Research in Transportation Economics 89 (2021) 101136

Contents lists available at ScienceDirect

Research in Transportation Economics

journal homepage: www.elsevier.com/locate/retrec

Access protocols for railroads: Reframing the infrastructure

separation debate
John G. Allen a, *, Gregory L. Newmark b
a
Transportation Consultant, 5518 S. Harper Ave, Chicago, IL, 60637-1830, USA
b
Department of Civil Engineering, Kansas State University, 1701C Platt Street, Manhattan, KS, 66506, USA

A R T I C L E I N F O A B S T R A C T

JEL classification: The Manichean framing of organizational arrangements for railroads as either vertical integration or infra­
R structure separation does not encompass the full array of access options in the industry. This paper views the
Keywords: conversation about access to rail infrastructure as an institutional matter substantially independent from eco­
Railroad ownership and control nomic debates regarding rail organization.
Institutional issues This research identifies a taxonomy of six access protocols which describe distinct arrangements for the use of
Vertical integration
railways, considers the conditions conducive to each institutional structure, and illustrates each category with
Infrastructure separation
Rail access protocols
actual examples from across the world. This effort seeks to expand the debate about railroad institutional
structures to include the full panoply of existing access arrangements.

1. Introduction separation should be implemented (Cantos et al., 2010; Cremer et al.,

2009), and infrastructure separation entered the US rail policy conver­
This paper takes a new look at an issue often taken for granted: the sation, also fueled by shipper concerns about the pricing power of the
ownership and control of railways (Moyer & Thompson, 1992; Posner, largest (Class I) freight railroads (De Vany et al., 1997; Jahanshahi,
2001; Gómez-Ibáñez 2004). Historically, railroads were vertically inte­ 1998; Allen, 2000; Evans, 2002; Preston, 2002; Sonstegaard, 2003).
grated, with track functions (e.g., dispatching, maintenance) and train But there is more to the institutional variety in the rail industry than
operations conducted by the same entity. When railways were brand a dualistic split between vertical integration and infrastructure separa­
new in the 1830s, there was some thought of opening the tracks to any tion. This Manichaean dichotomy unduly narrows the debate about
user willing to pay to operate on the line, much like a turnpike for road optimal rail policy by overlooking the institutional variation in rail ac­
traffic; but, before the invention of the telegraph, it was difficult to cess possibilities (Thompson and Helene, 1998; Hylén, 2001). This paper
communicate to train crews what trains were authorized to be on what seeks to expand that discussion by presenting a wider array of existing
tracks (an issue not arising for highways). Furthermore, no other access arrangements. This research draws on institutional and operating
would-be operators emerged, perhaps due to the capital cost of cars and experience from railroads in North America (Allen & Newmark, 2019;
locomotives. Therefore, track ownership, capacity management, and Winston et al., 2011) and railways in other parts of the world (Dunn
train operations became closely linked, establishing vertical integration et al., 2001; Lu, 2002; Gómez-Ibáñez 2004; Gómez-Ibáñez & de Rus,
worldwide (Jahanshahi, 1998, p. 73). 2006; Drew, 2006; Friebel et al., 2007; Van de Velde, 2018; Finger &
However, starting in the mid-1980s, vertical integration has been Montero, 2020) to identify six different access protocols. This work
contested by another paradigm, infrastructure separation, which di­ defines each protocol, considers its origins and attributes, and illustrates
vorces track ownership from train operations (Gibb et al., 1996; Irvine, this experience with examples. Finally, uses for the access protocol
1987, p. 20; Tye, 1987). First implemented in 1988 in Sweden as a cost concept are proposed for salient questions in the rail industry today.
transparency measure (Larsson, 2001), infrastructure separation
became European Union (EU) policy in 1993 when Directive 91/440 2. Analytical dimensions
took effect (Ludewig & Brinckman-Salzedo, 2008). Following this de­
cision, robust debate emerged in the EU as to how infrastructure Two basic analytical dimensions are needed for a more nuanced

* Corresponding author.
E-mail addresses: [email protected] (J.G. Allen), [email protected] (G.L. Newmark).

https://doi.org/10.1016/j.retrec.2021.101136
Received 25 May 2021; Received in revised form 16 October 2021; Accepted 7 November 2021
Available online 14 November 2021
0739-8859/© 2021 Elsevier Ltd. All rights reserved.
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

understanding of the ownership and control of railroads. One is the Under vertical integration, unitary access is found on most
fundamental organizational structure, regardless of public or private government-owned railways outside Europe, most North American
sector ownership. This dimension, termed organizational regime, is based private-sector regional railroads (Class II), and short lines (Class III), and
on the relationship between track ownership and train operations, with on smaller Japanese railways. Normally a factor of ownership, unitary
the options being vertical integration and infrastructure separation. The access may also be implemented through long-term concessions, as in
second analytical dimension, termed access protocols, describes the Mexico, Panama, Brazil, and Argentina (Barbosa, 2019; Campos, 2001;
specifics of what entities use the railway in practice. Access protocols Kogan & Thompson, 1994; Thomas, 2006).
offer a framework for understanding subtle nuances in rail institutional Somewhat counterintuitively, de facto unitary access is possible
arrangements. under infrastructure separation if no operators other than the incumbent
seek to use the railway. For instance, the island of Ireland has a track
3. Access protocols gauge not found elsewhere in Europe. Due to its geographic isolation,
Iarnród Éireann (IÉ), also known as Irish Rail, was exempted from the
This research identifies six existing access protocols – unitary access, requirement to split into separate operating and infrastructure com­
disaggregated unitary access, negotiated access, mandated access, joint panies until 2013. In operating terms, IÉ still functions under de facto
access, and competitive access – all of which may be found under both unitary access because no other firms have contested such a small
organizational regimes. Table 1 summarizes the six access protocols, market. The same is true of Uruguay’s rail system, the Administración de
their origins, their outcomes, and issues they raise. Ferrocarriles del Estado (AFE) (Marmouget, 1969), which although
nominally split into separate operating and infrastructure entities
(Thomas, 2007) continues to function under unitary access (World Bank,
3.1. Unitary access 2008).

Unitary access is the simplest of all access protocols and means that
only one operator has the right to use the railroad. This right is usually a 3.2. Disaggregated unitary access
factor of ownership but may also come about through lease or conces­
sion. Unitary access is normally associated with vertical integration, Disaggregated unitary access occurs when a full-service railroad
although it is (very rarely) found in infrastructure-separated settings. (operating freight, intercity passenger, and/or commuter trains) dis­
Historically, unitary access has been the organizational basis for the tributes these functions to different entities. In this arrangement, mul­
rail industry, and it offers the major benefit of simplicity as access is tiple operators use the same lines, but there is only one operator for each
limited to a single entity. But unitary access becomes problematic when type of service.
connecting operators are involved, because simply exchanging cars with Disaggregated unitary access increases cost transparency for
other railroads at junction points may not be satisfactory. Unitary access different services and has consequently emerged in response to financial
is also hampered by accounting inefficiencies, particularly for large losses on railroads performing multiple functions. Disaggregation in­
government-owned railways, because costs are often aggregated rather creases institutional complexity (Allen & Lu, 2010, pp. 106–108). It may
than allocated to specific lines and services, thus making cost control also increase costs when rolling stock is no longer shared and new yards
difficult. and other facilities are provided for different services, but these cost

Table 1
Access protocols: Characteristics, origins, outcomes, and issues raised.
Access Protocol Characteristics Origins Government Role Outcomes Issues Raised

Unitary Access One exclusive operator on a Original form of Prevent abuses of Works best where operations and Depending on system geography,
given railroad or line. organization for market power. geography are simple. As size and may result in controversy about
railroads as they came complexity increase, other access market power of freight
into service. protocols may become desirable. railroads. Especially in public
sector, may conceal serious
inefficiencies.
Disaggregated More than one user, but Occurs when different Prevent abuses of On any given route, a dominant May require provision of
Unitary only one user per sector types of trains are market power. carrier becomes the track owner and separate locomotives, yards,
Access (freight, commuter, distributed among runs its own trains. Other carriers work equipment, etc., although
intercity passenger). different operators. become tenants. efficiency gains usually far
outweigh these capital
investments.
Negotiated One railroad agrees to let One railroad needs to Provide legal Results in tenant railroads expanding Requires the parties to come to
Access another use its tracks for use another’s lines to framework for their geographic reach and host mutually beneficial terms.
their mutual benefit. reach an important voluntary access railroads receiving additional income Asymmetries of size may put
destination. agreements. for use of surplus track capacity. smaller carriers at a
disadvantage.
Mandated Legislation (or regulatory Sometimes, market- Provide for access May help mitigate anticompetitive If regulators (or legislatures) are
Access conditions for a merger) based interactions fail to where necessary, effects of large mergers. Thus, when seen as taking rights away from a
gives one or more other provide socially desired while minimizing Railroads A and B merge, regulators railroad, private investment in
operators the right to use levels of competition. financial effects on rail may grant Railroad C limited access the rail sector may evaporate
the railroad. industry. to certain specific lines. while the issue is litigated.
Joint Access Facilities, lines, and May be more Ensure equitable Provides greater physical reach at Joint access usually provides
sometimes entire railroads economical to share access among users. reduced costs by sharing assets. Has clear benefits to customers as
are owned on a joint basis. facilities than for historically included union passenger well as railroads.
railroads to build and stations; now usually involves jointly
operate separate lines. owned terminal freight railroads.
Competitive Any qualified operator may Implemented to provide Enforce evenhanded If designed and executed properly, Economic interests of track-
Access use the railroad. multiple operators on a conduct by access- may lead to stable rail freight duopoly owning and train-operating
single line. granting track owners. with well- funded infrastructure. entities must be aligned. If this
does not happen, safety may be
adversely affected.

2
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

increases are seen as minor compared with the broader efficiency gains example, on one major European railway, the dominant operator
of disaggregation. “remained the manager of all assets required for the operation of
Under vertical integration, disaggregated unitary access is evident in transport services, such as train stations, freight terminals, marshalling
many places. In North America, Amtrak and Via Rail Canada took over yards, warehouses or storage facilities” (Blutel, 2015, p. 211), limiting
intercity passenger trains in 1971 and 1979, respectively, from railroads the infrastructure manager’s ability to act impartially – a situation later
anxious to shed these costly services (Guerrero et al., 2011; Iacobacci, ruled illegal by the European Court of Justice in 2013.
2010; Wilner, 2012; Zarembski, Blaze, and Patel 2012). States, prov­
inces, and regional agencies assumed (or inaugurated) commuter service 3.4. Mandated access
on existing rail lines (Allen & Levinson, 2014; Liu et al., 2005). In Japan,
passenger operations dominate, and freight – a sector of lesser priority – Mandated access occurs when certain specific operators are granted
was disaggregated from the core passenger business (Aoki et al., 2000, access to another company’s tracks, often as part of a merger applica­
pp. 171–193; Ishikawa & Imashiro, 1998; Mizutani & Fukuda, 2020). tion. For example, US regulators have required that if two companies
Prior to Great Britain’s infrastructure separation starting in 1994, the combine, they must sell portions of the combined system, or allow
1983 reorganization of British Rail (then vertically integrated) from potentially adversely affected third-party railroads to operate on por­
geography-based operating regions to business sectors provided another tions of the combined system. Merging railroads view mandated access
example of disaggregated unitary access (Railway Gazette International, as undesirable and typically reach negotiated access agreements with
1984). third parties to prevent a less-advantageous regulatory ruling or to head
Under infrastructure separation, an interesting example of dis­ off costly legal action (Wilner, 1997, p. 228). Regulators also tread
aggregated unitary access occurs on Australia’s interstate lines, where a carefully in mandating access to avoid even the appearance of unfairly
single franchised operator provides intercity passenger service, and redistributing one railroad’s rights to other carriers or to shippers (Sidak
freight services operate under competitive access (as discussed below). A & Spulber, 1997). A poorly reasoned decision to mandate access may
similar form of disaggregation occurs, more controversially, in Great leave the government itself liable to litigation.
Britain, where intercity passenger and commuter services are franchised Mandated access has almost always occurred under vertical inte­
and freight operates under competitive access (Preston 2001, 2002). gration. In the US, it commonly results from regulatory proceedings.
Thus, as a condition of Union Pacific’s (UP) acquisition of Southern
3.3. Negotiated access Pacific, the STB granted competitor Burlington Northern Santa Fe
(BNSF) limited trackage rights on UP between Denver and northern
Negotiated access allows one railroad to use another’s tracks to reach California.
important destinations. Tenant railroads expand their geographic reach, A variation on this theme is found in Canada in the form of reciprocal
and host railroads monetize excess capacity. Of course, the parties must switching, also known as interswitching. Although it does not actually
come to mutually beneficial terms. Asymmetries of size and financial give one railway access to another’s tracks, its requirement for inter­
resources may put smaller carriers at a disadvantage when negotiating change is tantamount to mandated access, and is best understood in this
with larger ones. light. In this arrangement, the regulator can, under certain circum­
Under vertical integration, North American railroads use negotiated stances, require that cars being sent from one point to another on
access on a widespread basis. In the US, negotiated access comes in two Railway A be interchanged onto competing Railway B for the long haul,
forms: trackage rights and haulage rights. With trackage rights, a tenant to minimize market abuse by the first railway (Nolan et al., 2020, pp.
pays the host to use the host’s tracks between specified points. The 3–4). Interswitching is an administrative last resort, available only if
tenant provides its own locomotives and arranges for its employees to origin and destination are both within 18 miles of an interchange, and
qualify for operating on the host railroad. Depending on the nature of even then is very rarely invoked. Nevertheless, a 2016 STB proposal to
the agreement, the tenant may or may not serve customers along the require reciprocal switching in the US met with strong opposition from
host’s line. Trackage rights agreements are registered with the Surface railroads.
Transportation Board (STB). With haulage rights, the tenant pays the There was a brief example of mandated access under infrastructure
host railroad to provide train crews (and sometimes locomotives) to separation in Estonia. The Baltic state privatized its national railway,
serve customers along an agreed-upon line or segment. The tenant Eesti Raudtee (ER), in 2001 (Koppel, 2006). The sale separated infra­
markets the service to shippers as though it, rather than the host rail­ structure and operations, as required under EU law, and “allowed for
road, were the service provider. These agreements are not registered open access – initially 20 percent of its capacity” (Grant, 2015 p. 178).
with the STB. Then in 2003, a new government unilaterally expanded that share to
Two railroads may arrange for directional running, a form of nego­ 100 percent of capacity – essentially mandating access on the other 80
tiated access on parallel or nearby lines, giving each road trackage rights percent. ER contested this mandate in court as an uncompensated taking
over the other’s line in one direction, thus turning two single-track lines contrary to EU law. The court sided in ER’s favor, resulting in Estonia
into the equivalent of a double-track line. This maximizes track capacity renationalizing the railway in 2006 and ending this short period of
without adding more sidings for opposing-direction trains to pass. Ca­ mandated access under infrastructure separation.
nadian National and Canadian Pacific have largely single-track lines on
opposite sides of British Columbia’s narrow Fraser River Canyon, where 3.5. Joint access
adding tracks is not feasible. The two railways have had directional
running between Kamloops and Hope since 1999 (Del Grosso, 2001 p. Joint access occurs when two or more railroads, sometimes com­
55). petitors, find it mutually advantageous to jointly own or manage facil­
Negotiated access may also occur under infrastructure separation on ities, lines, or even railroads that extend their geographic reach. This
a de facto basis when the infrastructure manager and the dominant partnership represents a full sharing of access rights rather than the
operator are both government owned, such as in France and Italy. If the more limited sharing under negotiated access. Joint access allows rail­
infrastructure manager favors the dominant operator in allocating ac­ roads to share the cost of reaching an important location, expanding
cess, this functions analogously to negotiated access because the com­ opportunities for them and their customers. Of course, the price of joint
bination of the infrastructure manager and the dominant operator can access is that one or more other railroads also enjoy this access.
preclude fair access for other entrants (Benedetto, Smith, and Nash Under vertical integration, joint access in North America sometimes
2017). One situation that can lead to such favoritism is when the takes the form of a jointly-owned terminal freight railroad providing
infrastructure management entity relies on the dominant operator. For switching for two or more Class I railroads in and around a major city.

3
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

Examples include the Belt Railway of Chicago, Indiana Harbor Belt, the integrity, and accountability.
Terminal Railroad Association of St. Louis, and the Ferrocarril Terminal
del Valle de México. 4. Using access protocols
Joint access occurred on a widespread scale until 1970 between
Chicago and the Pacific Northwest. Northern Pacific and Great Northern Fig. 1 provides a sorting algorithm, using a step-by-step decision tree
were competitors between Minneapolis/St. Paul and the Pacific North­ to help determine which access protocol or protocols apply to any given
west, but they jointly owned connections at each end: Chicago, Bur­ line or railway. Unlike organizational regimes, more than one access
lington & Quincy between Minneapolis/St. Paul and Chicago, and protocol may exist on the same line, e.g., competitive access for
Spokane, Portland & Seattle at the western end. All four merged into Australian interstate freight but disaggregated unitary access for long-
Burlington Northern in 1970. Similarly, union passenger stations, distance passenger service. Thus, when using Fig. 1, it may be neces­
owned jointly by participating railroads and managed for the benefit of sary to specify the sector being examined.
all holders, were common before Amtrak’s founding in 1971. Access protocols may be used to facilitate analysis of existing and
Sometimes, arrangements for joint access come about through proposed institutional arrangements for railways. Generally, the first
contentious acquisitions. The 1999 division of Conrail between CSX three protocols (unitary access, disaggregated unitary access, and
Transportation (CSX) and Norfolk Southern (NS) was preceded by negotiated access) lack safeguards against the creation of rail freight
intense acrimony over which railroad would get what parts of Conrail monopolies. Competitive access runs the risk that the interests of the
(Loving, 2006). As part of the final agreement, the two acquiring rail­ track owners and train operators may be misaligned, resulting in
roads created a jointly-owned terminal railroad, Conrail Shared Assets insufficient funding to maintain the physical plant (perhaps less of a
Operations (CSAO), owning tracks and yards in the Detroit area, danger with mandated access, depending on the specific arrangements).
northern New Jersey, and southern New Jersey and Philadelphia. The situation with joint access depends on the larger context outside the
Joint access is normally associated with vertical integration, but is jointly-owned or operated lines and facilities.
widespread on a de facto basis in infrastructure-separated Great Britain, Access protocols alone do not provide answers as to which institu­
where different train operating companies serve rail passenger stations tional arrangements are best for any given situation, as other factors are
(Hylén, 2001, p. 27) in the manner of union stations in the US prior to necessarily involved. Geography plays an important role (Vance, 1995).
Amtrak. In North America, there are two major Eastern, Western, Canadian, and
Mexican freight railroads, operating under various access protocols on
3.6. Competitive access different parts of their systems, providing better service to shippers,
more efficient operations, and more competition than would be possible
Competitive access (sometimes referred to as “open access” in North with unitary access alone (Allen & Newmark, 2019). This nuanced
America) gives all qualified entrants the right to use the same rail line. approach was not applicable to Australian interstate freight, with its
Any firm meeting various financial, training, safety and other re­ single transcontinental line. Therefore, competitive access was the only
quirements specified under law may arrange to operate trains on specific way to enable more than one entity to operate freight trains.
lines with the track owner (Haydock, 2013). These protocols do not explain all differences between railroads,
If implemented carefully, competitive access can work well, which may be due to such factors as national politics (Zahariadis &
assuming access charges are sufficient to fund a state of good repair. But Allen, 1995), system geography, traffic levels, competitiveness with
if train operators and track owners have conflicting incentives, safety other modes, the balance between freight and passenger services, and
may be seriously compromised (Bagwell, 2004; Bartle, 2004; Fowkes & private or public sector ownership.
Nash, 2004; Hope, 2002). France and Germany operate under similar access protocols with
Competitive access is found almost entirely on infrastructure- comparable infrastructure and rolling stock yet have different attitudes
separated railways. Interstate rail freight in Australia comes closer toward freight. Given the relatively short distances and the passenger-
than anywhere else to the theoretical ideal of a steel-wheeled toll road oriented physical plants of European railways relative to American
for trains (Starrs, 1999; Bureau of Transport & Regional Economics, railroads (Vassallo & Fagan, 2007), France provides a supportive insti­
2006; Fagan, 2007; “Review of rail access regimes” 2018). Australian tutional environment for rail freight (Grant, 2015 pp. 200-203; De La
National Railways (ANR), a federal government corporation, formerly Rosa, 2020 pp. 37–39). Germany, by contrast, has rejected EU efforts to
owned the country’s sole transcontinental railway. But when the federal prioritize freight, expressing concerns that this might disrupt the
government adopted a macroeconomic policy emphasis based on execution of carefully-crafted timetables, causing passengers to miss
competition, ANR came under critical scrutiny as a monopoly. As there their connections (Railway Gazette International, 2009).
is not enough traffic to justify another transcontinental rail line, Some thoughts are offered below on how access protocols can help
competitive access was the only way to give shippers a choice of carriers. structure research into optimal rail policy.
ANR’s successor, Australian Rail Track Corporation, owns and maintains
the physical railway for the use of fee-paying operators. European rail­ 4.1. North America
ways are also moving toward competitive access for freight, as in Great
Britain (Wilner, 1998; Jack, 1998; Woodburn, 2001; Fowkes & Nash, North American railroads operate under all access protocols
2004) and France. Similarly, the Alameda Corridor between Los Angeles (including competitive access on Los Angeles’ Alameda Corridor).
and Long Beach, California is owned by a public authority which collects Despite occasional exceptions (Nolan et al., 2020), mandated access,
trackage fees from users BNSF and UP (Railway Gazette International, both legislative and regulatory, has received less attention than it de­
2002). serves. With the STB considering reciprocal switching for US railroads,
Perhaps counterintuitively, there is at least one known instance of the industry needs more research, accessible to decision makers, on how
competitive access under vertical integration. Australia’s corporatized (and whether) it has affected pricing and revenue for Canadian railways
Queensland Rail (QR), owned by the state government, is nominally – and what the implications are for US railroads.
vertically integrated, but allows other qualified operating companies to Little has been done with joint access at a level that is useful to in­
use its tracks (Bilyk, 2003). Obviously, this requires robust internal dustry professionals about the economics and operations of terminal
organizational firewalls so that information provided by firms seeking freight railroads, as opposed to trackage rights under negotiated access
access is not leaked to QR’s operating division to give it unfair com­ or mandated access to mitigate possible anticompetitive effects of
mercial advantage. Therefore, this option is appropriate only in juris­ mergers. Should regulatory policy encourage more joint access, such as
dictions with very high levels of professionalism, political and judicial Conrail Shared Assets Operation, as opposed to piecemeal mandated

4
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

Fig. 1. Decision tree for classifying examples of access protocols.

5
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

access in conjunction with mergers? protocols, ranging from unitary access, the simplest (and most restric­
tive), to competitive access, which approximates the concept of a steel-
4.2. Europe wheeled toll road, open to all qualified users. Examples of all these
protocols have been identified and discussed for both organizational
Several European national railways have fallen short of the full re­ regimes.
quirements of the European Railway Packages, the institutional suc­ It has become clear since the turn of the 21st century that the vertical
cessors to Directive 91/440 (Ludewig & Brinckman-Salzedo, 2008). integration versus infrastructure separation dichotomy is not enough to
Access protocols might help evaluate opposing claims that railway of­ adequately describe the full range of possible forms of track access.
ficials are protecting inefficiencies versus protecting market stability Scholars, analysts, and rail industry professionals need a more
and the financial health of large operators from potentially predatory comprehensive lexicon for describing institutional arrangements.
behavior. Indeed, access protocols may be more important than the vertical inte­
At present, European regional railways and short lines must choose gration – infrastructure separation dichotomy for ensuring desirable
between the default protocol of unitary access and full competitive ac­ policy outcomes.
cess. Might other access protocols be better suited to the needs of Taken together, organizational regimes and access protocols provide
regional railways, such as Italy’s Ferrovie Nord Milano, than just these a versatile and fine-grained framework for analyzing the ownership and
two diametrically-opposed arrangements? control of railroads. Adding other factors such as passenger vs. freight
The access protocols concept may also help structure other research orientation and public or private ownership may provide even more
on the EU’s rail sector. For instance, Estonia’s brief experiment with 20 detailed analysis, along with geography, traffic density, and the political
percent competitive access and 80 percent unitary access for Eesti environment.
Raudtee’s capacity suggests a potentially promising way to bring in­ The access protocol concept is helpful for understanding arrange­
vestment capital into railways subject to infrastructure separation. Can ments as they exist on the ground, not just on paper. When brought to
this be reconciled with EU regulations, or would governments need to bear on railway policy, access protocols may help governments craft
seek exemptions? institutions and regulations that provide a better fit with the operating
and commercial realities that railroads face.
4.3. Australia
CRediT authorship contribution statement
Australia is known for implementing competitive access smoothly.
Research might be helpful into possible differences between access ar­ John G. Allen: Conceptualization, Methodology, Formal analysis,
rangements on the infrastructure-separated Australian Railway Track Investigation, Writing – original draft. Gregory L. Newmark: Visuali­
Corporation (ARTC) and on the vertically-integrated Queensland Rail zation, Resources, Formal analysis, Validation, Writing – review &
(Bilyk, 2003; Talebian et al., 2016). editing.
Access protocols might also help analyze the largest controversy in
Australian rail freight. The Pilbara region of Western Australia has References
several heavy-haul iron ore mining railroads not connected with the
general railway system and operating under unitary access. Government Allen, J. G. (2000). Commuter rail, freight railroads, and the open access debate.
Transportation Research Record: Journal of the Transportation Research Board, no.
authorities and smaller mining companies have sought competitive ac­ 1704, 35–41.
cess to these lines. The companies that built them argue that these Allen, J. G., & Levinson, H. S. (2014). Accommodation of long-term growth on North
railroads, built to meet their firms’ production needs, are fundamentally America’s commuter railroads. Transportation Research Record: Journal of the
Transportation Research Board, no. 2419, 40–49. https://doi.org/10.3141/2410-05
different from common-carrier lines owned by state and federal gov­ Allen, J. G., & Lu, A. (2010). Organizational regimes for commuter rail: Improving train
ernments. Court judgments have been contradictory and uneven in their service worldwide. Transportation Research Record: Journal of the Transportation
logic (Railway Gazette International, 2008; Railway Gazette Interna­ Research Board, no. 2146, 100–108. https://doi.org/10.3141/2146-13
Allen, J. G., & Newmark, G. L. (2019). Sustainability without subsidy: The public case for
tional, 2010). Is competitive access the best policy for the Pilbara mining vertically-integrated rail oligopolies for freight. Transportation Research Record:
railroads? Or might negotiated access be preferable? Journal of the Transportation Research Board, no. 2673, 204–214. https://doi.org/
10.1177/0361198119843861
Aoki, E., Imashiro, M., Kato, S., & Yasuo, W. (2000). A history of Japanese railways, 1872-
5. Comparisons across settings 1999. Tokyo: East Japan Railway Culture Foundation. https://trid.trb.org/view.as
px?id=584971.
Access protocols could be used to compare how trackage rights work Bagwell, P. (2004). The sad state of British railways: The rise and fall of Railtrack,
1992–2002. The Journal of Transport History, 25(2), 111–124.
in practice for less-dominant sectors, such as Japanese rail freight or
Barbosa, F. C. (2019). Brazilian freight rail concessions overview: Current outcomes and
intercity passenger trains in the US, Canada, South Africa, and Australia. perspectives. In 2019 Joint Rail Conference. Snowbird, Utah: American Society of
Finally, access protocols may help analyze the potential of institu­ Mechanical Engineers Digital Collection.
tional arrangements other than the unitary access prevailing on many Bartle, I. (2004). Britain’s railway crisis: A review of the arguments in comparative
perspective. Centre for the Study of Regulated Industries, University of Bath.
national railways serving large countries, such as South Africa (Chauke Benedetto, V., Andrew, S., Smith, J., & Nash, C. A. (2017). Evaluating the roles and
& Maluleke, 2005), China (Cui, Pittman, and Zhao 2021; Pittman, 2004; powers of rail regulatory bodies in Europe: A survey-based approach. Transport
Yin-Nor, 2020), Russia (Pittman, 2004; Saakyan, 2020), and Ukraine Policy, 59, 116–123. https://doi.org/10.1016/j.tranpol.2017.07.003. October.
Bilyk, P. (2003). Third party access under an integrated structure: Lessons from QR’s
(Pittman, 2017). What is the potential for introducing competition access undertaking. In ACCC Regulation, Industry Structure and Market Power
through allowing other entrants to have rail freight access in such Conference. August.
countries? Which mix of access protocols would be most appropriate? Blutel, J.-M. (2015). France: The independence of the French rail infrastructure manager:
Is the reform on the right track? European Networks Law and Regulation Quarterly, 3
(3), 211–214.
6. Conclusion Bureau of Transport and Regional Economics. (2006). Optimising harmonisation in the
Australian railway industry. Canberra, A.C.T, Australia: Bureau of Transport and
Regional Economics.
When examining the ownership and control of railroads, most re­ Campos, J. (2001). Lessons from railway reforms in Brazil and Mexico. Transport Policy,
searchers focus on the choice of organizational regime, i.e., vertical 8, 85–95.
integration versus infrastructure separation. Yet important though that Cantos, P., Manuel Pastor, J., & Serrano, L. (2010). Vertical and horizontal separation in
the European railway sector and its effects on productivity. Journal of Transport
choice is, this research highlights another crucial analytical dimension –
Economics and Policy, 44(2), 139–160.
access protocols, which establish in more detail (officially or otherwise)
what operators may use the railway. This research identifies six access

6
J.G. Allen and G.L. Newmark Research in Transportation Economics 89 (2021) 101136

Chauke, S., & Maluleke, J. (2005). Separation of rail infrastructure from operation: Is it a national railway restructuring. In Research in transportation business & management,
feasible model for the South Africa rail transport industry?. In Proceedings of the 24th modal shift and logistics integration in intermodal transport networks’ in research in
southern African transport conference (SATC 2005). Pretoria, South Africa, 2005. transportation, business and management, 35. https://doi.org/10.1016/j.
Cremer, H., De Donder, P., & Cremer, J. (2009). Costs and benefits of vertical divestiture. rtbm.2020.100484. June, Article Number: 100484.
Communications and Strategies, 68, 1–16. Moyer, N. E., & Thompson, L. S. (1992). Options for reshaping the railway. Report WPS
Cui, S., Pittman, R., & Zhao, J. (2021). Restructuring the Chinese freight railway: Two 926. Washington, D.C.: World Bank Publications.
scenarios. Asia and the Global Economy, 1(1). Nolan, J., Su, C., Logan, P., & Peterson, S. (2020). “Parallel or converging? A comparative
De La Rosa, S. (2020). The ongoing process of liberalization of the railway sector in analysis of the grain and rail transportation systems in Canada and the United States.
France. In Handbook on railway regulation. Cheltenham, England: Edward Elgar AgEcon Search, 2, 2020. https://doi.org/10.22004/ag.econ.307243. November.
Publishing. Pittman, R. (2004). Chinese railway reform and competition: Lessons from the
De Vany, Arthur, S., & David Walls, W. (1997). Open access to rail networks. experience in other countries. Journal of Transport Economics and Policy, 38(2),
Transportation Quarterly, 51(2), 73–78. 309–332.
Del Grosso, R. C. (2001). Canadian National Railway: North America’s railroad. Bonner’s Pittman, R. (2017). Reforming and restructuring Ukrzaliznytsia: A crucial task for
Ferry, Idaho: Great Northern Pacific. Ukrainian reformers. Наука и Прогресс Транспорт а. Вест ник Днепропет ровского
Drew, J. (2006). Rail freight: The benefits and costs of vertical separation and open Национального Университ ет а ЖелезнодороЖного Транспорт а, 1(67), 34–50. https://
access. In Proceedings, European Transport Conference. doi.org/10.15802/stp2017/92775
Dunn, , Jr., James, A., & Anthony, P. (2001). Toward a ‘new model railway’ for the 21st Posner, H. (2001). Global strategies for railroading. Trains. April 2001.
century: Lessons from five countries. Transportation Quarterly, 55(2), 43–57. Preston, J. (2001). Franchising and refranchising of passenger rail services in Britain.
Evans, J. (2002). The case for separation of infrastructure. Brussels, Belgium: European Transportation Research Record: Journal of the Transportation Research Board, no.
Railway Circle Luncheon. September 25, 2002. 1742, 1–8.
Fagan, M. L. (2007). Introducing competition into natural monopoly industries: An evaluation Preston, J. (2002). Railtrack - problems, solutions and absolutions. In 34th annual UTSG
of mandated access to Australian freight railroads. Report RPP-2007-05. Cambridge, conference. Scotland, January 2002: Edinburgh.
Massachusetts: John F. Kennedy School of Government, Harvard University. Railway Gazette International. (1984). Sector management (article series). April 1984.
Finger, M., & Montero, J. (2020). Handbook on railway regulation: Concepts and practice. Railway Gazette International. (2002). Freight trains running on Alameda Corridor. May
Cheltenham, England: Edward Elgar Publishing. 2002.
Fowkes, A. S., & Nash, C. A. (2004). Rail privatisation in Britain - lessons for the rail Railway Gazette International. (2008). Pilbara lines declared open access. December 2008.
freight industry. In ECMT Round Table 125. Leeds, England. Railway Gazette International. (2009). DB torpedoes a freight-oriented network. March
Friebel, G., Guriev, S., Russell, P., Shevyakhova, E., & Anna, T. (2007). Railroad 2009.
restructuring in Russia and central and eastern Europe: One solution for all Railway Gazette International. (2010). No peace in the Pilbara. August 2010.
problems? Transport Reviews, 27(3), 251–271. Review of rail access regimes. (2018). Australia: Department of Infrastructure, Regional
Gibb, R., Lowndes, T., & Charlton, C. (1996). The privatization of British Rail. Applied Development and Cities. n.p.
Geography, 16(1), 35–51. Saakyan, Y. Z. (2020). Rail sector reforms and regulation in Russia. In In Handbook on
Gómez-Ibáñez, J. A., & de Rus, G. (2006). Competition in the railway industry: An railway regulation. Cheltenham, England: Edward Elgar Publishing.
international comparative analysis. Cheltenham, England: Edward Elgar Publishing. Sidak, J. G., & Spulber, D. F. (1997). Deregulatory takings and the regulatory contract: The
Gómez-Ibáñez, & José, A. (2004). Railroad reform: An overview of the options. In competitive transformation of network industries in the United States. Cambridge
Conference on railway reform. Madrid, Spain. University Press.
Grant, H. R. (2015). Railroaders without borders: A history of the Railroad Development Sonstegaard, M. H. (2003). Competitive access to North American rail. Transportation
Corporation. Bloomington: Indiana University Press. Quarterly, 57(4), 61–67.
Guerrero, S., Juan Argote, Carrel, A., & Mazare, P.-E. (2011). Policies to address conflicts Starrs, M. (1999). Rail reform models. In Proceedings, 13th Australasian Transport Research
between passenger and freight rail service in the US. In ASME/IEEE Joint Rail Forum (pp. 59–75). Perth, Western Australia: Australasian Transport Research
Conference, 54594 pp. 443–452). Colorado: Pueblo. Forum.
Haydock, D. (2013). Want to set up a railway undertaking? Today’s Railways. July 2013. Talebian, M., Savelsbergh, M., & Moffiet, C. (2016). A new rail access charging policy:
Hope, R. (2002). Accidents raise fears about Britain’s fragmented railway. Japan Railway Hunter Valley coal chain case study. Transport Policy, 46, 101–108.
& Transport Review, 33, 32–40. December. Thomas, D. (2006). Boxes galore. Latin Tracks. July 2006.
Hylén, B. (2001). Access to the rail network in some European countries. Report 47a-2001. Thomas, D. (2007). Privatization? Yes, but not much. Latin Tracks, 2007.
Stockholm, Sweden: Swedish National Road and Transport Research Institute. Thompson, L. S., & Helene, S. (1998). Infrastructure separation: What have we learned so
Iacobacci, M. (2010). Shared corridors, strange bedfellows: Understanding the interface far? Railway Gazette International, 1998.
between freight and passenger rail. Ottawa, Ontario: Conference Board of Canada. Tye, W. B. (1987). Competitive access: A comparative industry approach to the essential
Irvine, K. (1987). The right lines. London: Adam Smith Institute. facility doctrine. Energy Law Journal, 8, 337–379.
Ishikawa, T., & Imashiro, M. (1998). The privatisation of Japanese National Railways. Van de Velde, D. (2018). Changing trains: Railway reform and the role of competition: The
London: Athlone Press. experience of six countries. London: Routledge.
Jack, A. C. R. (1998). Competitive access: Lessons from Great Britain - a response. Journal Vance, J. E., Jr. (1995). The North American railroad: Its origin, evolution, and geography.
of Transportation Law, Logistics and Policy, 65(4), 359–369. Baltimore, Maryland: Johns Hopkins University Press.
Jahanshahi, M. F. (1998). The US railroad industry and open access. Transport Policy, 5, Vassallo, J. M., & Fagan, M. (2007). Nature or nurture: Why do railroads carry greater
73–81. freight share in the United States than in Europe? Transportation, 34(2), 177–193.
Kogan, J., & Thompson, L. (1994). Reshaping Argentina’s railways. Japan Railway & Wilner, F. N. (1997). Railroad mergers: History, analysis, insight. Omaha, Nebraska:
Transport Review, 2, 23–29. June. Simmons-Boardman.
Koppel, O. (2006). Rail transit policy of the European Union and Estonia: Objectives and Wilner, F. N. (1998). Competitive access: Lessons from Great Britain. Journal of
outcome. Transport (Lithuanian Academy of Sciences), 21(3), 1–7. Transportation Law, Logistics and Policy, 65(2), 182–191.
Larsson, S. (2001). Separation can Be made to work - with state support. Railway Gazette Wilner, F. N. (2012). Amtrak: Past, present, and future. Omaha, Nebraska: Simmons-
International, 157(10), 701–704. Boardman.
Liu, R., Yang, F., & Chen, M. (2005). Understanding the shared operation of commuter Winston, C., Maheshri, V., & Dennis, S. M. (2011). Long-run effects of mergers: The case
rail transit and freight railroads. Journal of the Transportation Research Forum, 44(1), of US western railroads. The Journal of Law and Economics, 54(2), 275–304.
157–171. Woodburn, A. G. (2001). The changing nature of rail freight in Great Britain: The start of
Loving, R. (2006). The men who loved trains: The story of men who battled greed to save an a renaissance? Transport Reviews, 21(1), 1–13. https://doi.org/10.1080/
ailing industry. Bloomington: Indiana University Press. 014416400750059257
Lu, A. (2002). Vertical integration versus infrastructure separation for railroads: World Bank. (2008). Policy options for improving the efficiency of Uruguay’s railway sector.
Different optimums for different settings?. In 81st Annual Meeting. Washington, DC: Washington, D.C.: World Bank. Report 41884-UY.
Transportation Research Board. January 2002 https://stuff.mit.edu/afs/athena.mit. Yin-Nor, L. T. (2020). China’s railway development: A cyclical process of de-regulation
edu/course/urop/uic/www/OpenAccess.032602.pdf. and re-regulation. In Handbook on railway regulation. Chel: Edward Elgar Publishing.
Ludewig, J., & Brinckman-Salzedo, D. (2008). In European railway legislation handbook Zahariadis, N., & Allen, C. S. (1995). Ideas, networks, and policy streams: Privatization in
(2nd ed.). Hamburg, Germany: DVV Media Group. Britain and Germany. Review of Policy Research, 14(1–2), 71–98.
Marmouget, L. M. (1969). Los transportes. Montevideo, Uruguay: Editorial Nuestra Tierra. Zarembski, A. M., James, B., & Patel, P. (2012). Shared corridors, shared interests. In
Mizutani, J., & Fukuda, S. (2020). Issues on modal shift of freight from road to rail in 2011 Joint Rail Conference, 527–30. Pueblo, Colorado: American Society of
Japan: Review of rail track ownership, investment and access charges after the Mechanical Engineers Digital Collection. https://doi.org/10.1115/JRC2011-56095.