Sunk Costs and Sustainable Competition in the Economics of Railroad Transportation / by Robert Gallamore

No one doubts that fixed and sunk costs in the railroad industry are as great as found anywhere in American industry.
— William B. Tye (1990, p. 86)

Among the most fundamental observations about railroad costs is that large
capital expenditures are required to construct a new railroad and launch operations.
These initial costs, sometimes called threshold costs, are “sunk” into the railroad’s
construction, infrastructure, and initial operation. They are required for obtaining a
business charter; surveying and purchasing right-of-way lands or easements; leveling
grades, bridging rivers and streams, and sometimes tunneling mountains; installing
crossties, rail, and ballast; constructing crossings, sidings, yards, and terminals;
developing train control, signaling, communications, and information systems; building
maintenance shops and general offices, and so on. In recent decades, an important
example of sunk costs for a major new project is that of preparing an environmental
impact statement (EIS) and obtaining necessary construction permits.

These infrastructure sunk costs might in themselves pose a substantial barrier to
the entry of a new competitor desiring to serve the same market, especially if such a
potential entrant is not already in a related part of the industry -- and is thus able to
leverage its operational know-how, its own maintenance shops, and its established
access to connecting traffic. Substantial barriers to entry might also be faced by a
potential market entrant proposing to compete against incumbents with a different
technology -- the possibility of a coal slurry pipeline competing against railroads moving
unit trains of coal, for example.

The circumstances of large threshold costs and presumed barriers to entry
conditioned traditional analysis of railroad-to-railroad horizontal competition in this
venerable industry, largely restricting market and pricing studies to consideration of
“who goes where?” and “how good are their routes?” Even normal antitrust analyses
applied in other industries – definition of relevant markets with cross-elasticities of
demand, investigation of product and geographic competition, and national branding
and price-maintenance (Robinson-Patman Act) strategies – have had relatively minor
importance for railroads. Minor, that is, as compared with the effects of traditional
“ICC-style” regulation of rates and services conducted by the Interstate Commerce
Commission (Surface Transportation Board after 1995); “ICC-style” regulation included
competitive analyses of mergers and trackage rights, and approval or not of line
abandonments and passenger train discontinuances.

Only in the last three to four decades (and as a result of railroad rate deregulation
effectuated by the Staggers Rail Act of 1980) have economists made major advances in
broadening microeconomic and industrial organization principles -- other than with
respect to mergers and abandonments – in application to analyses of railroads’
competitive efficiency and its impact on pricing. The new attention of economic
theorists has turned to the question of whether or not, in the absence of ICC regulation,
markets served by railroads would be sufficiently competitive to serve the public’s
important interest in economic efficiency. Further, this attention has been underscored
by the reality that the post-Staggers Act period has been characterized by a large and
funneling wave of railroad mergers – like high tides entering the Bay of Fundy. Today’s
railroad industry is clearly more concentrated than it was prior to mid-century; is it still
sufficiently competitive to perform well under deregulation?

This is the context in which the Theory of Contestability1 was developed, and in
which a rather surprising reevaluation of the timeworn notion of sunk costs has come
about. The reevaluation has produced some interesting wrinkles that are relevant in the
modern railroad context.

Railroad Networks, Sunk Costs, and Contestability
 

What do we mean by “barriers to entry,” “fixed costs,” and “sunk costs?” As
pointed out by consultant William B. Tye, some of the best ideas come from (without
fully coinciding with) the Chicago School of economics and the theory of contestability.
By “contestability” was meant dynamic conditions in an industry with single– or
multi-product firms normally realizing increasing returns to scale, such that market
power was present in sufficient amount to enable firms to earn adequate revenues.
Such combinations would permit the firm to remain viable (cover average total costs),
while rates were constrained to reasonable levels by credible threats of competitive entry
into the relevant market. “Captive shippers” (those not benefiting from adequate
competition) expected to see their rates capped by application of the Staggers Rail Act’s
Stand Alone Cost (SAC) test to levels a hypothetical dedicated, efficient, new market
entrant would charge. In fact, the ICC ruled in its Coal Rate Guidelines decision that
railroads are not sufficiently contestable overall, but that the SAC test could simulate
contestable market conditions in captive shipper cases.

Of course such a tightrope walk was controversial and gave rise to more than the
usual amount of learned academic discussion (lay people might characterize it
differently) -- as well as further regulatory hand-wringing. Captive shippers were not
satisfied that the Staggers Rail Act’s protections against their exploitation had been
honored adequately, and continued to seek legislative relief. But both deregulation of
rates and services and mergers tamed with competitive conditions could go forward,
and they did. More extreme remedies, such as competitive access compelled by regulatory
authorities to lower entry barriers (by reducing the sunk cost risks of entry and exit), or
nationalization and subsidy of bankrupt railroads, were not approved.

It is true that investments in railroad networks (as described above) are sunk.
They cannot be moved easily to other locations (time and place utility is everything in
transportation – see Box 2.3 in Gallamore and Meyer, American Railroads), and thus these
fixed and sunk costs have limited alternative-use value. This also means that network
operation is definitely not “perfectly contestable.” But service operation is a different
kettle of fish, since rolling stock and motive power can be redeployed, at least to some
extent. Also, since passage of the Staggers Rail Act of 1980, railroads have become
extraordinarily successful in developing intermodal (highway/rail) services in both
domestic and international transportation. These are based on unit trains of double-
stacked containers operating over railroads between major terminals (“ramp-to-ramp”),
and on public highways or waterways beyond. Intermodal services, because they
typically use road transport for origin and destination connections to extend railroad-
managed services off the fixed track networks, have liberated rail markets from some of
their former rigidity.

These are risks an incumbent railroad faces, but the risks are even more clear and
present for a new entrant. All those recent costs sunk in new construction and
competitive start-up are at risk and would have to be swallowed if a failed entrant must
later give up on the enterprise and exit the market. Thus, sunk costs are a large
impediment to new competition in the relevant market.
Next comes the issue of barriers to exit as well as entry. Taking what might be
assumed to be the usual case, incumbent firms have served their markets and
communities reasonably well. They have hired employees, given local shippers
opportunities to participate in distant markets, and provided ancillary businesses
locational advantages. They want to stay in that market as long as they can, but if any
combination of external circumstances and new competition makes an incumbent
railroad’s business unprofitable, it may try new pricing initiatives and may even have a
going-out-of-business sale. Ignoring the sunk costs of original investments, the railroad
may revert to out-of-pocket (short term variable cost) pricing for survival – any traffic it
can continue to attract will help pay the bills. These lower prices increase barriers to
entry, but may or may not injure overall social welfare, depending on the viability of
incumbent firms and their efficiency in serving the market.

If pricing remedies fail, an incumbent railroad may apply to abandon the line,
look for a merger partner, or even seek bankruptcy protection of its remaining assets.
Bowing to public pressure to preserve service and jobs, the regulatory authority may
disallow abandonment – increasing the costs of exit. (For American railroads in the
1960s and 1970s, this was the case for tens of thousands of miles of little used lines -- and
for hundreds of deficit-producing passenger trains.) The cost of exit for passenger
service, for example, could be tallied up by counting years of delay in discontinuing
deficit trains and by the initial outlays and ongoing opportunity costs of turning over
services to Amtrak after its establishment in 1971. (See Chapter 11 of American Railroads.)
In the rail world, the full import of sunk cost theory would have applied only to
network operations. It would not have applied to operation of “foreign” trains on
trackage rights (a formal agreement qualifying a tenant operator to use an owner
railroad’s facilities on a fee for service basis), or to flexible intermodal services. Nor
would it have applied to much of the remainder of network operations if the statutes
and regulators had been willing to structure an alternative competitive universe; this
hypothetical could have resulted from a requirement that incumbent railroads submit
to compelled open access -- the use by a competitor of a privately owned railroad’s
track and structures under a regulatory mandate for the express purpose of breaking an
incumbent’s monopoly power over prices. (See American Railroads, Chapter 9.)

Now economists had to ask if, under these circumstances -- that is, after passage
of the Staggers Act and the rise of competitive modes (e.g. larger combination trucks on
the Interstates) and efficient modalities (e.g. double-stack containerized intermodal
service) -- rail markets had become sufficiently contestable to relax the old constraints of
adversarial ICC rate and service regulation and to allow a transition to deregulation.

Barriers to Entry and Exit of Competitors
 

In contestability theory, sunk costs count as a barrier to entry. There are several
ways to think about why this is so. Economists, including Baumol and others often
associated with the Chicago School, emphasized that the need to incur large costs to enter a
market is not itself a barrier to entry, as long as incumbents do not have large cost advantages
over new entrants. But where sunk costs create cost advantages to incumbents, they may
be thought of as a barrier to competitive entry. In the Chicago School’s perspective,
incumbents have already incurred sunk costs, but potential entrants have not yet done
so. If newcomer firms can enter an industry on the same terms as incumbents, after
entry they can compete on equal terms and merely replicate whatever functions
incumbents had carried out before.

A railroad industry example of potential entry with large sunk costs might be the
cost of assembling a contiguous right-of-way for a pioneering transcontinental railroad.
More contemporaneously, the Dakota, Minnesota & Eastern Railroad sought permission
to construct a new rail line into the Powder River Basin (American Railroads, Chapter 10,
Box 10.3) enabling its entry as a third competitor for eastward movement of low-sulfur
coal; while the line is not yet built and its future is unclear, clearly the proposal
illustrated sunk cost entry barriers.

There are other ways in which sunk costs of railroad establishment and
operation might not prevent new market entry – new competition that could challenge
a monopolist’s power to restrict output and realize rents. For example, if the potential
entrant is well capitalized, it might be able to afford to build or buy infrastructure
facilities comparable to the incumbent’s. If the market is large and demand is booming,
there may be room in the market to support an efficient new competitor that will be
able to price at levels covering average costs – uninhibited by an incumbent’s implicit
threat of predatory pricing. That is, if the market is big enough to support several
competing regional networks -- the sunkness of investment costs may not do much to
keep competing networks from emerging. In this case, market size is everything.
If the market is small, however, the incumbent may have substantial existing
advantages making it costly or economically impossible for the new entrant to get a
foothold in the business. Indeed, there may be no more available routes over which a
competitive line could be built. Or regulatory authorities may believe there is
insufficient potential traffic for two railroads in the market -- fearing, for example, that
one or both competitors may be unable to cover fixed charges stemming from the new
entry or its impact, and thus stranding investments. In such a case, the regulatory
authority may be unwilling to grant a construction charter or certificate of convenience
and necessity to the potential entrant.

There is another line of thinking that leads to similar conclusions. Even without
regulatory intervention, the incumbent may have significantly declining costs with
greater volume, and thus, as a natural monopoly, may be able to set and sustain prices
lower than any challenger could. Sunk costs also might create opportunities for
incumbents to punish new entrants by cutting prices. In contravention of antitrust and
regulatory law, an incumbent might use predatory practices to thwart a newcomer,
deliberately lowering prices in the near term to discourage entry, then returning them
to monopoly levels when the threat of new competition is gone. With necessarily sunk
costs for any industry participants, incumbents can tailor prices to their advantage
faster than entrants can exit the market. Thus, if there is a risk to incumbents that new
rivals might upset their profit-maximizing pricing, there is even more risk to the
potential entrant, namely that perceived prevailing prices ex-ante might melt away ex-
post.

As pointed out by specifically by Professor John H. Brown of Georgia Southern
University, the real impact of sunk costs is that incumbents cannot recover their sunk
costs if they exit the market, and therefore only incremental costs matter in their
decision-making. Saudi Arabia might say (and mean) that it is cutting prices to punish
new entrants to world oil markets that are using fracking technology to recover crude oil
from tight reserves. As has been said, no cost is quite as sunk as a hole in the ground!
Further, imagine a different kind of case in which a single incumbent in a market
has no sunk costs to recover in subsequent product pricing, and can therefore exit the
market at no cost. According to contestability logic, nevertheless, the incumbent cannot
earn monopoly profits -- because if it tries to do so, potential entrants also with no sunk
costs can engage in “hit and run” entry. By definition, this means that the new entrant
can undercut the incumbent’s price and erode its market position. While rare in the
U.S., an example might be a case in which a commuter authority franchises operators to
provide service using the authority’s track and rolling stock; these contract operators
can come and go at the authority’s bidding without the need to sink costs on entry or
recover them on exit, because these are provided by the commuter authority. Still, the
contract operator might be in competition with another incumbent, such as Amtrak.

In sum, sunk costs are important in the microeconomic theory of competitive
ratemaking. These are what prevent natural monopoly or natural oligopoly markets
from being perfectly contestable. Violations of perfect contestability conditions
through the presence of sunk costs become a barrier to new entry, enabling incumbents
(again in theory), to set prices that are above marginal and average costs. Likewise,
sunk costs are also a barrier to exit from the industry, because they cannot be amortized
by moving assets to another market. Similarly, there is no salvage value in the assets
that the firm might recover or take with it in exiting the industry. Consequently,
according to the theory, the villain is sunk costs -- those that cannot be recovered by
exiting the market.

- # # # -

When students were studying economics in the 1960s, most all were told that
truly sunk costs could safely be ignored, because “bygones are bygones” and what
matters are investments going forward. That old learning may have subconsciously
guided me to leaving the discussion just presented out of American Railroads; I quite
literally “ignored” sunk costs. Like the railroads, so often “out of sight and out of
mind” but now hailed as an engine of economic growth and productivity, it is a bit
strange that the concept of sunk costs has moved from irrelevance to center stage in
important contemporary theories of the economics of transportation!

Where Are Railroads on the Cost Curve?
 

Critical to the theory of contestability and the preceding illustrations is the
assumption that incumbent railroads are operating at less than minimum efficient
scale. This means that each firm is producing on a declining portion of its average total
cost curve, as illustrated in Figure 1 [at page 20 infra]. It currently has declining
incremental costs with volume (scale of operations) and, necessarily, excess capacity.
The firm would benefit from increased demand, which would enable improved
utilization of its assets. Further, due to technical indivisibilities, or “lumpiness” (in
that the value of each unit output of the product is large relative to the size of the
relevant market), some of both incumbent’s and potential entrants’ costs are fixed over
output but not sunk over time. Think of a contractor who builds a few houses in the
space of a year, but has working capital tied up in inventory until those units are
finished and sold – or perhaps a rail line renewal project that takes two construction
seasons to complete and cut into service.

By contrast, perfect contestability is not violated by the actuality of fixed costs,
because these are simply cost indivisibilities trapped by the incumbent being forced to
operate at less than minimum efficient scale and therefore being forced to endure (and
absorb the costs of) excess capacity. For example, our illustrative building contractor
has one-half of a house left incomplete and unsold at the end of a season; he might have
avoided this inefficiency if higher demand had permitted scaling his operation that
season to a larger and evenly divisible output. Similarly, the rail renewal project might
have been completed in one season if more and larger equipment were brought to the
site and put to work – which in turn might have been cost-effective if the project were
planned for multiple tracks or longer track segments.

The importance of the cost issues addressed in this paper and illustrated in
Figure 1 (and the policy consequences that go with them), has everything to do with
where experts believe railroads are on the U-shaped curves. For years, students of
railroads have assumed that railroads operated on the left side of the chart, that they
were natural monopolies, that they had declining costs with scale, that marginal cost
pricing would bankrupt them, that value-of-service (VOS) pricing would be the manner
in which firms could afford to stay in business and fulfill their common carrier
obligation to serve all comers, and lastly, that only price differentiation could bring
revenue adequacy. John Meyer warned that just as ICC-imposed railroad VOS
ratemaking in the early decades of the twentieth century drove business to competing
modes (American Railroads, Chapter 2, p. 28 ff.), differentiated inverse-elasticity pricing
could do the same -- unless marketing officers are very skillful and have good
information on which to base their pricing decisions.

In contrast to the well-known history, recent experience in the industry gives
many indications that railroad firms and their markets have moved to the right side of
Figure 1. Here railroads need not try so hard to expand traffic, as to ensure that services
are priced to earn adequate revenues, win contract rate competitions with rates and
service quality that will perpetuate participation of customers in the market, and plan
capacity expansions to handle volume increases with better service.

The pricing paradigm that permits these things is not Ramsey inverse-elasticity
pricing, really, but Vickrey congestion pricing -- use of demand-based prices to ration
available capacity to the users most willing to pay for it. These are still differentiated
prices, but they have a different rational and perhaps a greater need for quick and
flexible application. The example provided in American Railroads is a set of trackage fees
private freight railroads might optimally receive from public passenger service agencies
for operation of their trains over (congested) freight infrastructure. Ideally, the Vickrey
congestion-based fees would be the same as prices based on opportunity costs, but
calling the trackage fees by that name is unpopular – perhaps it sounds too exploitative.
Meanwhile the public is getting more and more used to the idea – and the necessity --of
congestion tolls, so perhaps that is a better appellation than the economist’s term,
opportunity costs.

Thus is the complexity of railroad costs – sunk and fixed, declining with output
or rising with constrained capacity, and optimally priced only when demand elasticities
and their location on the U-shaped cost curves are properly estimated. These must
guide government and industry policy. Revision of received wisdom to fit these newly
understood circumstances is a necessary endeavor.

Contestability and Railroad Mergers
 

The previous discussion follows from and fits naturally with the dilemma of
railroad mergers described in American Railroads, including the historical policy debates
over the welfare advantages of industry consolidation vs. competition in relevant
markets. In the last few decades of the twentieth century, academic and regulatory
economists began to address the dilemma as a contest between the social virtues of
deregulation, on the one hand, in contrast to continuation of the drive for economies of
scale and scope, on the other. Depending on many circumstances, merger economies
presumably were available from either horizontal (parallel – “cost-saving”) or vertical
(end-to-end – “single-line service”) combinations. With jobs and competitive service
always at risk in parallel consolidations, end-to-end merger into a railroad reaching new
markets might be seen as the most palatable restructuring alternative from the
perspective of either the incumbent railroad or the regulatory authority. Williamson’s
Lemma, we called it in American Railroads.

Having passed through the enormous financial and personal costs, dislocations,
and uncertainties of the “Northeast Rail Crisis” (American Railroads, Chapters 6 and 7),
many observers hoped that the new theory of contestability would enable and
encourage regulators to approve both beneficial railroad mergers and deregulation. If
successfully executed, these would together free railroad managements from the
artificial and resource-distorting economic costs of traditional ICC regulation – as well as
serve the longstanding but unproven belief that railroads should be consolidated into
fewer, larger, and more streamlined firms.

Meyer and Tye on Contestability and the Transition to Deregulation
 

In the 1980s, John R. Meyer (1927-2009) and William B. Tye jointly authored an
important article on the applicability of sunk costs and contestability theory to railroad
deregulation in the latter part of the Twentieth Century.
be summarized as follows:

Under regulation, both railroads and their shippers sink costs in their respective
parts of the transport enterprise, but ICC regulation quite unnaturally outlawed normal
business-to-business contracts (see American Railroads, Box 9.2). If allowable, these
contracts would have protected subject businesses from random or opportunistic
behavior harmful to efficient rail service.

Tye recalls that his co-author Meyer viewed the presence of sunk costs as the
primary threat to a successful transition to deregulation, and he foresaw the need to
develop transition mechanisms. Under railroad regulation, both carriers and shippers
sank costs in their respective industries, but ICC regulation displaced the usual business
contracts existing in other industries to protect the parties from opportunistic behavior
that would otherwise be possible. Anticipating that law and policy would allow
loosening of the ICC’s grip on railroads, Meyer and Tye nonetheless were concerned
that simply going “cold turkey” from regulation would fail to account for the fact that
there is a legacy of carrier and shipper sunk costs from the prior regulatory regime.
(John Meyer called the phenomenon “regulatory overhang.”)6 At the same time, Meyer
and Tye reasoned, abrupt deregulation would not convey the contractual protections
that would have been in place in other capital-intensive and publicly beneficial
businesses before sinking costs.

The primary objective during a transition to deregulation, the two economists
argued, is to develop safeguards against rogue opportunism (misbehaving while the
teacher’s attention is diverted), but nonetheless to encourage switching from traditional
rate and service regulation to contract ratemaking as permitted by the Staggers Rail Act
of 1980. Importantly, the new contracts that ought to be put in place should take the
form of normal business contracts that would have been signed prior to the parties
sinking costs, (not spot contracts that were more likely under the legacy of sunk costs
from the prior regulatory regime).

To repeat, Meyer and Tye urged that railroad-shipper ratemaking agreements in
the transition to deregulation should mimic arrangements that would have been signed
in the absence of sunk costs. The proposed contracts would look like the kind
businesses would have put in place had they been signed prior to sinking costs.
Sponsorship and funding of airports is a particularly interesting application of the
“contract–before–sinking costs” model. Perhaps another useful analogy would be the
counterfactual outcome resulting if, in agreement with the British Parliament, the
American Colonials had imposed modest tea import taxes on themselves rather than
suffering from, and having to rebel against, King George’s imprudent levies.
Deregulation was like overthrowing colonialism; it would necessarily be disruptive, but
it needn’t lead to casting valuable tea overboard followed by armed rebellion -- and
certainly not the French Revolution’s Reign of Terror.

What Would John Meyer Say Today?
 

I learned from John Meyer that virtually all businesses, large and small, have
some market power, or they would be unable to survive. Economies of scale in
transportation businesses like pipelines, railroads, and airlines are part of their nature,
but could not by themselves create and sustain substantial market power; rather, the
industry and the firm needed durable, relationship-specific capital investments as well,
and these assets would be both large and sunk.

Professor Meyer illustrated this observation with discussion of airline economics.
Airlines enjoy pronounced economies of scale in aircraft size, and these indivisibilities
alone might constitute a barrier to entry allowing incumbents to profit from charging
rates above marginal costs. Still, planes are mobile and not sunk in specific markets, so
the threat of “hit and run” entry competition is credible. Like the important
observations made earlier regarding portable (not sunk) railroad operating equipment
such as locomotives and freight cars versus network sunk assets, contestability theory
applied to aviation became part of the justification for airline deregulation.

Such circumstances would be true for railroads as well as airlines if they were
organized along the European model of vertical separation – i.e. public ownership of
rail track and facilities, with franchised, above-the-rail operating companies competing
for deregulated traffic. Thus, one can imagine a regime, not unlike that in the
privatization of British railways, where (subject to price, access, and service regulation)
policy has granted to a named monopoly franchise long term responsibility for the
national publicly-owned railroad infrastructure (Network Rail), but permits and
facilitates competition among various franchised freight and passenger operators -- all
of whom pay fees for licensed access to operate over portions of Network Rail.

John Meyer was quick to point out, however, that this vertical separation
(“above-the-rail” versus “below-the-rail”) model for railroads has many other problems,
including the difficulty of making optimal trade-offs between capital improvements and
operating expenses; these are such as for maintenance of way, coordination of train
operations with track maintenance and renewal activities, setting dispatching priorities
among users, determining correct cost allocation and levels of user fees, and lastly the
fact that public ownership of infrastructure can perhaps too easily be used as a vehicle
for subsidy of operations. As noted in American Railroads, similar issues continue to
complicate American railroad economics in the one key area in which it relies on vertical
separation: long distance Amtrak passenger service outside the Northeast Corridor (this
on track owned by private profitmaking firms, not public infrastructure agencies).

From the point of view of practical railroaders, the academic arguments presented
in this paper were all much ado about very little, but the implications for policy toward
mergers and competitive access were real and worrying. Railroaders had seen cost
structures in their businesses change quite radically even before deregulation. Variable
costs were increasing as a share of total costs (sunk costs were being depreciated away or
written off), and the old ICC cost formulas such as Rail Form A had become hopelessly
obsolete. The ICC’s preoccupation with such things as setting divisions for interline
rates, prescription of car hire (per diem) rates, and productivity adjustments to the
prevailing inflation-adjusted railway charge-out price and cost indexes, were held in
contempt by the railroads – and were fast fading in relevance.

What railroads needed were adequate revenues, which before the turn of the
Millennium, they rightly believed might be achievable with Ramsey pricing and
differential rates, even if perfect contestability conditions had to be relaxed. Indeed,
William Tye, citing the historic dispute between Pigou and Taussig over whether
revenue adequacy and intramodal rail-rail competition are mutually exclusive (“the
oldest regulatory debate in the railroad industry” [Tye, 1990, p. 94]), wrote that markets
become unsustainable as circumstances move toward perfect contestability, and that
Ramsey pricing and perfect contestability are diametrically opposed concepts. In the
new Century, however, railroads have in fact been able to get their rates up to “revenue
adequate” levels for the most part, and traffic is filling up most excess capacity; the large
railroads are now investing unprecedented amounts in annual capital spending to
increase capacity.

All of this says that the times, they are a-changing. It is an important lesson to be
learned from study of an industry with sunk costs, but also one with historically
changing cost patterns and demand characteristics. This observation brings me back to a
reconsideration of John Meyer’s skepticism regarding extreme price discrimination as a
key answer to the rail industry’s problems. His objections arose chiefly from the strong
incentives that differentiated prices would create for adversely affected shippers to find
alternatives. It is now quite apparent that changes in the market for coal as a fuel for
power generation are resulting from inroads made by natural gas generation, plus
anticipated coal plant retirements due to cost and environmental concerns. These are
evidence of contestabilities quite literally replacing captivity. While declining coal traffic
volumes seem to have been cushioned by increasing rail traffic generated by fracking
and growth of other commodities as well as intermodal containers, John Meyer’s
concerns regarding coal traffic seem to be coming to pass.

Meanwhile, as detailed in American Railroads, the rail renaissance has continued.
Since the turning of the Millennium and especially since the end of the Great Recession,
large mergers overcame their teething problems, safety indices have been improving,
operating ratios showed efficiency gains, and new investments were adding to industry
capacity. Shippers were receiving lower rates overall, short lines were prospering,
employees were getting higher compensation, and no one was concerned about line
abandonments. With larger volumes of traffic moving under long term contracts having
inflation adjustment clauses, railroads saw competition for renewal of these contracts as
much more important than discussions of sunk or avoidable costs for marginal
businesses. How could anyone be thinking about reregulation when the Staggers Rail
Act was working so well?

In these remarkable ways, American public policy toward competition and
contestability in rail markets had reached a fairly stable equilibrium by the early years of
the new Millennium. How long it will endure remains to be seen.