1994; concessions were quickly let in three of FNM's five operating divisions.

With ratification of NAFTA, the Mexican rail system became attractive to foreign investors. Traffic immediately began building on the FNM system, and, beginning in 1996, the Mexican government began the process of concessioning the right to operate over portions of the FNM infrastructure. Major U.S. railroads have purchased minority stakes in the concessioned operations that must, by law, remain under the control of Mexican nationals. Transportación Ferroviaria Mexicana (TFM), affiliated with KCS, now operates the 4,200 kilometres of the Northeast Railway, which reaches Mexico City from Laredo, Texas, and co-operates with, even if it is not formally allied with, the CN-IC-KCS alliance. (For example, CN personnel located in Mexico make customer-specific joint sales with KCS and TFM.) Ferromex's 6,500 kilometre Pacific-North system reaches multiple border crossings, and has been active in expanding international intermodal capabilities. Ferrosur, the third and final independent carrier, began operations at the end of 1998, and serves the Mexican heartland.

Reasons for Rail Consolidation

In any industry, railways included, the strategic purposes in seeking mergers is to obtain some combination of intercorporate synergies and enhanced market power. The governmental interest lies in ensuring that the public and private benefits from synergies, particularly those that are manifested in cost reductions and more effective development and deployment of innovations, outweigh the public disbenefits of reduced competition.

Rationalisation of Plant

Consolidation of railways has been promoted by government policy, off and on, since the early days of the 20th century. The U.S. Transportation Act of 1920 affirmatively sanctioned mergers as a means to reduce system overcapacity; this position was questioned but never fully reversed until the STB concluded in March 2000 that merger policy established by the Congress in the 1970s had become counterproductive and was no longer in the public interest.23

The public justification for mergers has usually been offered in terms of achieving various forms of economies. Merger applicants have invariably claimed that consolidation will result in a lower cost structure through the abandonment of duplicative or excess facilities and improved utilisation of those retained. Operating efficiencies resulting from reduced circuity, combined terminal operations, paring and pairing of track, etc. have been among the claimed "public benefits" mergers are expected to produce in order to obtain regulatory approval. The STB's proposed rules describe potential benefits as follows:

Proposed § 1180.1(c)(1): Potential benefits. By eliminating transaction cost barriers between firms, increasing the productivity of investment, and enabling carriers to lower costs through economies of scale, scope, and density, mergers can generate important public benefits such as improved service, enhanced competition, and greater economic efficiency. A merger can strengthen a carrier's finances and operations.24

Historically, these benefits have been either small or illusory.25 It is unquestioned that a declining cost structure, with some significant step functions as capacity is reached, exists with respect to line density. However, full advantage of economies of density is often obtainable through steps short of merger.26 Untrammelled pursuit of economies of density, the spreading of fixed infrastructure costs over more traffic, can lead to two, opposing negative consequences. First, as made evident by the UP-SP "meltdown," determination of infrastructure capacity is not an exact science, nor is estimation of future demand. When a mismatch of capacity and demand does occur, resultant congestion can quickly spread havoc throughout a network. Second, the benefits of a strategy of rationalising low-density lines are self-limiting; the elimination of feeder services contributes to the reduction of density on main lines. This realisation led to a shift in railway and government policy away from abandonments and in favour of spinning off lines to short line carriers wherever possible. But treatment of short lines subsequent to their formation has become of serious concern. The ability to finance short lines in the future will turn on their ability to maintain some independence from connecting Class I carriers; the purchase prices of "captive" branch lines may be expected to directly reflect service and rate guarantees provided by the connecting Class I carrier.

Under any circumstance, as the opportunities to rationalise plant are primarily associated with parallel mergers, and the current structure of railways in Canada and the U.S. can be characterised as a triple duopoly - CN and CP in Canada, BNSF and UP in the Western U.S., and CSX and NS in the East - no more major mergers are available that can lead to large scale plant rationalisation without also generating extensive rail monopolies.

Economies of Scale

Even less certain than the potential of mergers to generate economies of density is the ability of mergers to produce economies with respect to rail organisation scale. The performance of KCS and IC at the top of the U.S. Class I railway industry for the past several years brings into sharp focus the question of the existence of such economies. Kent Healy in 1961 found distinct diseconomies of scale in systems with over 10,000 employees.27 This finding has been attacked, with Frank Wilner making the claim that "experience has found these assertions to be grossly incorrect"28 - a statement written just as the UP-SP and Conrail service disasters were about to commence. In fact, Healy's observations - read carefully - were and are amply supportable. Healy did not reject economies of density - the relationship between line-specific traffic volume and operational efficiency is unchallenged - but he did question the tendency of large organisations to create distance between management and customers and management and forces on the ground. Bureaucratisation causes inefficiencies, shipper disaffection and other ailments, which were in strong evidence, particularly with the UP-SP merger. The relatively successful CN-IC merger, as CN has been at pains to point out, has benefited in part as a result of a well-planned strategy of contemporaneous centralisation and decentralisation of functions; a conscious attempt has been made to keep most marketing functions closer to customers and functions which could profit by economies of scale - such as purchasing and administration - have been consolidated.29 CN-IC has also confronted a far less daunting challenge in implementation phasing; it was faced neither with absorbing a large and largely decrepit system as was UP nor with the need to integrate systems on the first day of merged operations - the unprecedented task faced by NS and CSX.

Legitimate reasons for seeking mergers include repositioning carriers to accommodate changes in trade flows, a motivation evident in both the consummated and failed CN merger proposals, and to extend opportunities for single line service. The North American rail industry consists of several major overlapping and interconnected networks, producing fierce competition along with strong requirements for co-operation between carriers to provide geographic scope to service areas. Motor carrier networks, on the other hand, are not interdependent, and trucks have access to all possible origins and destinations through the courtesy of government-provided infrastructure. Thus, motor carriers have significant operational (and information flow) advantages that railroads cannot replicate beyond their own geographically constrained infrastructure. Movements requiring a high level of service quality and extended hauls are lost by railroads to trucks despite rail's cost structure which is more advantageous with increasing length of haul. Mergers increase the scope of railroads' ability to perform single-line service. The vision claimed by railroads is to harness the benefits of a fully-integrated continental network and circumvent the "artificial" constraint of limited geographic scope. At issue in any future merger proceeding is the extent which arrangements short of merger can attain the same goals as consolidation.30

Reactive Mergers, Cramdowns, and Weak Sisters

Other reasons for mergers are often more powerful and more suspect. Mergers have induced counter-mergers, as railway officials fear being out-manoeuvred by competitors; whether this is a real threat or not is difficult to ascertain. The consequences to UP of not acquiring SP following the BNSF merger may have been less severe than the consequences of not acquiring it. But there is no empirical means of proving either case. NS may have felt compelled to enter into a bidding war with CSX for Conrail, but it may have achieved defensive aims with less cost by negotiating trackage rights deals with CSX, sparing both carriers billions of unnecessary expense. But the long-term history of rail mergers again and again evidences the overriding role of defensive, reactive mergers in the process of rail consolidation.

Mergers also produce two important ends for railway management. One, railways have been able, in the U.S., to use mergers as a means of abrogating collective bargaining agreements, under the so-called "cramdown" authority. It is alleged by labour that:

[F]ollowing past transactions, carriers have been pressured by the financial markets to produce immediate savings, and to react to plunges in their stock prices, by cutting costs; that carriers have responded to this pressure by laying off employees.31

Notably, the U.S. Department of Transportation concurred with this assessment in urging the end of "cramdown" authority. The role of financial markets also enters into the merger equation another way: the railway industry's share of GDP has been declining for decades, but growth is essential to courting investor favour. The only consistently effective way of "growing" a Class I railway is through acquisition; internally generated growth is extremely rare and of limited consequence. Tables 5 and 6 on the following page illustrate the decline in numbers of Class I railways in the East and West U.S. respectively. Over time, as the four-firm concentration ratio grows, eventually to 100 percent in both territories, note that operating revenue of the four largest firms, measured in 1999 US constant dollars, changes very little. Nor are the new "mega-carriers" outsize in comparison with their forebears. Conrail's 1980 operating revenue exceeded that of any eastern carrier in 1999; in the West, it was not until 1995 that a carrier's revenues were greater, in real terms, than was BN's in 1980. (SP and Santa Fe were the second and third largest western carriers in terms of operating revenue in 1980; UP was number four.)

Finally, one reason occasionally proffered for merging railways is to improve service and financial strength of a weak carrier by joining it with a strong one. Pre-Staggers experience with "weak sister" mergers had been sufficiently poor as to impel changes in merger rules that severely restricted regulatory authority to force inclusion of competitively harmed railways. Nevertheless, UP effectively resurrected the "weak sister" rationale it in its bid to acquire SP. Of the 1,300 statements of support from shippers for its proposed merger, a sizeable majority cited problems with SP as a reason to join the railroads.32

Table 5

Eastern Region Class I Railroads: Operating Revenue In Descending Order

Constant 1999 $US (000s)

Table 6

Western Region Class I Railroads: Operating Revenue In Descending Order

Constant 1999 $US (000s)

Performance of the U.S. Class I Rail industry in the Process of Restructuring

Significant improvements in railway productivity have occurred in the post-Staggers era, a fact well-advertised by Class I railways.33 Not only the railways themselves, but the STB, despite its putative role as regulator, has repeatedly served as an advocate of the railways in praising the success of the deregulatory and restructuring programme, and implicitly, its own role in it.34 Linking the improvement in cost per unit of output (generally, revenue ton-mile) to restructuring and deregulation is to a large extent spurious, as rail productivity increases have been in the same order of magnitude as other basic industries.35 Further, the constant-dollar rail revenue per ton-mile decline (45.3 percent from 1984 to 1999, according to the STB's rate report) was only slightly superior to real cost declines in pre-Staggers periods. The corresponding decline from 1932 to 1947 was 40 percent; from 1953 to 1969, the decline was 36 percent. This phenomenon can be traced back to the early days of railroading. Nominal revenue per ton-mile dropped by more than half - from 1.88 cents to 0.73 cents from 1870 to 1900.36 This trend preceded the passage of the Interstate Commerce Act (1887), that regulated railroads - and continued after it. From 1890 to 1900 - after enactment - revenue per ton-mile declined 23 percent in nominal terms - versus the eight percent nominal drop in rates as measured by the STB in the corresponding period (three to 13 years) following enactment of the deregulatory Staggers Act.

Following the passage of the Staggers Act, profit margins have fairly consistently been higher than in earlier periods. Railroads have, in some easily measured respects, become more efficient It cannot be ignored that the improved financial health of the industry has come concomitantly with a drastic downsizing in its relative economic importance. Operating revenues have declined 37.1 percent in real terms from 1980 to 1998. In the same period, real U.S. GDP increased 63.6 percent. The share of GDP represented by rail revenues had, in 18 years, declined to 38 percent of the 1980 level.

The AAR's annual Analysis of Class I Railroads indicates that between 1980 and 1998 shareholders' equity had declined 8.0 percent in real terms. Total assets to the latter year (1998) increased 6.0 percent - but were down 17.1 percent between 1980 and 1994, before inflated merger acquisition costs helped inflate reported assets. If not for merger purchase premiums, there would be no increase in net investment in the Class I industry over the past 20 years.

Crossborder Issues

In the period 1992 to 1998, Canada's trade with the U.S. grew at a rate of 13.5 percent per year versus 4.0 percent for domestic trade and 4.6 percent for other international trade.37 In that period, the share of exports destined for the U.S. grew to $270 billion - 85 percent of the nation's total exports. While recessions and financial crises in Asia and Latin America partially explain the relatively weak growth of non-U.S. trade, there can be little question that NAFTA has invigorated cross-border movements and will continue to do so.

Canadian railways responses to increased Canada - U.S. trade include:

The CN-BNSF merger, if consummated, would have resulted in a single carrier that would have directly and through strategic alliances blanketed much of the North American continent. Other possible mergers could do the same. Without here passing judgement on whether such mergers serve valid commercial interests of railways and shippers, the creation of a true transcontinental carrier in the U.S. has long been a dream of railway managers, and presumably attempts to realise this vision, with or without Canadian participation, will arise in the future.

Several salient cross-border issues were raised in the STB review of merger procedures. Given the different commercial and regulatory regimes existing in Canada and the U.S., a merger approved by one or the other countries may confer national advantages; important commercial decisions involving the merged railroad could conceivably be based upon national, rather than economic, considerations. In Ex Parte 582, The U.S. Department of Transportation (DOT) expressed concern that there might be an effort to influence routing of rail traffic to/from foreign ports at the expense of U.S. ports. The U.S. Department of Agriculture believed it to be essential that mergers be conditioned so that shippers in both countries will be assured of equal access to rail transportation and that the influence of state trading enterprises, particularly as respects the distribution of railcar capacity among U.S. and Canadian agricultural shippers, be monitored closely.

DOT was also concerned with the impact of foreign law in the area of corporate structure and management. Canadian law appears to tie corporate control to nationality or citizenship, raising questions of reciprocity. (CP and CN have both countered with the fact that they are majority owned by U.S. shareholders.) Similarly, Canadian shippers' have expressed concern over transnational mergers and routing rights. CRSC explained to the STB that Canadian shippers' right to choose their own routings would be endangered by crossborder mergers.

CRSC indicates, by way of illustration, that, under Canadian law, a Canadian shipper has the right to specify, in a Bill of Lading, that its traffic is to be carried by the originating carrier to an interchange point, and thence via a connecting carrier to destination. CRSC warns that a U.S./Canadian rail merger (CRSC cites, as an example, a BNSF/CN merger) would effectively eliminate the ability of shippers to seek competitive rates in this manner at points where the merger partners connect. The effect, CRSC adds, would be a loss of existing rail competition at those points for traffic destined to U.S. markets.38

Resolution of these issues is likely to require the creation of a Joint Canada - U.S. Commission which would be responsible for synchronising the terms of approval of a transborder merger as between those ordered in Canada and the U.S.

Future Restructuring: Duopolies and the Public Interest

A concern of several observers and interested parties in response to the CN-BNSF merger proposal was that it would trigger the final round of North American major railroad restructuring, inevitably culminating in an ocean to ocean, tropics to tundra duopoly.39 This is a possible but hardly preordained outcome.40 With approval of the next major merger proposal, continental duopoly could become a self-fulfilling prophecy. Prospects for a transcontinental duopoly are heightened by the tendency of railways to try to prevent competitors from substantively exceeding them in size. Concerns of non-participating carriers include the loss of friendly connections, inferior market coverage and reduced leverage in negotiating rate divisions. Failure of remaining carriers to forestall a merger is therefore frequently followed by applications for responsive mergers. The path of least resistance in such a process leads to continental duopoly.

Assuming that a duopoly does eventuate, it is not entirely clear as to what the consequences would be.41 Economists have never conclusively established under what circumstances duopolies compete nor when they don't or won't. The U.S. cellular telephone duopolies established in the 1980s raised interesting questions as to whether identical prices were indicative of pure competition or pure collusion. Canadian railroading is considered to be a classic example of aggressive competition and market-sharing coexisting as between the same two corporations.42 The extent of intramodal competition would in any circumstance be a function of pre-existing point-to-point competition,43 negotiated access agreements and imposed conditions attendant to successive consolidations. Given the wide range of competitive potential that differently structured duopolies may provide, it is entirely appropriate for a public interest investigation to be instituted to review major merger proposals. The investigatory body should be invested with the authority to impose conditions upon a merger with the purpose of maximising public benefits: that is, preserving potential service enhancements and innovations to the maximum extent possible while minimising removals and reductions of competition. Where trackage rights can provide adequate substitute competitive access to those shippers which otherwise would lose rail service alternatives, it should preferably be arranged through negotiation between included and non-included carriers. If questions of practicability of negotiated arrangements or the possibility of tacit collusion remain, the regulator's authority should allow it to modify and monitor the agreements, and perhaps impose competitive access requirements of its own. Where excess circuity or other operational factors make competitive access of limited usefulness, the regulatory authority should extend to imposition and enforcement of long-term service and rate guarantees that can be structured on most-favoured nation or pre-existing service bases.

Canadian and U.S. Regulatory Frameworks

Canada Competition Bureau

Mergers in transportation are subject to civil law as provided by sections 91-107 of the Competition Act. Merger Enforcement Guidelines prepared by the Competition Bureau of Industry Canada provide the public with details of how the Bureau would evaluate a proposed merger or consolidation.

The Commissioner has the authority to approve a merger or allow it to proceed subject to monitoring and further conditioning. In cases where the Commissioner does not approve a proposal, the Bureau submits an application to the Competition Tribunal challenging a merger. The Competition Tribunal based on evidence submitted by the Competition Bureau and other parties makes legal evaluation of a proposed merger as to whether it is likely to lessen or prevent competition. Evidence of concentration or market share may be evaluated in the light of other factors including health of the business, foreign competition, innovation and barriers to entry. The law provides an exemption for transactions where the efficiency gains exceed and offset likely anti-competitive effects (§96).

Regulation of Rail Mergers Prior to May 1996

The Canada Transportation Act proclaimed in July 1996 eliminated review of transportation acquisitions by the National Transportation Agency by repealing previous authority and restored the pre-eminence of the Competition Act. That agency had been granted prevailing authority over all other agencies by the National Transportation Act, 1987. The newly created National Transportation Agency was given overriding authority over the Director by §29 which indicates that where there is a conflict between the 1987 Act and any other Act in respect of a particular mode of transportation, "the rule, order or regulation made under this Act prevails."

Earlier, the National Transportation Act of 1966-67 required in Article 27(2) that the predecessor authority, Canadian Transportation Commission, give notice of any proposed acquisition involving a transportation company to the Director of the Competition Bureau. The Director submitted evidence and expert testimony for consideration by the Canadian Transportation Commission.

Part VII of the 1987 Act also addressed acquisitions of Canadian transportation undertakings. Jurisdiction was limited to transactions involving assets or annual revenues in excess of $10,000,000. The agency was directed (sections 256 and 257) to investigate and make a finding as to any objection that the proposed transaction is not in the public interest. Section 4 of the Act defines public interest as being what is consistent with the national transportation policy or with directions from the Governor in Council issued before notice of the proposed transaction (§23).

The 1987 Act did not specifically require notification of the Director by the agency.

Regulation of Railroad Mergers in the U.S.

Railroad mergers involving assets in the United States are subject to the approval of the STB. Federal regulations governing procedures for STB review of proposed transactions are published in the U.S. Code of Federal Regulations (CFR) Title 49 part 1180 entitled "Railroad Acquisition, Control, Merger, Consolidation Project, Trackage Rights, and Lease Procedures"44 General acquisition procedures are presented in Part A and consist of ten sections identified as 1180.0 through 1180.9. Procedures for transfer or operation of lines of railroads in reorganisation are presented in Part B consisting of one section (1180.20).

The announcement of the proposed combination of BNSF and CN in December 1999 as well as lingering concerns about rail service problems prompted the STB to initiate a rulemaking procedure to review and revise current regulations because the agency concluded that those regulations:

"are not adequate to address future major rail merger proposals that, if approved, would likely result in the creation of two North American transcontinental railroads".

By decision served October 3, 2000, the STB published for public comment new merger rules expected to be effective in the spring of 2001. The following discusses selected aspects of what the STB has proposed in Ex Parte No. 582 (Sub-No. 1): Major Rail Consolidation Procedures.

General Policy

At the heart of the new rules is what the agency describes as "a paradigm shift in our review of major mergers." More to the point, the STB initiated on its own authority a major reversal of two decades of a national transportation policy that sought to revitalise the financial condition of the rail industry through a process known as rationalisation that emphasised elimination of excess line capacity. The agency (and its predecessor) consistently had accepted railroad claims that this approach also improved service by expanding market reach of railroads and providing single system service to more shippers.

The STB declared that the rationalisation "process has now largely been completed" and that public benefits "to be realised from further downsizing of rail route systems are limited." The STB has proposed that it will be the stated policy that consolidations are no longer presumed to be in the public interest. The burden of proof has been shifted to merger applicants.

The STB's goal in evaluating proposed mergers will be to "ensure balanced and sustainable competition in the railroad industry." Thus, the standard of intermodal competition will be replaced with a standard of intramodal competition. Curiously, the STB views railroad networks in the context of "a broader transportation infrastructure". The "competitiveness of the infrastructure" is now offered as a standard to replace competing modes of transportation and competing carriers.

With regard to acknowledged anticompetitive effects, under the current policy "the Board does not favour consolidations that substantially reduce the transportation alternatives available to shippers unless there are substantial and demonstrable benefits that cannot be achieved in a less anticompetitive fashion." Under the proposed policy, no matter the nation's benefits, the STB would "not favour consolidations that reduce ... alternatives available to shippers unless there are ... benefits ... that cannot otherwise be achieved. In other words, anticompetitive effects no longer are in fashion. To the contrary, enhanced competitiveness is now an acknowledged public benefit, if not a requirement. Consistent with the shift in emphasis to competitive impacts, the new policy is silent on how pricing freedom relates to merger policy.

The current general policy statement concludes with an admonition that the analysis of competitive impacts must be sensitive to the fact that the U.S. Congress has "mandated ... greater freedom to price without regulatory interference." This indication that less weight could be given anticompetitive effects has been deleted from the proposed policy.

Consolidation Criteria

The STB acknowledges that merger evaluations are governed by federal statute, in addition to the regulations the agency publishes in the Federal Register, which establishes five criteria that the STB must consider at a minimum. However, in a marked departure from current regulations, the STB has proposed to replace the recitation of the statutory minimums with a list of six specific goals to be considered. Effective competition is listed as the first goal indicating that the STB has decided to upgrade its importance and give it more weight. It is interesting to note that whereas the statute requires the STB to consider "the effect of competition among rail carriers" a new standard has been set requiring a transaction to result in "effective competition."

Public Interest Considerations

STB evaluations of mergers have always been based on a balancing test weighing many complex factors. What is noteworthy about the proposed new regulations is that the STB would codify three significant findings about railroad consolidations; namely that:

First, "Other private sector initiatives, such as joint marketing agreements and interline partnerships, can produce many of the efficiencies of merger while risking less potential harm to the public";

Second, Additional consolidation is "likely to result in a number of anticompetitive effects, such as loss of geographic competition, that are increasingly difficult to remedy"; and

Third "Merger applications must include a provisions for enhanced competition." Otherwise, the regulations warn, it "will likely cause the STB to make broad use of powers ... to condition its approval to preserve and enhance competition."

Clearly, significant burdens of proof have been shifted to the shoulders of applicants in merger and consolidation proceedings and a substantial new requirement has been established.

In addition, the STB indicates it would give substantial weight to competitive gains that can be realised immediately, discount claims of future benefits using a current value approach, and perhaps even disregard benefits that are realisable through alternative business arrangements.

Merger Conditions

In explaining its policy on exercising its authority to impose conditions on consolidations, the STB reiterates its rationale for requiring applicants to include in the proposed transactions specific agreements that will result in increases in competition. The STB states that any merger "will create some anticompetitive effects that are difficult to mitigate with appropriate conditions." Consequently, the proposed policy is that applicants will voluntarily identify and implement access transactions to enhance competition to "mitigate or offset all types of threatened merger harms to the public interest."

Cumulative Impacts and Crossover Effects

Just as the STB's proposed general policy would reverse federal policy toward the railroad industry, so too is the STB proposing to reverse the fundamental nature of the merger planning and evaluation processes. Current policy limits issues addressed in merger proceedings to known events that are directly attributable to a proposed transaction and are "reasonably certain to occur." In order to "curb speculation", "the impact of potential or hypothetical transactions" currently are excluded from consideration and deferred to a later proceeding as necessary.

The STB proposes to require merger applicants to calculate "the likely public benefits that their merger will generate ... in light of the anticipated downstream mergers." "The Board expects applicants to anticipate with as much certainty as possible what additional ... merger applications are likely to be filed in response to their own application and explain how these applications, taken together, could affect the eventual industry structure and the public interest."

Policy has switched from prohibiting speculation to requiring it. What once was the purview of academics, think tanks and public planning agencies such as the U.S. Department of Transportation has been made the responsibility of railways seeking to execute business transactions.

Transnational Issues

The STB has proposed a new policy that would require railroads to submit "full system" analyses incorporating information about marketing plans, customers, employees and operations in Canada and Mexico.

Proposals such as the one advanced by CN and BNSF also would involve additional burdens on applicants to fathom the hypothetical. The STB proposes that, among other issues, in the case of "foreign control" of a railroad, "applicants must assess the likelihood that commercial decisions ... could be based on (foreign) national or provincial rather than ... economic considerations." Further, the risk of foreign control has prompted the STB to propose codifying another new policy that rail mergers "not detract from the ability of the United States military to rely on rail transportation to meet the nation's defence needs." Applicants would have the burden of assessing the impact of such a proposed merger on the defence of the United States.

Other Issues

The STB has proposed strengthening its policy with respect to employee protection and uses language intended to pressure railroads to bring ongoing collective bargaining negotiations concerning implementing mergers to conclusion before it finalises the proposed merger rules. Specifically, under the revised policy the STB would "look with extreme disfavour on overrides of collective bargaining agreements."

The proposed merger rules also codify current practice with regard to the STB's oversight process during implementation of a merger. Additional planning and monitoring requirements addressing post-merger service have been proposed. In particular, major new analytic inputs will be required under the rubric of a Service Assurance Plan. Current service levels must be documented and their adequacy demonstrated. Expected service improvements must be identified in terms of precise steps. The plan will included provision for problem resolution procedures and an ad hoc group to be known as a Service Council that will bring railroads, shippers and interested parties together. Specific contingency plans are to be included in the application "to address the negative impacts if projected service levels do not materialise in a timely fashion." ... "Those plans, based upon available resources and traffic flow and density, must identify potential areas of disruption and the risk of occurrence." Open-ended planning accounting for the universe of uncertainty would be mandated by federal regulation.

Issues in Rail Merger Evaluation

Fundamental questions raised by the rail industry restructuring process include 1) would the restoration of some degree of regulatory power over rail mergers by the CTA be necessary to serve the public interest and, if so, 2) should such regulatory authority be limited to simple approval or disapproval of mergers as proposed, or should it encompass the right to modify specific terms and conditions of proposed transactions? Also, are there compelling reasons to promote harmonisation of U.S. and Canadian regulatory principles and remedies, and to what extent would such parallelism be beneficial?

Limitations of Competition Bureau Methodologies as Applied to Railways

In Canada, where CTA's jurisdiction over mergers was terminated in 1996 in favour of the Competition Bureau, very serious issues may be raised as to the appropriateness of applying standards developed for non-network industries to networks. As the Commissioner of Competition has noted in its submission to the CTAR Panel, "competition law cannot address typical problems associated with a natural monopolist, such as: high prices, insufficient supply, inadequate service or type of services, high or low profitability, absence of entry into the industry and insufficient investment, etc."45 These and other issues associated with rail merger proceedings have traditionally been addressed in the domain of economic regulation, by agencies with industry-specific expertise. It may well be that appropriate railroad industry oversight requires that greater weight be given to the principle of transparency and less to the principle of confidentiality than can rightfully be accorded by an agency with economy-wide responsibilities, such as the Competition Bureau. As an example, the Competition Bureau's policy towards intervenors is insufficiently broad to allow a full airing of issues which may well impact on the public interest.46

Little concern would be voiced if the situation had persisted wherein there was small possibility of a major Canadian railway merger, but it has been several years since this has been the case. All manner of permutations of joint CN/CP ownership of eastern lines were discussed as realistic possibilities in the years immediately preceding the termination of CTA authority, and since 1996, as noted in the discussion of forces driving consolidation, mergers with U.S. carriers have become a strategic focus of both major Canadian railways.

The Competition Bureau's area of special competence - determination of where corporate practices act to restrain or preclude competition - would not appear to extend to the ability to evaluate, in the context of a major rail proceeding, the virtues of each the hundreds of ameliorative measures likely to be argued for by the possible hundreds of intervenors. Weighing the whole sack of requests for conditions, determining what combination allows for both viability of the merger and viability of competition requires a very specialised set of technical competences, which the Competition Bureau cannot be expected to maintain in-house. As noted above, the U.S. experience in assigning plenary authority to rail mergers to one agency, with the stipulation that a second agency's viewpoint be accorded substantial weight, has failed abysmally in practice. The STB's unrestrained approval of merger applications over the objections of the Attorney General is a proximate cause of the reevaluation of merger proceedings now underway in the U.S.

Railway Mergers and the Public Interest

Forthcoming major merger proposals will involve matters of great public interest. The structure of the rail network will have grave implications for development; it will strongly effect the economic viability of industry now in place and locational decisions of industry in the future. Strong arguments have been made that regional development is not a proper concern of Federal policy-making, that the market is the appropriate mechanism to determine the efficient allocation of resources, and "second guessing" by bureaucrats is more likely to lead to economic stagnation than economic vibrancy. However, arguments made by the railways themselves strongly call into question whether the railways' perceived self-interests are in line either with the national interest or with their own, long-term survival, and specifically raise the issue whether unconditioned or weakly-conditioned mergers promote or hinder efficient resource allocation.

A prime example of the danger of inefficient resource allocation lies in railways' assertion that market or product competition adequately restrains their market power. CN, in its submissions to CTAR, claims that 18 percent of all railway traffic in Canada is subject to such competition, including 94 percent of coal traffic (the remaining six percent is subject to intramodal competition) and virtually all other export traffic.47 The concept of market or product competition was, until recently, employed in determining whether a railway in the U.S. exercised "market dominance" over a shipper, and thus potentially subject to maximum rate regulation. The existence of such "competition" was also commonly accepted by the Interstate Commerce Commission as evidence of attenuated market power exerted by railways proposing consolidation. But such application of the concept as put forward, with some success, by railways, badly distorts the intended role of product and market competition as defined and as applied in merger evaluations in both Canada and the U.S.

Part 3 of the Competition Bureau's guidelines discusses market and product competition in terms of defining the relevant market for analysis of competitive effects. Specifically:

The assessment of whether a significant and nontransitory price increase would likely be made unprofitable involves an examination of likely responses from sources of product and geographic competition, on both the demand and supply sides of the market. On the demand side, it is necessary to evaluate the extent to which:

(i) buyers would likely switch to substitute products; and,

(ii) buyers would likely switch to the same product sold in other areas. On the supply side, it is necessary to evaluate the extent to which:

(iii) new entry would likely occur through the construction of facilities, (or as a result of sellers of other products adapting existing facilities, to commence production) of the product or a substitute; and,

(iv) sellers of the product or of a substitute who are located in distant areas would likely divert their product into the area in question.48

Similarly, the U.S. Horizontal Merger Guidelines, employed by both the DOJ and the Federal Trade Commission, expounds:

Absent price discrimination, a relevant market is described by a product or group of products and a geographic area. In determining whether a hypothetical monopolist would be in a position to exercise market power, it is necessary to evaluate the likely demand responses of consumers to a price increase. A price increase could be made unprofitable by consumers either switching to other products or switching to the same product produced by firms at other locations. The nature and magnitude of these two types of demand responses respectively determine the scope of the product market and the geographic market.49

What is striking about both Canadian and U.S. applications of source and product competition is that they apply directly to the products produced by the merger applicants. They do not apply to factor inputs. In the context of a railway merger, the question presented is "can a shipper obtain transportation from another source," for it is transportation that the railways are selling. The fact that a monopolist railway will only price up to, and not beyond the level that will put a shipper out of business is not something that the guidelines intend to condone. The railway's argument could as easily be applied to any factor input. World competition for grain constrains delivery price of grain. Therefore, a fertiliser monopolist will constrain its prices to that which will keep farmers in business, if only marginally. A monopolist steel producer would set its prices applicable to North American automobile manufacturers based upon prices of imported cars. Taking the analogy one step further, railways would have no reason to object to the monopolisation of the railway supply industry. A single North American producer of rail, ties and ballast, being aware of the thin profit margins of its customers, and the fact that they face truck competition in some markets, would take care only not to price above that which would endanger railway solvency.

Source and product competition as argued by the railways means no more and no less than that the railways are entitled to act as equity partners with their shippers, extracting all possible rents above that needed to keep the businesses afloat. Railways argue that given their self-interest in not pricing in a manner lethal to the golden goose, there is no need to constrain pricing that may merely be toxic. The resource misallocation that derives from this attitude is obvious: who wishes to invest in any otherwise economically sound venture when all profit is susceptible to capture by a service supplier? What of the farmer who actually faces source competition? What incentives does he have to become more efficient than is minimally necessary to face world competition when the value of any breakthrough is subject to immediate capture by the railway? The long-term interest of the railways is implicated in the misuse of source and product competition. Profit-maximising pricing by a monopolist - as are railways constrained only source and product competition faced by their shippers - differs depending on whether a long or a short term perspective is taken. Pressure exerted by financial interests on railways inevitably focuses on short term results; pricing to maximise railway profits in the short run, to satisfy the need for strong quarterly results, can have a devastating long-term effect by discouraging development of rail-dependent industry in the monopolists' service territory. The effects of such strategies, of course, do not usually come into full bloom until well after the current management teams have moved on to other pursuits.

The Competition Bureau is certainly capable of properly applying its own guidelines with respect to product and source competition. However, the railways' arguments are correct to the extent that railway services provided to many shippers subject to these forms of competition are not highly profitable, and cannot be made highly profitable, at least in the short-term. Therefore, the Competition Bureau may be likely to find that merging carriers would be unable to sustain non-transitory price increases to such shippers above the five-percent threshold. Further, the experience with most railway mergers has not been that they have resulted in widespread rate increases, but only selective ones where opportunities for price discrimination have been enhanced. Thus, product and source competition as proffered by the railways may enter the merger evaluation process through the back door. It will require the special expertise of transport analysts to recognise where recent history of market-constrained pricing may not be indicative of the efficacy of those same constraints several years in the future. And it will require expert input of representatives of shipper industries to provide evidence of whether and how resource misallocation may transpire, a form of evidence which appears to be admissible only on a discretionary basis under current Competition Tribunal rules. Under any circumstance, it is worth considering whether a broad, public-interest oriented review of major mergers proposals may be required to "legitimise" consolidations and thereby fend off reregulatory legislation. In the U.S., the STB's stiffening of merger rules, supported by all major U.S. railways other than the BNSF, is generally understood to be a prophylactic measure, blunting the strengthening movement towards increased federal oversight of the industry.

A second area in which the public interest is implicated but which may not be adequately served by Competition Bureau rules alone involves competitive access as it pertains to modification of merger agreements. In the U.S., the STB has generally endeavoured to ensure that individual shippers which had competitive options prior to a merger retain some form of access to at least two carriers subsequent to a merger; trackage rights have become a favoured, if not always fully effective, means of providing substitute access. The STB has generally declined to require access to additional carriers by exclusively served "bottlenecked" shippers whose sole carrier sought to merge with one of several connecting carriers, on the theory that that such shippers could not be competitively harmed. But the reason that the shipper cannot be competitively harmed - procedurally treated as a rebuttable presumption - is that since the shipper is already subject to a monopoly, all monopoly profit - "one lump" - is already extracted or extractable by the directly serving carrier. Without addressing the issue of whether this is true, either in theory or in fact, the railways' embrace of the "one lump theory" raises additional concerns that should be examined in the context of any proposed major merger.

How reasonable, in almost any other circumstance, would it be for a corporation to take the public stance that it is treating its customer as badly as possible, and if the customer believes that they could be treated any more badly than they are, they should have the burden of proof to establish it? The rebuttable presumption applied until now by the STB proceeds from the logic that there is no competitive harm because the railway already has the shipper in a hammerlock. The railroads' impassioned defence of the one-lump theory illustrates not only the willingness to sacrifice the long term for short-term profits, it evidences an adversarial attitude which is bizarre given its direct target being the industry's very own clients. This behaviour is not accommodating of customers' requirements, and surely the railroads understand that abused customers will take the first opportunity to relocate, build out, substitute factor inputs, anything to escape the clutches of the carrier, even if it takes decades. From an economic perspective, the railroad has effectively decided to disinvest in the bottlenecked route. It has determined to extract all the money it can while shippers are captive and then forfeit the business. The proper way for regulatory authorities to analyse any merger which creates bottlenecked customers, closes gateways, etc. is to find that it constitutes, to that extent, a disinvestment in the rail industry. It ultimately portends a reduction of connectivity, a reduction of the effectiveness of the railway system to serve as a network.

True, in any given circumstance, millions of dollars may be at stake in a bottleneck situation. And true, railroads must price differentially because the ratio of average revenue to marginal cost on directly competitive service is far from adequate to cover fixed costs. But, in most efforts to restrict customer choice, the railroad is in fact disinvesting in the property, whether intentionally or not. Railway strategies that accelerate shippers' relocation plans or result in shippers' reducing output below otherwise optimal levels should be recognised for the disbenefit it is, with claimed merger benefits reduced accordingly.

Another matter in which Competition Bureau procedures may not afford a sufficient basis for evaluation of rail mergers is the determination of the relevant market. Identification of the relevant market(s) is critical as it is the power of the merged entity to raise prices within the market that is determinative of anticompetitive effects. Nor, generally, will the Competition Bureau challenge a merger that does not first pass the threshold test that the post-merger market share of the merged entity will exceed 35 percent.

But 35 percent of what? Measured by revenue share, 100 percent of the Canadian class I Railway system will constitute less than 35 percent of the Canadian transportation market. The narrower the definition of relevant markets, the more likely that the threshold for further investigation will be breached. For example, in 1996, the Acting Director of Investigation, in opposing the acquisition by Canadian Pacific Limited of Cast North America defined the market by specific origin-destination pairs.50 But how would the markets be measured in a major railway consolidation? Would it be commodity specific on an origin to destination basis? If not commodity specific, the fact that there is substantial motor carrier or water competition for some products but not others would certainly impact against the threshold tests, as well as the overall ability to raise rates by five percent or more.51 Industry-specific rules are required to arrive at some level of predictability and consistency in railway consolidation cases.52

Summary of Conclusions and Recommendations

In the event of further proposed major rail consolidations, protection of Canadian interests requires a re-examination of the current regulatory framework. In an era where mergers promise few benefits from improved operational efficiency, the extent of benefits accruing to carriers from foreclosure of competitive services must be subject to close inspection. Issues include the acceptability of continued cession of oversight authority to U.S. regulatory agencies and the necessity of long-term oversight of merged carriers.

Reinstitution of dual merger oversight authority, with competitive and public interest issues reviewed by the Competition Bureau and the CTA, respectively, would properly raise the bar for prospective Class I railway merger applicants. Given the gravity of the decisions to be made - likely determining the structure of the North American rail network for decades to come - the costs of added circumspection are certainly justified. A sequential process is desirable; railways should submit proposals to the Competition Bureau which would then evaluate competitive issues sans detailed input by a multitude of interested parties, and could choose to challenge the merger or not before the Competition Tribunal. A positive decision by the Tribunal or failure to challenge by the Bureau would be followed by a second stage proceeding in which the CTA would evaluate public interest issues, and in which the Bureau of Competition would be expected to be an intervenor. CTA would be expected to approve those mergers in which the public interest can be satisfied through non-debilitating conditioning of a merger application. If conditions necessary to preserve the public interest and satisfactory to the merger proponents can not be agreed upon, the merger would have to be denied.

Further, as multinational rail consolidations may be conditioned based on the divergent and quite possibly conflicting views of Canadian and U.S. authorities, the institution of a Joint Canada - U.S. Commission, with the authority to resolve inconsistencies in merger decisions, should be invoked when necessary.53

The driving force behind the STB's current merger procedure re-evaluation proceeding has been the failure to anticipate impacts of recent mergers. This amply justified regulatory exploration has, in some respects, turned into an exercise in "fighting the last war," in which solutions are being crafted which may have rectified the UP-SP merger but may not turn out to be pertinent to continually evolving conditions. Canada should modify its merger rules in a manner which is neither obstructionist nor liable to codify an industry structure which ultimately could be detrimental to long-run national interests. This requires formulation of rules which, in general, are directed towards promoting incrementalism and reversibility in restructuring initiatives.

Appendix: U.S. Railway Restructuring as Reflected in the U.S. Department of Justice's Market Concentration Methodology

Excerpts from U.S. Department of Justice and the Federal Trade Commission; Horizontal Merger Guidelines, Issued April 2, 1992, Revised April 8, 1997: §4.2 Market Shares and Concentration

Market concentration is a function of the number of firms in a market and their respective market shares. As an aid to the interpretation of market data, the [using] Agency will use the Herfindahl-Hirschman Index ("HHI") of market concentration. The HHI is calculated by summing the squares of the individual market shares of all the participants. Unlike the four-firm concentration ratio, the HHI reflects both the distribution of the market shares of the top four firms and the composition of the market outside the top four firms. It also gives proportionately greater weight to the market shares of the larger firms, in accord with their relative importance in competitive interactions.

The Agency divides the spectrum of market concentration as measured by the HHI into three regions that can be broadly characterised as unconcentrated (HHI below 1000), moderately concentrated (HHI between 1000 and 1800), and highly concentrated (HHI above 1800). Although the resulting regions provide a useful framework for merger analysis, the numerical divisions suggest greater precision than is possible with the available economic tools and information. Other things being equal, cases falling just above and just below a threshold present comparable competitive issues.

General Standards

In evaluating horizontal mergers, the Agency will consider both the post-merger market concentration and the increase in concentration resulting from the merger. Market concentration is a useful indicator of the likely potential competitive effect of a merger. The general standards for horizontal mergers are as follows:

a) Post-Merger HHI Below 1000. The Agency regards markets in this region to be unconcentrated. Mergers resulting in unconcentrated markets are unlikely to have adverse competitive effects and ordinarily require no further analysis.

b) Post-Merger HHI Between 1000 and 1800. The Agency regards markets in this region to be moderately concentrated. Mergers producing an increase in the HHI of less than 100 points in moderately concentrated markets post-merger are unlikely to have adverse competitive consequences and ordinarily require no further analysis. Mergers producing an increase in the HHI of more than 100 points in moderately concentrated markets post-merger potentially raise significant competitive concerns...

c) Post-Merger HHI Above 1800. The Agency regards markets in this region to be highly concentrated. Mergers producing an increase in the HHI of less than 50 points, even in highly concentrated markets post-merger, are unlikely to have adverse competitive consequences and ordinarily require no further analysis. Mergers producing an increase in the HHI of more than 50 points in highly concentrated markets post-merger potentially raise significant competitive concerns. Where the post-merger HHI exceeds 1800, it will be presumed that mergers producing an increase in the HHI of more than 100 points are likely to create or enhance market power or facilitate its exercise.

Table A-1

Changes in HHI of Eastern Class I Railroads

1980 - 1999

Source: Association of American Railroads, RLBA analysis

Table A-2

Changes in HHI of Western Class I Railroads

1980 - 1999

Source: Association of American Railroads, RLBA Analysis

1 Under Part VII of NTA 87, the NTA had the right to review the acquisition of any transportation undertaking (with assets or sales over $10 million) engaged in any transportation activity under the legislative authority of the Federal Parliament. The NTA could disallow the proposed acquisition if were found to be against the public interest. The NTA had 120 days after receipt of notice to render a decision. Section 92 of the Competition Act provides that the Competition Tribunal can dissolve a merger or dispose of assets or shares where the Tribunal finds that a merger prevents or substantially lessens competition. The Tribunal can take up to three years to conduct its investigations and come to a decision.

2 NTARC; Competition in Transportation; Policy and Legislation in Review, Vol. II, p. 190. (1993)

3 Ibid. Vol. I, p. 181.

4 Ibid.

5 "Nothing in or done under the authority of this Act affects the operation of the Competition Act." CTA, 96 ¶4(2).

6 Statistics Canada; Rail in Canada, 1991

7 From 1988 to 1992, total rail tonnes originated in the East decreased from 116 to 106 million, continuing a secular decline, especially pronounced in the Atlantic Provinces, extending over the previous two decades.

8 The question asked and answered was:

9 The following explanation has been given for the Bureau of Competition's Inter-departmental policy activities:

10 STB Ex Parte No. 582 (Sub-No. 1) Major Rail Consolidation Procedures October 3, 2000; Notice of Proposed Rulemaking. The STB announced that the new rules "represent a paradigm shift in our review of major mergers." The STB's summary describes the effect of its rules thusly:

11 CN November 2000 Submission to CTAR, Section 2.1.4.

12 The ICC Termination Act of 1995 instructed the STB, in making findings with respect to the competitive aspects of proposed mergers, to "accord substantial weight to any recommendation of the Attorney General [i.e., Department of Justice (DOJ)]" 49 USC § 11324. In the UP-SP case, the DOJ argued that the anticompetitive effects of the proposed merger were so great as to require that the application be denied. The STB devoted one third of a page of its 290 page decision to its denial of all DOJ requests - a space exactly equal to the 21 lines it used to explain its response to the Save The Rock Island Committee. The STB characterised one of DOJ's concerns as "remarkable" and concluded its discussion with the words: "We strongly disagree." Decision No. 44, Decided August 6, 1996, p.198. Such admonishments were reserved for a very select minority of intervenors. As noted by Edward Emmett, President of the National Industrial Transportation League "[I]n recent memory, every time the Secretary of Transportation or the Department of Justice weighs in with an opinion, the Board ignores them..." interview in Railway Age, February, 1997.

13 Canadian Shippers' Summit, Enhancing Rail Competition in Canada, Submission to CTAR, November 2000. Identical language appears in the Canadian Pulp and Paper Association Submission, October 6, 2000, p.8.

14 "Recommendation 11: MoTH recommends a public policy review of any proposed class 1 merger should be required under the CTA. In addition to general consideration of the public interest, part of this review should consider the impact of a merger on regional railways access to markets. With regard to specific issues related to rail markets and competition, MoTH would like to ensure the application of the Competition Act to railways. Recommendation 12: MoTH recommends the CTA apply the Competition Act to railways." British Columbia Ministry of Transport and Highways Submission to CTAR, December 2000, pp. 24, 25.

15 "CCPA believes that the Competition Bureau should continue to review the effects of proposed rail mergers or combinations. We also believe that there is a need for the Agency to have a role as well, while avoiding overlap with the Competition Bureau to the maximum extent possible." Canadian Chemical Producers' Association Submission to CTAR, October, 2000 p.27...

16 Saskatchewan Highways and Transportation Submission to CTAR, undated.

17 Understanding the issue has never been the railways' problem. Numerous articles have been published over the years on marketing and pricing issues that have emphasised the relationship between service reliability and value of rail service to customers. For example, in 1968, the Railway Systems and Management Association sponsored a symposium and published a document, The Measure of Railroad Freight Service, in which many of the leading lights of the rail industry addressed with great self-awareness the institutional, operational and economic aspects of railway service and the imperatives to improve it. In 1979, R.R. Latimer, Vice President and Senior Executive Officer of CN, discussed several of the same issues at a conference moderated by Dr. James C. Nelson, co-author of Railway Pricing Under Commercial Freedom. Latimer's comments were reprinted in Railway Age, May 1979. Thomas T. Nagle's The Strategy & Tactics of Pricing, Prentice Hall, 1987, which emphasised factors affecting customers' price sensitivities, was used as a guidebook by CSX Intermodal marketing managers.

18 U.S. General Accounting Office; Railroad Regulation: Changes in Railroad Rates and Service Quality Since 1990, April 1999. The report notes:

19 Some shippers suggest that the CN-BNSF merger would erode the efficacy of access provisions such as CLRs, on the theory that BNSF would be eliminated as a potential participant for shipments originating or terminating on CN lines. As the Canadian Resource Shippers' Corporation (CRSC) told the STB:

20 STB Ex Parte No. 582, decision served March 17, 2000.

21 CN, in particular has been responsive to the opportunities and imperatives of the trade agreements. As it told CTAR:

22 Statistics Canada, Rail in Canada, 1998, pp. 18, 21,22.

23 The STB summarised about 40 years of industry history by expressing the belief that through the merger process "railroads have now reduced most or all of their excess capacity, and have greatly improved the efficiency of their operations." This begs the question as to how the majority of plant rationalisation has come into being - of the 106 thousand kilometres of rail lines dispensed with by U.S. Class I's over the past 20 years, half have been acquired by short lines and regionals and much of the remainder abandoned outside of the merger process. In Canada, the entire rationalisation process has taken place without the intrusion of a major merger.

24 STB Ex Parte 582, op cit. Decision of October 3, 2000.

25 An evaluation of seven mergers performed for the U.S. Department of Transportation in 1975 concluded that "mergers did result in some improvements in certain cost categories. However, improvement was neither as great nor as widespread as might have been expected on the basis of projections made during merger proceedings... An evaluation strictly limited to changes in the sum of all performance measures examined supports a finding that three mergers could be judged successful and two unsuccessful, with inconclusive results for the remaining two. James Sloss, Thomas Humphrey and Forrest Krutter; An Analysis and Evaluation of Past Experience in Rationalising Railroad Networks; U.S. Department of Transportation Program of University Research, 1975, p.123. Twenty years later, an examination of U.S. rail mergers for the Ontario Ministry of Transportation found that little had changed, and benefits from four 1980s mergers were ephemeral. R.L. Banks & Associates and KPMG; Rail Merger Initiatives - the U.S. Experience, 1995. The STB has, however, found, less than a year after its approval, that the original projected merger benefits of the BNSF merger were understated, and used that finding as reason to reject challenges to the savings claimed - $627 million annually as restated by the STB - for the UP-SP merger. STB, FD 32760 (UP-SP Merger) Decision No. 44, August 6, 1996, pp. 109,110.

26 As examples, CN and CP currently share rail lines through the Rockies and CP has use of CN's Sarnia tunnel. In the U.S., as of 1998 Class I railways operated 30,962 kilometres of road - one fifth of the total owned - under trackage rights agreements, up from 22,215 kilometres in 1980. The predominant purpose of trackage rights has been to more efficiently utilise infrastructure; employment of the device to maintain competitive access in conjunction with mergers has been only a secondary purpose.

27 Kent T. Healy, The Effects of Scale in the Railroad Industry, Yale University, 1961

28 Wilner, Railroad Mergers; History Analysis Insight Simmons- Boardman, Omaha Nebraska, 1997, p.121.

29 U.S. railways have, on the other hand, frequently reorganised themselves sequentially - as "centralised" then "decentralised" and "centralised" again, vainly attempting to come to terms with inherently suboptimal management structures.

30 Comments submitted by railways to the STB in Ex Parte 582 evinced widely differing views on the ability of devices such as alliances and track sharing to generate benefits similar to those purportedly obtainable via merger, with newly merger-averse carriers NS, CSX and UP insisting that many merger benefits may be obtained without the benefit of clergy. CN and CP, both of whom have been actively reviewing restructuring opportunities in past months, maintained that merger proceedings should not become embroiled in questions of whether merger benefits could otherwise be obtained through speculative arrangements. CN claimed that if railways were able to realise by means of short of merger greater efficiencies than they have already realised by means short of merger, they would have done so on their own. Regulatory authorities should not be engaged in second-guessing of business judgements, management initiatives, and shareholder votes. CP added that arrangements which may work best for carriers in one set of circumstances may not work well for carriers in differing circumstances. The viewpoint that public agencies should not "second guess" the means by which railways intend to capture benefits is inapposite; the regulatory authority's function should be to determine if the benefits sought consist also in the purely private advantages obtainable through increased market power. It should be of specific concern to regulators that the reason that particular co-operative arrangements have not been already transacted by railways "on their own" as suggested by CN, is because a merger may be necessary to permit full exploitation of the arrangements' abilities to foreclose competition.

31 STB, Ex Parte 582, Op Cit., characterising Comments of Rail Labor Division of the Transportation Trades Department, AFL-CIO.

32 For example: ABC Rail Products "I can tell you that several times in the recent past, ABC has experienced shortages in the availability railcars for shipment on SP lines..." ACE Cogeneration "As a shipper dependent on SP service, we strongly support this merger that would result in bringing stability to management and ensuring the necessary capital to improve service." Allied Extruders: "we are concerned that SP has had financial problems for the past several years and has been unable to make necessary improvements in infrastructure." UP-SP Merger Application, Vol. I, Part 1.

33 See, e.g., Statement of Edward R. Hamberger, President of Association of American Railroads, before CTAR, October 2000. Hamberger highlights U.S. Department of Labor statistics indicating a 171 percent gain in rail productivity "in the post-Staggers era." (p.4.) Neglected is that the U.S. Bureau of Labor Statistics' Railroad Freight Producer Price Index, which "reflects price for shipping a fixed set of commodities under specified and unchanging conditions" increased from 75.3 in 1980 to 113.4 in 1998. (U.S. Statistical Abstract, 1999 Table 1066.) This 51 percent rise in nominal rates contrasts with an 18 percent decline in nominal revenue per ton-mile over the same period.

34 For example, the STB's Office of Economics, Environmental Analysis and Administration most recent rate study, released in December 2000, uses the inherently defective measure of "revenue per ton-mile" as a surrogate for rail rates, and put emphasis on the "additional $31.7 billion" rail shippers would have paid in 1999 if rates - as measured by revenue per ton-mile - had remained the same as they were in 1984. That shippers would have paid the extra billions is of course, highly unlikely; the STB's reasoning implies that demand for rail service is completely inelastic in the long run, completely the opposite of what railways argue when they claim lack of monopoly power.

35 Basic industries typically exhibit productivity improvements superior to that of the economy as a whole as they are far more receptive to process-related efficiency enhancements than are many service industries. Funeral parlours, symphony orchestras, universities, law firms, etc. are examples of industries that tend to have lower-than-average productivity improvements.

36 Frank N. Wilner; op cit., p.12.

37 Transport Canada, Transportation in Canada 1999, p.63ff.

38 STB Ex Parte 582, Decision of October 3, 2000, Op cit. p.335.

39 As pronounced in the STB's March 30, 2000 order in Ex Parte 582:"[W]e cannot close our eyes to the fact that the mere consideration of any major merger now would likely generate responsive proposals that, if approved, could result in a North American duopoly."

40 Certainly, if recent consideration of a CN-CP merger were to reach fruition, a very different configuration could eventuate. Given current developments in U.S. merger policy, a combined CN-CP, unless very liberally conditioned, would in all likelihood be foreclosed from merging with a major U.S. carrier. Such a follow-on consolidation would be anathema to non-included carriers and to a wide array of U.S. shipper interests; political and legal opposition would be formidable. A CN-CP merger would therefore be seen as contrary to current CN and CP strategies aimed at building a more effective transborder rail network. Such a merger would also raise issues of reregulation as noted by NTARC.

41 The STB succinctly summarised KCS's conception of the effects of duopolization:

42 Many shippers are doubtful; CPPA and the Western Canadian Shippers' Coalition are among the Canadian interests that argued to the STB that "experience with CN and CP has taught that, in the railroad industry, a duopoly is, in reality, a dual monopoly... capable of frustrating competition merely by declining to compete with each other..." (STB Ex Parte 582 Op cit. p.333) Similar sentiments were voiced by chemical companies BASF Corporation, OxyVinyls, LP and Williams Energy Services, adding that as each duopolist would have an interest in the survival of the other, for monopolisation would inevitably induce reregulation, there would be minimal rail-to-rail competition even for the minority of shippers which retained access to both systems. However, shippers' own studies demonstrate that the presence of direct rail-to-rail competition results in rates significantly below those faced by captive shippers; such evidence of differential pricing, not denied by railways, is central to the arguments for competitive access. This same evidence would tend to support the proposition that rail duopolies do face some very real market-based constraints that monopolies would not.

43 CN's submissions to CTAR indicate that rail-to-rail competition exists for only 41 percent of Canadian rail traffic.

44 Statutory authority is found in 49 U.S.C. §11323 et seq.

45 Submission to CTAR of the Commissioner of Competition, November 17, 2000, p.12. These are all classic issues in railway regulation, and the movement towards duopolization and monopolisation refocuses attention on "natural monopoly" questions as they pertain to the rail network.

46 In the Reservac/Pegasus case (a merger challenged by the Competition Bureau on March 4, 1988) the issue arose as to

47 "18 per cent is subject to product source competition. Newsprint produced in Quebec for markets in New York, Pennsylvania and New Jersey vies for market share with newsprint produced in Georgia and Alabama. Ontario and Alberta plastics and petrochemical producers compete for markets with producers in Louisiana and Mississippi. The above forms of competition are not mutually exclusive as more than one form usually applies. Canadian railroads, therefore, must price their transportation services so as to ensure that the products they carry can be sold competitively throughout North America. The same is true for virtually all export traffic that the railroads move to world markets." CN Submission to CTAR, November, 2000, p.16.

48 Competition Bureau Merger Guidelines, Section 3.1, Conceptual Framework.

49 U.S. Horizontal Merger Guidelines, Section 1.21, General Standards. Note that in the U.S., the absence of price discrimination ("differential pricing" in railway argot) is prerequisite to determination of relevant markets in which market and product competition come into play.

50 Monteiro and Robertson, op cit., p.514. The Competition Bureau has not, however, intervened in any railway freight mergers since prior to enactment of NTA 87, and what its current policy would be is difficult to discern.

51 Measurement of prospective increases in market power in terms of the ability to maintain non-transitory price increases may be of only limited applicability to railways, which may exercise increased power through reduction in quality of service.

52 The threshold test market share test used in the United States also fails to overcome the relevant market problems as faced in Canada. The Herfindahl-Hirschman Index ("HHI") is a statistical technique employed by the Department of Justice in analysing industries pre- and post merger to determine the impact of a merger/acquisition on competition. HHI indices, are calculated by summing the squares of market share for all market participants, and can range from slightly above zero to 10,000 (where one participant holds a 100 percent share - 100 X 100 = 10,000). The pre- and post merger change in the index is of principal concern. Nevertheless, it is a point of interest to show how the U.S. railway industry has progressed since 1980 with DOJ's test applied. Progression of market concentration measured by HHI under the simplifying assumption that two markets - Eastern and Western railways - exist in the U.S., is displayed in the Appendix, Tables A-1 and A-2, respectively.

53 Recommendations favouring some form of Joint Commission were made by both U.S. and Canadian governmental units to the STB in Ex Parte 582. For example, Saskatchewan advised, the "[c]reation of a Joint Commission between the U.S. Surface Transportation Board and the CTA Agency and/or Competition Bureau to discuss procedural issues" - U.S. Department of Transportation submitted that the STB should "engage in consultations with the pertinent foreign agencies charged with oversight, in whole or in part, of that transaction. Such consultations, DOT suggests: would minimise the possibility of misunderstandings respecting foreign law; and would provide information to decisionmakers of each agency on the status and prospects of other reviews of the same transaction." Representing Canadian shippers, CRSC requested that, with respect to a U.S./Canadian rail merger, STB regulations should "include formal consultation with the Canadian Minister of Transport and the Canadian Commissioner of Competition." STB Ex Parte 582, Decision of October 3, 2000.