Rail Structures, Ownership and reform-UK

BRIEFING PAPER
Number CBP 7992, 27 July 2017
Rail structures, ownership By Andrew Haylen

Louise Butcher
and reform
Contents:
1. Structural options and case
studies
2. Structural reform in GB
3. Structural reform in the EU
www.parliament.uk/commons-library | intranet.parliament.uk/commons-library | papers@parliament.uk | @commonslibrary

2 Rail structures, ownership and reform
Contents
Summary 3
1. Structural options and case studies 4
1.1 Vertically integrated 5
Fully integrated, publicly-owned monopoly 5
Multiple, vertically integrated operators 8
Vertically integrated concessions 10
In-market competition between integrated rail operators 11
1.2 Intermediate models 13
Horizontally separated and devolved 13
Holding company model with mandated access 13
1.3 Vertical separation 15
1.4 Other factors at play 20
2. Structural reform in GB 23
2.1 What did the rail market look like before the 1990s? 23
2.2 Privatisation 25
2.4 McNulty Rail Value for Money Study, 2011 31
2.5 Developments post-McNulty 32
Shaw review into Network Rail, 2015-16 33
Current Government view, 2016- 36
2.6 Labour’s alternative: public ownership 36
3. Structural reform in the EU 40
3.1 The Railway Packages 40
3.2 Structural models in the EU 42
Cover page image copyright Manchester Victoria Station by Department for Transport
[cropped]
3 Commons Library Briefing, 27 July 2017
Summary
This paper explains the various ways railways are structured around the world and sets out
how rail structures in Great Britain have developed to date and prospects for the future.
In the early 1990s the Major Government initiated a radical structural transformation of
the British rail system, one that is yet to be repeated internationally (the system in
Northern Ireland is entirely separate, state-owned and state-run). Over a relatively short
period British Rail was split up into over 100 private companies. The infrastructure
ownership and management was privatised and rebranded under Railtrack, the rolling
stock companies and freight sectors were also privatised and the passenger rail operations
were let to private operators under 25 separate franchises. Over the next fifteen years
there followed further reforms, partly due to the change of government but also due to
issues following the Hatfield disaster and the collapse of Railtrack.
There are different views about how successful GB rail privatisation has been. Some argue
that it never delivered on its initial promise to introduce on-track competition. Some assert
that while private sector involvement has been beneficial, the lack of vertical integration
between track and train has been a mistake. Whereas others maintain that the whole
thing has been a failure – leading to inefficiencies, spiralling costs and high fares – and the
railway should be brought back into public ownership.
Great Britain was the only country to adopt its specific model of operation and was the
only country to completely divest itself of any public stake in passenger operations
(Sweden would be the closest European comparator). A host of other models have
emerged following the wave of international reform in the 1990s. Despite this, there is
still no consensus on the optimum structural model for the railway industry.
Indeed, logically a single model could never be suitable across all regions. The World Bank,
for example, has argued that rail restructuring is a “pragmatic search for a model that
works in specific markets and in which railway management objectives are reasonably
aligned with national policy objectives for railways”. That said, cross-country evidence
shows that there is a clear presence in almost all markets of a state-owned incumbent
railway operator and infrastructure manager, particularly if passenger demand and service
dominates the national rail network. Even in Sweden, where market opening has recently
been taken to the point of including all train services, the state continues to be a major
stakeholder in both operations and infrastructure.
Railway restructuring remains a controversial issue in Great Britain, twenty years after
privatisation was completed. The Conservative Government is broadly supportive of the
current structure, although Secretary of State Chris Grayling has said that he intends to
bring back together the operation of track and train in some form. This is not a new idea
– Sir Roy McNulty suggested it in his 2011 report on rail value for money. In contrast, the
Labour Party is committed to bringing private rail companies back into public ownership as
their franchises expire, as part of a wider renationalisation programme.
Information on other aspects of rail policy can be found on the Railways briefings page of
the Parliament website.
1. Structural options and case

studies
During the second half of the twentieth century, the most common
structure for national rail delivery was a single, publicly-owned firm
entrusted with the unified management of both infrastructure and
operations. 1 The freedom to compete in supplying rail services was
weak or absent from national rail industry structures in most countries.
Under this protective environment, during the 1970s and 1980s, many
national rail companies incurred growing financial deficits. Cantos et. al
(2011) explained that:
The main problems associated with the traditional policies for
railways were increasing losses, which were usually financed by
public subsidies, a high degree of managerial inefficiency and
business activities oriented exclusively toward production targets
rather than commercial and market targets. 2
The sector’s decline sparked a restructuring movement around the
world, with the core objectives of improving service quality and cost
reduction at the heart of these reforms. 3 Reforms were in some cases,
such as in Japan and Sweden, implemented over many years; while
others occurred in only a few years, as in Britain. There was no single
template for reform, though changes can be identified around three
core, interrelated building blocks:
1 Business organisation – the degree to which delivery institutions
are to be structured in a business-like or commercial manner,
including the option of private sector ownership;
2 Market competition – the degree to which the railway transport
services it produces are to be competitive, as between different
rail service providers; and
3 Separation (either horizontally 4 or vertically) – the degree to
which its monolithic nature should be broken down and some of
its sub-businesses be separated and decentralised. 5
Countries have typically either opted to retain infrastructure in public
hands, creating government management agencies to regulate private
train operators; or have established nominally independent but
government-owned enterprises to manage tracks and trains. In terms of
market competition, reforms have been very different all over the world
and, where introduced, have been done by directly facilitating the free
entry of new companies into the railway network; or to foster
1
It would typically operate as a department of a ministry, or a public entity with an
administrative reporting relationship to that ministry
2
Cantos et. al, Universitat de València, Quantifying the benefits from structural reforms
in railway transport markets in APEC economies, 2011, p103
3
Nash, C., “Passenger railway reform in the last 20 years – European experience
reconsidered”, Research in Transportation Economics 22, Volume 22, Issue 1, 2008,
p67
4
Institutional separation between passenger operations and freight operations; also
institutional separation between (vertically integrated) passenger operators, as
implemented in Japan
5
World Bank, Railway Reform: Toolkit for Improving Rail Sector Performance, 2011, p66
competition for the market by means of a franchising or concessions

system in which companies compete for the right to use and/or operate
the infrastructure during a certain period of time.
Because of these changes railways now come in all shapes and sizes:
vertically integrated, vertically separated, with or without competition,
public and private, passenger or freight dominated or mixed, supported
by subsidies or fully self-reliant. This section of the paper outlines these
models and their variations, as well as their broad costs and benefits
with reference to specific case studies internationally.
It should be noted from the outset that there is no model of operation
that is universally accepted as being optimal. The preferred model is
highly dependent on a country’s circumstances and its success is
dependent on a number of other local political, economic and social
factors which are discussed at the end of the section.
1.1 Vertically integrated

For a given rail service and piece of infrastructure, the term “vertically
integrated” is used to refer to the situation in which the owner of the
track infrastructure is allowed to provide the given rail service over that
infrastructure. 6 It remains the case that, outside Europe, most railways
retain the vertically integrated structure but have moved away from the
traditional model of a single, publicly-owned railway.
Fully integrated, publicly-owned monopoly
This is the traditional model of railway organisation where a single,
publicly-owned entity controls all or the vast majority of the
infrastructure facilities as well as the operational and administrative
functions. A number of countries still operate this model of railway,
most notably China and India. There are, however, variances of this
model based on the business organisation.
Many of the archetypal railways prior to the reforms of the late 1980s
and early 1990s, were run as and by public departments and
authorities which, according to the World Bank, were “ill-equipped to
compete in a tough external business environment”. 7 This is because
they were often captive to ‘bureaucratic pressures’ rather than
commercial incentives, described by the World Bank as:
• accountability measured-by-process (‘box-ticking’) rather than
results;
• vulnerability to short-term national budgeting processes that
destabilise longer-term business and investment planning;
• public service employment norms and procedures that impede
commercial operations; and
• political patronage or seniority as a basis for selecting board and
senior management, rather than merit; and other constraints. 8
6
“Structural reform in the rail industry”, OECD Journal: Competition Law and Policy
No.3, Volume 8, Issue 2, 11 October 2006, p86
7
Op cit., Railway Reform: Toolkit for Improving Rail Sector Performance, p67
8
Ibid., pp67-68
These railways can also be bound by political pressures that lead to

what might be considered perverse commercial outcomes (whatever
other rationale there might be), such as maintaining artificially cheap
fares for passengers, running trains where passenger demand is very
low, or reallocating investment to areas of greatest political visibility.
This led Russell Pittman to conclude that a broad consensus had
emerged that the traditional arrangement of a state-owned monopoly
railway was “inefficient and unworkable”. 9
To overcome this, other models of business organisation have been
developed. 10 For example, a state-owned enterprise (SOE) may be
established by a specific railway law or under a general SOE law which
specifies an enterprise’s commercial orientation, objects and freedoms,
and channels for political influence. As pointed out by the World Bank,
this form of organisation needs “shoring up” to create a more business-
like structure with the following measures:
• a professional and independent board of directors;
• merit-based management selection;
• management accountability based on short- and medium-term
business planning targets;
• creating business management structures geared to markets and
focusing on core functions;
• greater pricing freedom;
• use of internationally recognised commercial accounting and
auditing standards; and
• contractual agreements between enterprises and government for
reimbursement of public service obligations imposed by
governments. 11
Alternatively, if a government wants an arms-length relationship, it can
establish and register a state-owned company (SOC) using a formal
joint-stock company constitution drawn up according to corporate law;
this would include a board of directors (selected for reasons such as
their business skills and industry experience, plus independence from
the policy ministry) to establish and monitor the company’s direction
and strategy to enhance profitability or otherwise ensure a return on
shareholders’ funds. Arm’s length distance from government can be
achieved by vesting all or part of the shareholding in another
Department separate from say a country’s transport department and
without line-ministry political accountability. This would make it less-
likely to succumb to political pressure or avoid unpopular decisions. 12
9
Pittman, R. “Options for restructuring the state-owned monopoly railway”, Research in
Transportation Economics, vol. 20, issue 1, 2007, pp179–198; Mr Pittman is Dir. of
Economic Research at Antitrust Division, U.S. Dept. of Justice
10
It should be noted that the specific choice of corporate form are complex and will vary
from country to country based on legalities regarding asset holding, accounting
methods, taxation and transfer of staff to new entities
11
12
The principle argument in favour of having a fully integrated, publicly-

owned monopoly, and which can be applied more widely to other
integrated structures, is that the relationship between the services and
rolling stock, as well as the quality, quantity and technical characteristics
of the infrastructure, is so close that both aspects need to be planned
together. Instead of planning, operation and investment decisions being
split among multiple organisations, under a single, vertically integrated
system, “all decisions can be taken within one company by one line of
command”. 13 It is thus argued that vertical integration enables
optimisation for the system as a whole which “is difficult to achieve in a
vertically separated railway.”14
It is also sometimes argued that having a single operator creates for
economies of scale in operations. 15 However, there is good evidence
that rail systems are not subject to major economies of scale (i.e.
whether track traffic volume is carried by a single operator or several)
because of greater management complexity and loss of corporate agility
as the scale increases. 16 They do experience economies of density,
arising from declining average unit costs of additional traffic over a fixed
railway infrastructure until capacity is reached. 17 That is, whoever
provides the traffic, the more there is, the lower the unit infrastructure
cost.
A fully-integrated system may also be favourable because it is more
conducive to interchangeable ticketing. Additionally, there is evidence
that a well-planned and integrated timetable offering regular interval
services and good connections can considerably enhance traffic,
revenue, and benefits. 18 Thus having a single operator might be better
in terms of costs and attracting traffic than a fragmented passenger
network with a variety of operators. 19
As mentioned though, with a single, integrated and government-owned
operator there is often a lack of clear commercial focus and lack of
incentives to improve the quality of services. Independent regulation, in
terms of setting performance and commercial objectives, is often
proposed as a way of overcoming this lack of incentives. Yet, in the
absence of clear commercial incentives, the use of financial rewards and
penalties to incentivise good performance is, according to the OECD,
“difficult, if not impossible.” 20
13
van de Velde, D., C. Nash, A. Smith, F. Mizutani, S. Uranishi, M. Lijesen and F.
Zschoche for CER, EVES-Rail - Economic effects of Vertical Separation in the Railway
Sector, 2012, p26
14
Op cit., “Options for restructuring the state-owned monopoly railway”, pp179-198
15
The average costs decrease as the size of the company increases
16
17
Nash et.al, “Comparing three models for introduction of competition into railways,
Journal of Transport Economics and Policy”, Journal of Transport Economics and
Policy Vol. 47, No. 2, May 2013, pp191-206
18
Johnson, D, Shires, J, Nash, C, and Tyler, J, “Forecasting and appraising regular
interval timetable”, Transport Policy Volume 13, Issue 5, September 2006, pp349-66
19
Op cit., “Comparing three models for introduction of competition into railways,
Journal of Transport Economics and Policy”, pp191-206
20
Op cit., “Structural reform in the rail industry”, p96
Another objection to this type of model is that there is no intra-modal

competition. According to the OECD, “historical experience of many
OECD countries is that in the absence of any form of intra-modal
competition, the rail sector experiences low productivity, low service
quality, under-investment and lack of responsiveness to consumer
demands”. 21 The World Bank found that where competition was
present, either in or for the market, “there were significant
improvements in efficiency, quality and value-for-money.” 22
Multiple, vertically integrated operators
It is sometimes possible to retain vertical integration, and the previously
stated benefits, but to have it horizontally separated across regional
administrative units of the national railway. This tends to work best
when there are clearly separable business units with discrete geographic
focus. For example, larger countries have multiple railway markets (e.g.
heavy-haul freight in a mining region, major urban centres, and regional
networks), with each owned, managed and financed separately. This
according to the World Bank, can “sharpen market focus and
management accountability, and allow for specialised operations to be
devolved, divested, or compete with one another”. 23
Japan is the most notable example of this model when its previously
singular structure was split into six separate business units following
reforms initiated in 1987. It does, however, have a mixed ownership
structure with the three largest rail businesses having been progressively
21
Ibid., p90
22
23
Ibid., p76
privatised during the 1990s and 2000s, and a fourth more recently in
2016 (see box below).
Case Study: Japan

In response to major financial and efficiency pressures, the state agency Japan National Railway
was divided into six regional, vertically integrated passenger companies (JR Hokkaido, JR East, JR
Central, JR West, JR Kyushu, and JR Shikoku – known as the JRs) and one nationwide freight
company (JR Freight) in 1987.
Three of the JR companies (JR Central, JR East and JR West) were progressively privatised,
between 1993 and 2006, and are now privately owned, joint-stock companies listed on the
Tokyo Stock Exchange. Up to one third of the shares are held by foreign bodies. None of the
three companies receive state subsidies. The Shinkansen line, being almost insolvent, was
privatised and geographically divided. Ownership of each of the lines was distributed to the three
private operators according to geographic location; though there is still some cooperation
between these companies to provide Shinkansen services. In October 2016, all the shares of JR
Kyushu were also listed.
The state has retained ownership of JR Hokkaido and JR Shikoku but the remaining government-
owned passenger railways represent only a small share of total passenger kilometres. These
public enterprises nonetheless act as private operators and seek to earn a profit like private
companies. However, neither have been profitable and they receive public subsidies.
The six JR passenger lines own 87% of all Japanese railway track. The remaining 13% is divided
among other privately owned railways, which operate mainly as regional commuter lines. Freight
services are a secondary user and are provided by a company which does not own track
infrastructure of its own. Only in very few cases do tracks of different passenger lines overlap.
Intra-modal competition is therefore very low. Only in urban regions, such as the suburbs of
Tokyo, does intra-modal competition arise from other privately operating railways that focus on
regional transport. The main competition thus stems from inter-modal competition (i.e. road-
based transport).
The government has a role in jointly planning rail infrastructure with private operators and
approving the fares set by the JRs. The safety regulator is the Ministry of Land, Infrastructure,
Transport and Tourism.
The results of the JNR reform have, according to Kurosaki (2016), been “outstanding” because
of increasing transport volume, productivity, and sustainable management of the JRs, who have
focused on their markets and specific regional needs. Although the transport volume (passenger-
km) decreased 6% in the decade prior to JNR reform, the trend changed significantly, increasing
to 27%, in the decade after the reform. According to Kurosaki “this success can mainly be
attributed to privatisation and regional division, both of which solved the problems underlying
JNR’s failure”.
Sources: Ishida, Y. (2011) Japan, Reforming Railways: Learning from Experience, CER, 23-30 and
Kurosaki, F. (2016) Reform of the Japanese National Railways, Network Industries newsletter,
18(4), 8-11
As observed by the World Bank, “almost all private operation of

previously state-owned railway services has improved market and
commercial performance”. 24 Despite this, railway network privatisation
24
Ibid., p71
or concessions have proven more daunting and less attractive as a

public policy choice in countries where national railways have a strong
passenger base. In Canada, Great Britain, New Zealand, and parts of
Australia, some or all main-line railway infrastructure was transferred to
full private ownership. Since then, Great Britain and New Zealand have
essentially brought railway infrastructure back into public ownership,
although train operations are still in the hands of private companies. 25
Vertically integrated concessions
For a railway that is mostly or solely a passenger operation, the choice
for creating competition and enabling private sector participation in
restructuring would seem to be between:
• auctioning a concession 26 for the integrated railway (that is, for
infrastructure and train operations performed by the same
franchisee); and
• auctioning a concession for train operations while making other
arrangements for the infrastructure (as is the case in Great
Britain). 27
Under a typical concession contract the state maintains ownership of
the land under the railway while transferring most other infrastructure
and rolling stock assets and the right to operate rail services to a private
company during the contract period. Concessions are usually longer-
term arrangements which, according to the World Bank, “can take
advantage of private sector investment and commercial management
practices”. 28 Railway concessions can encompass the whole enterprise
or specific enterprise components – freight operations, commuter
services, or long-distance passenger services. As discussed further by the
World Bank:
Concession contracts that include rail infrastructure are typically
25 to 40 years to allow the concession operator to invest in long-
term assets to improve its performance. A concession contract can
include government investment in assets, such as infrastructure or
passenger rolling stock. Infrastructure concessions are exclusive –
the concession operator has the exclusive right to invest, maintain,
and operate the infrastructure. Sometimes concessions can also
allow operating exclusivity, or they can require the concession
operator to provide access to other train operators providing
specific transport services (passenger, freight, or both). 29
Fully integrated concession arrangements, particularly for passenger rail
services, have not been widespread internationally, 30 particularly in
Europe where current regulations require accounting separation
between the infrastructure manager on the one hand and between
25
Ibid., p71
26
A concession is a contract between a government owner and private parties to
provide some agreed rail-related services
27
Op cit., “Options for restructuring the state-owned monopoly railway”, pp179–198
28
29
Ibid., p201
30
Gangwar, R. et. al. “Framework for structuring public private partnerships in
railways”, Case Studies on Transport Policy, Volume 3, Issue 3, September 2015,
pp295-303
freight and passenger services on the other. 31 Where it has been done,
it has seen “generally positive results”. 32 Latin America and Africa have
led the way internationally in railway privatisation through concessions.
Eight Latin American countries – Argentina, Bolivia, Brazil, Chile, Costa
Rica, Guatemala, Mexico and Peru – developed private rail concessions
to one extent or another in the 1990s. 33 In assessing the effects of
concession arrangements in Latin America, the World Bank made the
following observations:
Now over a decade since rail concessioning in Latin America
began, the overall assessment of its results is positive, particularly
for freight railways. Railway traffic volumes have climbed, with
some improvements in surface transport market share. Although
numerous data problems exist, measures of productive efficiency
almost uniformly show post-concession improvements in cargo
transport. Effects on rail rates and service levels have generally
received positive reviews. Evidence is less extensive for passenger
services, mostly because concessioning was largely limited to
commuter services in Argentina and Brazil and because such
concessions must be evaluated in terms of complex subsidy and
regulated pricing regimes, rather than as market-based private
enterprises. Railway concessions have not revived uneconomic
intercity passenger services, nor has there been much effort to do
so. 34
In-market competition between integrated rail

operators
Conventional “in-the-market” competition between integrated rail
companies is possible under certain circumstances. For example,
competition for rail services over an origin-destination pair is possible
when the origin and destination happen to be connected by two
different rail routes, over infrastructure belonging to two competing rail
companies. This is sometimes called competition over “parallel tracks”. 35
31
The Netherlands procures its operations and infrastructure by way of concessions, but
they are vertically separated, for more information, see: Vosman, Q, Dutch
concessions: the ministry asserts control, March 2015
32
33
Martin, B. for the International Transport Workers’ Federation, Railway privatisation
through concessions – the origins and effects of the experience in Latin America,
2002, p2
34
World Bank, Results of Railway Privatization in Latin America, Transport Papers, TP-6,
2005, p.vii
35
Restructuring to promote competition between vertically-integrated rail

companies is more likely to be effective for freight, rather than
passenger-dominated railways. This is because passengers (especially
business travellers) tend to be more time sensitive and less able to
substitute for alternative origin-destination pairs. 36 Even in freight-
dominated markets, it may be difficult to restructure the network to
favour integrated competition.
Case Study: North America

In North America it has been possible to sustain competition in-the-market between
vertically-integrated railroads, but mostly in the freight context.
This type of competition can be important for freight transport between city pairs and for
freight transport involving a sea leg – in this case two railroads serving the same origin but
different ports will compete, to a degree, with each other. This approach has, at least in
North America, been able to deliver a degree of competition with relatively little regulatory
intervention.
Perhaps the clearest example of competition over parallel tracks can be found in Canada.
Canada has two major transcontinental railway systems, which are owned and operated by
two separate private companies, the Canadian National Railway (CN) and the Canadian
Pacific Railway (CP). Both cover the entire southern region of Canada from the Atlantic to
the Pacific. Most of the major Canadian cities, including Vancouver, Calgary, Edmonton,
Saskatoon, Regina, Winnipeg, Toronto, Montreal and Quebec, are served by both railways.
However, nation-wide passenger rail services are provided by VIA Rail Canada, a Crown
corporation (State-owned company providing commercial services), over the networks that
are owned and operated by CN and CP.
The United States has seven privately-owned, vertically integrated freight railways. There is a
substantial amount of literature from the US that the degree of competition between
competing vertically-integrated US railroad companies has a direct impact on rail freight
rates. As might be expected, studies have shown that the wider the choice of railroads
faced by any one shipper, the lower the freight rates. Intercity passenger rail services are
provided by the publicly-owned railway corporation Amtrak, which operates over the tracks
of the freight rail networks. Amtrak also owns some of its own track in the North East.
In order to have effective competition between railroads, different rail paths must be owned
and operated by different railroad companies. Careful restructuring of an integrated railroad
can increase the number of alternative rail paths over major traffic routes. In Mexico, for
example, the national railroad was split into three railroads. All three railroads serve Mexico
City. The two northern railways both serve ports on the Gulf of Mexico and on the Pacific
and both extend as far as the border with the US to the north.
Sources: OECD (2005) Structural Reform in the Rail Industry: Mexico and United States
Chapters, Policy Roundtables and Clausen, U, and Voll, R. (2013) A comparison of North
American and European railway systems, European Transport Research Review, 5(3), 129-33
Firstly, competition over parallel track requires the existence of two or

more train paths over the major traffic routes. In a country such as the
UK, or across Europe more generally, land-use density is quite high
which makes the reservation and acquisition of land for rail expensive.
Even where parallel tracks exist, separation of those tracks into
competing companies may reduce efficiency and increase
fragmentation, potentially undermining the ability of companies to offer
a single end-to-end service. According to Pittman writing in 2007 there
were probably only three countries in the world which were “obvious
candidates for a restructuring plan that would create multiple vertically
36
Ibid., p104
integrated railways competing among themselves in parallel fashion and

at common points: Russia, China, and India”. 37
1.2 Intermediate models

Horizontally separated and devolved
An alternative model is to horizontally separate rail operations. This can
facilitate policy decentralisation and involve devolution to sub-national
government authorities the funding responsibility for any separable
regional or suburban rail operations. This option gives responsibility and
accountability to those communities with the greatest stake in providing
services and finding the resources to sustain them. 38 There are several
variants of this option, which may or may not be vertically integrated:
1 Local authority may offer a concession or franchise through
competitive bidding for delivery of regional or suburban services.
In principle, the national public company might compete with
private train operators for the concession, which could include
local rail infrastructure. This type of train operating concession
would typically be re-bid periodically to ensure competition;
2 Train services can be divested to the local authority and run on the
centrally-owned network under network access agreements; or
3 Both the local network (if it is reasonably separable) and the train
services can be devolved to the local authority.
It should be noted that this option only works if sub-national
governments have the financial and administrative capacity to fulfil the
functions. A number of European countries have adopted this form of
governance and business organisation, most notably in Germany where
franchises are handled by thirty-three regional authorities. The most
prominent example of devolved services in Great Britain are those in
London, in which Transport for London, through its subsidiary London
Underground Limited, is responsible for operating the London
Underground train network and owns (in whole or in part) more than
250 stations. It also manages the London Overground network,
although Network Rail has responsibility for managing the track
infrastructure for that network.
Holding company model with mandated access
Under the holding company model, 39 infrastructure manager and
operator are distinct companies, but both are part of a same holding, or
somehow controlled by the same ‘parent’ company. This may take the
form of:
1 two separate state-owned companies (Spain, France, the
Netherlands);
37
Op cit., “Options for restructuring the state-owned monopoly railway”, p189
38
39
Internal separation of an infrastructure division from train operating divisions within a
railway company, or as companies within a holding structure, is not vertical
separation but a means of managing vertical integration
2 separate companies within a state-owned group of companies

(Germany, Italy); or
3 divisions of a single state-owned company (Switzerland). 40
It is argued that the costs may be lower under this model than a fully
vertically separated structure and that they may, more widely, be
reduced due to the avoidance of misaligned incentives because
transactions remain between parties whose ultimate interest is the
profitability of the company as a whole. 41 A problem with separated
structures, in that each company will seek to optimise its own
operations. 42 Mizutani et. al (2014) found that the holding company
model does reduce costs compared to vertical separation but this effect
is small (around 5%). 43
This model typically includes ‘mandated access’, where the
owner/operator of the infrastructure is required to provide access to the
infrastructure to an independent train operator at regulated terms and
conditions for the purpose of providing the given rail service over that
infrastructure. 44 The primary potential benefit of allowing independent
train operators access to the track is that it enhances the scope for
competition, thus increasing the probability that an efficient and high-
quality rail service is offered.
However, the scope for competition from conventional in-the-market
new entrants may be limited by a consumers’ preference for higher-
frequency services. This preference is particularly strong on short-to-
medium distance services, operating over distances of less than 200 km.
For services up to 800 km, frequency of service is likely to be a less
important consideration for end-users; though domestic route densities
have to be at a certain level for in-market competition to be feasible. 45
Thus, the likely level of in-the-market competition that will arise in a
regime of mandated access would be modest 46 and there are very few
examples of sustained in-the-market competition between passenger
train operators. For example, in Germany (which has been one of the
most successful at introducing in-the-market competition for passenger
services) the degree of competition from in-the-market passenger
services is “marginal at best” (see box, below). 47
One of the reasons this model may fail to deliver competition is the
perpetuation of perverse incentives on the train path allocation body to
be truly independent when allocating paths to new entrants, even in
spite of recent EU Directives. For example, at present there is separation
40
Transport for Quality of Life, Rebuilding Rail, June 2012
41
Nash, C., “Passenger railway reform in the last 20 years – European experience
reconsidered”, Research in Transportation Economics Volume 22, Issue 1, 2008, p63
42
Mizutani et. Al, “Comparing the costs of vertical separation, integration, and
intermediate organisational structures in European and East Asian railways”, Journal
of Transport Economics and Policy, 2014
43
Ibid., p31
44
Op cit., “Structural reform in the rail industry” p86
45
Ibid., pp112-113
46
Ibid., p113
47
Ibid., p113
between infrastructure and operations in Italy, but both parts are

owned by a single holding company. There have been complaints to the
European Commission over refusal by the infrastructure manager to
grant access to the track. Trenitalia still largely dominates the Italian
railway and intra-modal competition is extremely limited. 48
Case Study: Germany

The state-owned, incumbent rail operator in Germany, Deutsche Bahn AG, is generally
regarded as having retained its status as a vertically integrated company but since 1994 has
taken the form of a holding company, with distinct sub-divisions at the upstream level.
Under the current model, DB Netz (the infrastructure manager) and the state-owned train
operating companies such as DB Regio (non-commercial services), DB Schenker Rail (freight),
and DB Bahn Fernverkehr (long-distance passenger services) are part of the DB Holding
Company. There is open access for both passenger and freight services. Both DB and non-
DB operators must negotiate track access with DB Netz and also acquire access to other
infrastructure facilities operated by DB companies such as DB Station and Service and DB
Energy.
Franchises are handled by thirty-three regional authorities. They often franchise their
services in several different packages: service contracts have different contractual forms and
different degrees of service specification. They also vary with respect to contract duration
and network size, ranging from single lines up to whole regional network bundles, and
between gross and net cost contracts. Transport agreements do not entail any exclusivity so
that here in principle, open access is also possible.
The Bundesnetzagentur (the regulatory body for the utility sectors such as electricity, gas,
telecommunications, and postal services) has been responsible for supervising the rail
market since 2006, in particular in order to safeguard non-discriminatory access to rail
infrastructure.
Because of the high track access charges and dominance of a vertically integrated company
in Germany, one might expect that it would have low new entries, least market growth and
potentially the least pressure on operating costs. However, there appears to be a high
number of passenger operators in Germany, including a number of foreign companies.
As discussed in the Rail Liberalisation Index, “The organisation of the railway market with
the infrastructure and operation divisions combined under the umbrella of the DB Group is
not detrimental to the positive development of the competition conditions in the market”.
Even so, DB still dominates regional rail passenger services, with 88% of train kilometres.
Although currently Deutsche Bahn remains wholly owned by the Federal Republic, a
controversial privatisation of (at least part of) the Group has been on the political agenda for
several years.
Sources: Kirchner, C. (2005). Germany. In Reforming Europe’s railways. An assessment of

progress, Brussels: CER, 80–93 and Nikitinas, V, and Dailydka, S. (2016) The Models of
Management of Railway Companies in the European Union: Holding, the German
Experience, Procedia Engineering 134, 80-88
1.3 Vertical separation

As outlined in the European Commission’s 2011 White Paper,
competition to operate rail services is seen as an important way to
improve services and lower fares:
End customers of rail transport – passengers, forwarders and
contracting authorities – would benefit from this increase in
competition through lower prices, higher quality and greater
customer orientation. This would increase the attractiveness and
48
See European Commission outline of the Italian railway
market share of rail, which is a declared goal of European and

many national transport policies. 49
However, there remains a danger of conflict of interest where the
infrastructure manager is part of a vertically integrated structure, as they
retain an incentive to restrict or deny access to competitors. 50 The
incentive to discriminate against new entrants may be eliminated by
fully institutionally and organisationally separating the infrastructure
manager and the service provider. Thus, the principle advantage of
vertical separation is that, at face value, it ensures equal terms of
competition between different operators when no operator has more
control over the infrastructure than any other.
The model in Great Britain’s is one of complete vertical separation and
privatisation of all train operators. Sweden is the other notable example
of complete vertical separation, though it differs from Britain’s in a
number of areas, most notably that the state-owned operator still has a
significant presence in the market (see box, below). Norway, Denmark,
Spain, Portugal and the Netherlands also operate in a vertically
separated framework.
Even though complete separation is perceived to be more conducive to
competition, Van de Velde et. al (2012) found that “no single structure
seems more favourable than the others in terms of promoting market
entry”, 51 adding that:
Complete vertical separation will not remove the incentive for the
infrastructure manager to favour major customers over minor
ones. It should be reiterated that we have found no systematic
pattern whereby new entry, where permitted, is less likely with a
holding company than complete separation, so it is not clear that
further measures are needed. 52
The other notable advantage of having complete vertical separation is
that it provides financial transparency and enables specialisation,
particularly in terms of private companies providing public-facing
commercial services and leaving government to maintain its role in the
financing and investment aspects of the railway. 53 It is argued that:
…transparency may in itself be a way of increasing competition,
but it may also have other advantages in terms of helping
regulators and state authorities to make efficient decisions, for
instance regarding investment. In a franchising system it may
attract bids from more companies than those who would be
willing to take responsibility for both the infrastructure and train
operations in (regional) vertically integrated franchises. 54
49
European Commission, White Paper: Roadmap to a Single European Transport Area,
COM(2011) 144 final, 2011, p9
50
51
Op cit., EVES-Rail - Economic effects of Vertical Separation in the Railway Sector, p32
52
Ibid., p34
53
Op cit., “Passenger railway reform in the last 20 years – European experience
reconsidered”, p63
54
From the point of view of technical efficiency, 55 vertical separation

clearly generates a number of costs. Transaction costs, for example, in
terms of negotiation and enforcement of contracts between the
operators and the infrastructure manager, would be avoided by a
vertically integrated railway. 56 While transaction costs are higher in
separated systems, overall they are rather small, in the region of 2-3%
of total railway costs. 57
Case Study: Sweden

In 1988, rail infrastructure was vertically and institutionally separated from train operations
in Sweden under the Transport Policy Act of 1988. This came in response to the financial
deterioration of Swedish State Railways (SJ) in the 1970s and 1980s. Through this reform,
the state took full responsibility for track infrastructure investments and maintenance, by
means of a new authority, Banverket, while SJ was restructured into a train operating
company, paying fees to run on the tracks. The decision also included a decentralisation of
responsibility for the unprofitable local and regional lines to the regional transport
authorities.
In 1996, control of scheduling and access was shifted to Banverket from SJ, and open access
for freight was imposed. SJ continued to operate all passenger services, with support for
local and regional services negotiated with local authorities. By 1998, local authorities
started to put more and more local services up for competitive franchises and, over the next
few years, SJ lost many of the competitions because of its high costs and rigid management.
Following an influx of new passenger train operators in 2000, SJ was suddenly considered to
be in a disadvantageous situation compared with its competitors. To rectify this, but also
ensure equal access to functions and services for all operators, SJ's organisational structure
as a business administration (a state-owned public utility organisation but with an included
objective to make profits) was replaced in 2001 by several state-owned companies
concentrating on specified railway businesses: The passenger division formed one company
(SJ Ltd), the freight division another (Green Cargo), real estate became Jernhusen, vehicle
maintenance turned into EuroMaint and Swemaint, etc.
The reforms in the following years focussed on modernising laws and regulations to achieve
a regulatory framework in line with European Union directives. As a consequence of the
European Commission's first railway package, new legal measures were introduced in
Sweden, regulating the access to the state's railway infrastructure and the right to run
railway services. A new regulatory body, the Swedish Rail Agency, was also established.
In 2007 SJ Ltd lost its monopoly on night trains and charter trains. These changes
constituted a first step towards a new order of competition “on the track”, based on market
principles without subsidies. In the spring of 2009 further legislation set out the step-by-step
dismantling of the remaining SJ monopoly on long-distance passenger services. The national
passenger market was opened for competition in October 2010, taking full effect from the
implementation of the new timetable in December 2011. Finally, by means of a new law on
public transport in January 2012, affecting also local and regional bus and railway services,
open access is possible anywhere, although a large part of the market is still characterised by
competitive tendering.
Source: Alexandersson, G., & Rigas, K. (2013) Rail Liberalization in Sweden, Policy
Development in a European Context. Research in Transportation Business & Management,
6, 88-98
55
Technical efficiency is the effectiveness with which a given set of inputs is used to
produce an output. A firm is said to be technically efficient if it is producing the
maximum output from the minimum quantity of inputs, such as labour, capital and
technology
56
OECD, Recent developments in rail transportation services, DAF/COMP(2013)24,
2013, p28
57
Merkert, R, A. Smith and C.A. Nash, “The measurement of transaction costs –
Evidence from European Railways”, Journal of Transport Economics and Policy, Vol.
46, No. 3, September 2012, pp349-365
It should be noted that cost increases depend on the circumstances of

the country concerned and the way in which the system is managed. 58
Interestingly, a number of studies found that for more densely used
railways, of which Great Britain can be categorised, vertical separation
costs are higher than less for densely used railways. 59 Because of this
Van de Velde (2012) concluded that:
… where traffic is dense, there is a particularly strong need to
ensure close coordination. This may suggest complete vertical
integration, a holding company structure, with an active holding
company, or hybrid arrangements (such as joint ventures)
between the infrastructure manager and the leading train
operator. 60
In addition to transaction costs, costs can be driven up by a
misalignment of incentives between the various independent actors
under vertical separation. 61 This is because each actor is submitted to a
set of incentives given by the market and/or by the regulatory context.
Each of these actors is likely to optimise its own economic position
under these constraints, rather than those of the system as a whole. 62 In
practical terms:
… infrastructure managers may have too little incentive to help
train operating companies to increase revenue, by timely
implementation of improvements in capacity and quality, and by
failing to optimise timetables, slot allocation and day to day
operations. (for instance in terms of planning and undertaking
track maintenance and renewals and recovery from delays). 63
The quantitative evidence available to assess the concrete cost
consequences of misalignment is limited. 64 The 2011 McNulty Rail Value
for Money Study concluded that the total costs of the British rail system
could be reduced by 30% by 2018/19, 65 and that the principal barriers
to efficiency and cost reduction were “fragmentation of structures and
interfaces” and “ineffective and misaligned incentives”. 66 A supporting
consultancy report concluded that “while transaction costs are
significant in absolute cash terms, they do not in themselves appear to
contribute as much to net cost and general industry outcomes as do
incentive [costs]”. 67
The misalignment of incentives can be “reconstructed” using other
mechanisms besides the basic track access charging system, which
“itself proves to be insufficient to solve all misalignment issues created
58
Nash, C, and Matthews, B. for TS and CER, European transport policy: progress and
prospects, 2009
59
60
Ibid.
61
Op cit., Recent developments in rail transportation services, p28
62
Nash et. Al, “Structural reforms in the railways: incentive misalignment and cost
implications”, Research in Transportation Economics, 48, 2014, pp16-23
63
64
Ibid., p27
65
DfT/ORR, Realising the Potential of GB Rail: Final Independent Report of the Rail Value
for Money Study, May 2011, p45
66
Ibid, p5
67
Arup & Oxera for the VfM Review Team, Review of rail cross-industry interfaces,
incentives, and structures, September 2010, ppi-ii
by separation”. 68 Additional realignment mechanisms have been

developed in various countries and these tend to move towards hybrid,
cooperative arrangements (e.g. long-term contracts, strategic
partnerships and joint ventures), which set out targets and partially
apply bonus/penalties schemes to incentivise the actors over and across
the train infrastructure divide (see box below).
Great Britain: Aligning the incentives

From the beginnings of its rail reform programme in the early 1990s, Great Britain paid more
attention to the alignment of incentives than most other countries. In particular it developed
a sophisticated system of track access charges differentiated to reflect the different wear and
tear imposed by several hundred different types of vehicles. Later a ‘congestion charge’ was
added to reflect the greater potential for performance degradation (delays etc.) when lines
are operated close to capacity. There is also a performance regime in which operators are
compensated for delays attributed to the infrastructure manager or other operators, and in
turn pay for the delays they cause. These charges are reviewed and can be amended as part
of the Office of Rail and Road’s Periodic Review process, which are then transposed onto
Network Rail’s Control Periods.
The Adam Smith Institute described these charges as “a bewildering series of money transfers
[going] round the system taking in TOCs, Network Rail, the ROSCOs, the Department for
Transport (DfT), ORR and the Treasury inter alia”. McNulty considered that there were major
failings in the correct alignment of incentives. In particular, he concluded that operators had
no incentive to assist the infrastructure manager in reducing total costs, since only franchisees
pay more for marginal costs and they were protected from changes in track access charges
under the terms of their franchise agreements. He also stated that the infrastructure manager
(Network Rail) had inadequate incentives to assist operators to improve performance and
revenue.
McNulty concluded that “…industry relationships are based on contracts rather than
partnerships [which] can lead to inflexibility and confrontation [and] slow and ineffective
decision-making.” The Adam Smith Institute commented that the highly complex series of
legal contracts has “engrained bureaucracy within the railway industry”. The Transport Select
Committee also concluded in its 2017 Rail Franchising report that “the relationship between
Network Rail and operators is not as coordinated as it should be”, and that a major reason
for this is the misalignment between “the control periods and franchising schedules”, which
determine the timing of aforementioned incentive schemes.
According to Van de Velde et. al (2012), “the most effective alliance in terms of alignment of
incentives is the formation of a joint venture”. McNulty looked at case studies which
estimated that a joint venture might achieve substantial cost savings of up to 20%. Alliancing
has been tested in Britain but to date the results have been mixed, with the Transport Select
Committee in the same report commenting that “while [alliancing has been] a step toward
greater alignment between the operator and Network Rail, [it] has not achieved the desired
benefits that were initially envisaged for this programme”.
However, it is acknowledged that as “near-impossible” to design a track

access charging system that simultaneously provides for non-
discrimination, appropriate incentives for efficient development of the
network and appropriate incentives for its use. 69 Indeed, Van de Velde
et. al (2012) concluded that “whether the resulting set of mechanisms
will lead to a similar level of performance to what in principle can be
achieved in [integrated] regimes (e.g. Japan) is doubtful”. 70
There are other planning and coordination conflicts between operators
and the infrastructure manager that result from vertical separation,
particularly in terms of managing network changes, determining access
68
69
Ibid., p12
70
Ibid., p29
and timetables, as well as managing delays and disruption. These issues

can become even more problematic when traffic is dense. For example:
…on network changes, it is necessary to reach an agreement
regarding what investments are necessary, the financing of
investment which benefits more than one stakeholder, and how it
is to be undertaken. Similarly, regarding access and timetabling,
the key problem is to achieve the best possible compromise
between the requirements of the different operators, whilst
levying charges which provide appropriate incentives and raise the
necessary amount of revenue. Another area which might be
added is safety, where the allocation of responsibilities must be
clear and a balance is needed between costs and effectiveness. It
may be that an independent infrastructure manager has an
incentive to impose excessive safety requirements, where the costs
will be borne by third parties, such as in the introduction of new
rolling stock. 71
Despite the wide range of experience, the appropriate role of vertical
separation in the overall reform of the rail industry is not yet clear. 72
There does not seem to be a “one-size-fits-all” solution in term of
separation 73 and “there seems to be no evidence that vertical separation
is unconditionally superior or inferior to other structures”. 74 The success
of separation depends on the interaction with other reforms and the
sequencing of the reforms. Ultimately the decision maker will be making
the assessment that the benefits from competition outweigh the costs
involved with vertical separation.
1.4 Other factors at play

There remains a great deal of uncertainty about which structural
reforms work best and under which circumstances. There are
quantitative limitations in defining and measuring the “success” of any
structural rail reform, 75 and the relative success or failure of reform is
influenced, not only by the reforms themselves, but by a number of
other different economic, industry, social and political factors.
For example, there might be two similar countries in terms of
population, rail passenger and freight split, with near identical structural
frameworks. They may both undergo the same structural reform
process, implemented over the same period of time. Yet, there may be
71
Op cit., “Passenger railway reform in the last 20 years – European experience
reconsidered”, p63
72
Moreno, J, and Huertas, V., “Do reforms improve efficiency in European railways
sector? An empirical analysis with non-parametric approaches”, International Journal
of Recent Engineering Research and Development, Volume No. 02 – Issue No. 02,
2014, pp37-50
73
Op cit., “Structural reforms in the railways: incentive misalignment and cost
implications”, pp16-23
74
75
According to Nash et. al (2013), the existing evidence is unclear on the impact of rail
reform. Perhaps this is not surprising. Econometric studies suffer from the small
sample of countries to deal with, the significant differences in the way in which
reform has been carried out among them, and that few countries have seen
significant levels of competition, particularly in the passenger market. There are also
serious problems in ensuring data comparability
differences between these two countries in terms economic growth, 76

the cost and quality of the infrastructure and in the case of a vertically
separated railway, the ongoing relationship between operators and the
infrastructure manager. These and other differences can translate into
very different rail performance metrics despite undergoing broadly the
same structural reform.
An example of this is Japan, where despite the perceived success of its
rail reforms, rail passenger growth has remained relatively static. This
has in a large part been influenced by wider economic factors such as
stagnant population growth and an embedded social preference for car
use. Similarly Britain’s reforms, which are arguably viewed less
favourably than the Japanese reforms, have seen significant growth in
passenger numbers, though it is generally recognised that the major
causes for this growth are exogenous from the reform process. 77
Support by government for investment in infrastructure is also
important, given how poor infrastructure affects service quality,
reliability and, ultimately, the competitiveness of the whole rail sector. 78
This problem of identifying the full effect of reforms on rail sector
performance is explained by Alexandersson and Rigas (2013) with
reference to the reforms in Sweden:
Although the Swedish case, serving as an example of a liberalised
market, shows some impressive gains in terms of passenger
development (including market share) and to a lesser extent also
for freight, it is very difficult to link these improvements primarily
to vertical separation and/or market opening. It is likely that there
is a contribution from these policies that should not be neglected,
but one must also consider the simultaneous increase in
investments and the ambitions of regional authorities to expand
their networks (adding to the supply of relatively cheap
transport). 79
With respect to measuring whether structural rail reform has led to an
increase in rail’s share of the transport mode split, Holvad (2017) adds:
Overall, it is difficult to demonstrate a direct linkage between rail
regulatory reform initiatives and the rail modal share in the
passenger transport market. As such, the competitiveness of rail
vis-a-vis other modes as reflected in its modal share would be
determined by a range of influencing factors some of which are
within the domain of public policy, while others are linked to
railway industry practices as well as preferences and constraints
faced by individuals. This implies also that there would be limits to
how much can be expected to be delivered from railway
76
Wardman (2006) found economic growth to be particularly important, as well as
rising road congestion
77
Smith, A. and Nash, C. (2014), Rail Efficiency: Cost Research and its Implications for
Policy, International Transport Forum Discussion Papers, 2014/22, p15
78
Drew, J. and Nash, C.A., Vertical separation of railway infrastructure - does it always
make sense?, Institute for Transport Studies, University of Leeds, Working Paper
594, 2011, p2
79
Alexandersson, G., & Rigas, K., “Rail liberalisation in Sweden: Policy development in a
European context”, Research in Transportation Business & Management, Vol. 6,
2013, pp88-98
regulatory reform measures only as other connected measures

would need to be place as well. 80
There are other measurable outputs of rail sector performance,
including operating efficiencies, operating and capital expenditure, rail
fares, day-to-day service quality and punctuality and dependence on
government subsidies, just to name a few. There is a risk that such
benchmarks are viewed as authoritative metrics of the relative success
or failure of the structural reform. Any reforms have to be viewed in the
full context of the various economic, structural, political, and social
factors in a particular country.
80
Holvad, T. Universita Degli Studi Di Milano, Dipartimento Di Economia, Market
structure and state involvement: Passenger railways in Europe, , Working Paper
No.4, January 2017, p24
2. Structural reform in GB
Until 1994, the rail industry in Great Britain was organised in the form
of a single, publicly owned and integrated company, British Rail (BR),
which managed the infrastructure and provided passenger and freight
services throughout the country. 81
The fundamental objective of advocates of privatisation was to free the
nationalised industries from bureaucracy and political intervention and
to replace these forces with the disciplines of the market, in the
expectation that this would lead to greater efficiency, lower unit costs
and a better allocation of resources. The corollary of this, usually
welcomed by management, was that enterprises were freed from
constraints on investment and on funding imposed as part of public
expenditure controls.
Opponents of privatisation argued that it was primarily a convenient
way of abandoning the traditional social duties of the public enterprise,
and of renegotiating, to the disadvantage of employees, their terms of
employment. Further, the goals of reduced bureaucracy, greater
efficiency, lower unit costs and better allocation of resources would not
necessarily result from privatisation and could be achieved by other
means. 82
After much debate about options, including the possibility of complete
open access on the rail network, the Major Government opted for the
structure of a regulated monopoly infrastructure provider with
competitive operators using it. As a consequence BR was broken up into
around a hundred different companies and privatised. Many consider
this to be most radical reform of any national railway 83 and “on a scale
never contemplated anywhere else in the world before or since”. 84
2.1 What did the rail market look like before

the 1990s?
British Rail (BR) was a public corporation established under section 1 of
the Transport Act 1962 as a successor to the rail and shipping activities
of the British Transport Commission. 85 The British Rail Board operated
passenger and freight services within Great Britain and was almost
81
Although it should be noted that British Rail had its borders too: between regions,
sectors and divisions; for full discussion of British Rail’s structure and progress up to
privatisation see: Gourvish, T. (2002) British Rail 1974-1997 From Integration to
Privatisation; see also HC Library briefing paper SN1157
82
Transport Committee, Financing of Rail Services (third report of session 1986-87), HC
383, 13 May 1987, para 232
83
e.g., op cit., “Comparing the costs of vertical separation, integration, and
intermediate organisational structures in European and East Asian railways”, p10
84
Smith, A., Liberalisation of passenger services, Case Study – Britain, Centre on
Regulation in Europe, 6 December 2016, p3
85
the BTC was established under the Transport Act 1947 to provide "an efficient,
adequate, economical and properly integrated system of public inland transport and
port facilities within Great Britain for passengers and goods", excluding transport by
air and came into operation on 1 January 1948 when the various interests in
shipping, railways, hotels and road transport that were nationalised
entirely vertically integrated, that is to say it owned its own trains,

infrastructure and carried out almost all track and train maintenance
itself. Under state control, the railways were expected to run
economically but also to cater for not always clearly defined social
needs. In compliance with European rules it also operated passenger
services on a non-commercial basis where so directed by the Secretary
of State. 86 The Secretary of State paid BR a grant to cover the cost of
providing any loss-making services.
The Secretary of State for Transport exercised a number of controls over
the BR Board, such as appointment of the Chairman and Board
members; setting the external financing limit (EFL) and the public service
obligation (PSO) grant; determining the Board’s investment allocation
and approval of major investment projects; setting borrowing limits;
having the final say on station closures and service cancellations; setting
financial targets; and agreeing the business plans. He also approved
changes in the organisation of the Board; extensions of BR's fields of
activities; and the disposal of businesses and assets. Many of the powers
were procedural, or related to financial propriety. The Secretary of State
also had various safety and regulatory powers under railways
legislation. 87
During the 1980s BR was encouraged to develop greater commercial
awareness. It did this by reorganising its business and objectives,
privatising its ‘non-core’ activities and encouraging private sector
involvement. 88 These initiatives contributed to BR’s major organisational
transformation in the decade preceding its privatisation. The
Conservative Government's determination to reduce the level of subsidy
flowing to the public sector railway stimulated a drive within BR for
substantial efficiency improvements. Central to these reforms was the
partial replacement of the existing structure based on regions by one
formed around distinct rail business sectors, each with managers
meeting objectives in terms of marketing, cost allocation and
investment decisions. A further reform came in 1991 with the launch of
‘Organising for Quality’. Completed by April 1992, this initiative finally
abolished the old regions (which had continued to physically run the
trains) and defined separate profit centres within each of the business
sectors.
In some respects the substantial reorganisation and the accompanying
efficiency improvements undertaken by BR in its final decade generated
the conditions that made privatisation a more viable policy. The new
business sectors formed at least part of the basis for the privatised
railway and showed that the system could be operated as a series of
relatively independent components, rather than as a monolithic
86
called ‘public service obligations’ (PSOs), made under State Aid rules; more
information on EU rail policy can be found in HC Library briefing paper SN184
87
powers derived mainly from the Transport Act 1962, Transport Act 1968, Railways Act
1974, Transport Act 1980 and Transport Act 1981
88
for further information see HC Library briefing paper SN1157
structure. The reorganisation also led to a marked reduction in the PSO

payments paid direct to BR in support of loss-making services. 89
As a consequence of all these changes, by the late 1980s BR could be
said to be doing well, but after the good times of 1980s, BR's finances
collapsed in the early 1990s largely because of increased expenditure on
safety following the 1988 Clapham rail crash, the costs of improving
lines to the Channel Tunnel and perhaps above all because of the
recession, which saw usage fall by ten per cent. 90
2.2 Privatisation
Privatisation of the railways had been discussed intermittently since the
1960s and there was apparently an idea to float the Southern Region of
BR after the 1979 election, but this never went anywhere. Mrs Thatcher
finally agreed to the policy in late 1990, shortly before she left office.
John Major was initially more enthusiastic, though the driving force
behind the proposal appears to have been the Treasury. There was later
confusion between various models of privatisation proposed by the
Treasury and alternative suggestions or ‘hints’ from the then Transport
Secretary, Malcolm Rifkind and the Prime Minister. 91
The 1992 Queen’s Speech promised that “legislation will be introduced
to enable the private sector to operate rail services”. 92 On 7 May a
‘paving’ Bill was introduced to confer on the BR Board “powers to
participate in the implementation of proposals for the transfer of their
commercial activities to the private sector and proposals for the
establishment of new arrangements with respect to their other
functions”. 93
In July 1992 the Government published its White Paper outlining
proposals for privatising British Rail (BR). 94 The core of the Government's
proposals was the greater involvement of the private sector in the
running of the railways through the sale of some of the BR businesses
and the progressive contracting out of the management of passenger
services. The principal organisational means of achieving these
objectives was the separation of responsibilities for track and
operations. The Railways Bill 1992-93 was published on 22 January
1993 and had its Second Reading on 2 February. 95 The Bill was
essentially an enabling measure, leaving a large degree of discretion to
the Secretary of State, the regulator, and Franchising Director. Further,
89
Transport Committee, Railway finances (fourth report of session 1994-95), HC 206, 5
July 1995, para 9
90
ibid., paras 16-20
91
Wolmar, Christian, Broken Rails: How Privatisation wrecked Britain’s railways (2002),
chapter 4
92
HC Deb 6 May 1992, c51
93
the British Coal and British Rail (Transfer Proposals) Act 1993 received Royal Assent on
18 January 1993; the Secretary of State explained the purpose of the Bill at Second
Reading, see: HC Deb 18 May 1992, cc22-35
94
DoT, New opportunities for the railways: the privatisation of British Rail, Cm 2012, July
1992; this was accompanied by a statement to the House on 14 July, see: HC Deb 14
July 1992, cc971-72; there were further debates on the White Paper in October 1992
(HC Deb 29 October 1992, cc1160-1222) and on an Opposition motion opposing
privatisation in January 1993 (HC Deb 12 January 1993, cc771-869)
95
HC Deb 2 February 1993, cc156-255
there was no mention anywhere in the Bill of Railtrack – an indication of

the sweeping nature of the powers contained in Part II, which enabled
the Secretary of State to restructure BR in any way he thought fit.
The proposals excited much comment. The Transport Committee
published a report on the subject in April 1993 which concluded:
It is clear that in terms of previous international railways
experience, the form of privatisation adopted by the UK
Government is both novel and experimental (in the sense of being
untested). It is true that some elements of the Government's
proposals have been put into practice or contemplated in various
parts of the world. Yet in no country with a rail system of
comparable size and density of use is there an example, either in
operation or even under consideration, of a complete scheme
such as that contained in the Railways Bill. This does not of itself
mean that it cannot succeed. To take that argument to its logical
conclusion would mean that no innovation ever took place. What
it does mean, however, is that because of the lack of previous
experience to draw upon, the risk that something could go badly
wrong is that much higher. To put it another way, the system of
railway operation proposed by the Government probably can
work, but, in the words of one witness, it may need to be made
to work.
The onus lies firmly on the Government to demonstrate that its
plans will provide a better service to the travelling public. If all the
Government's assumptions are correct about such matter as:
1. the prospects for investment;
2. the practicality of the relationship between Railtrack and
operating companies;
3. the response of the private sector to the new opportunities
on offer; and
4. the feasibility of combining open access with franchising
then there may be the potential for an improved railway system.
Whether the Government is right in these assumptions is a matter
of political judgement. The final verdict will rest with rail users. 96
Railways Act 1993

The Railways Act 1993 received Royal Assent on 5 November 1993.
Part I set out the respective powers and duties of the Secretary of State,
the regulator and the Franchising Director. It also specified line closure
procedures, conferred upon the High Court powers to issue Railway
Administration Orders in the event of default by independent railway
operators, and established machinery for consumer representation. Part
II related to the Secretary of State's powers to direct BR to reorganise
itself and to form companies for various purposes, including franchising
and disposal. Part III of the Act contained miscellaneous provisions
relating to safety, railway heritage, the British Transport Police,
pensions, freight, financial assistance to BR staff in the preparation of
management buy-outs or franchise bids, and other matters. Many of
the principal changes were brought into effect on 1 April 1994.
96
Transport Committee, The Future of the Railways in the Light of the Government's White
Paper Proposals (second report of session 1992-93), HC 246, April 1993, paras 523-
524
The legislation radically changed the structure of the railway industry by

separating the responsibility for infrastructure and passenger service
operations. BR was divided into a body known as Railtrack on the one
hand, and a residual BR operating company to run all the other services
until they were sold or franchised. The Office of Rail and Road (ORR)
was set up to oversee the charges to be levied by Railtrack for the use of
the infrastructure. 97
2.3 What happened after privatisation?

There were a number of problems with the railways in the five years or
so after privatisation – some to do with the actual privatisation itself and
others due to secondary factors that were not directly linked to the
structure of the industry after 1993. Some of these might have been
anticipated in the run up to and during privatisation and dealt with but
for one reason or another were not.
For example, it is debatable as to whether there was sufficient
discussion about the structure of the privatisation – Railtrack was not
mentioned in the 1993 Act and Ministers were not clear what form the
new infrastructure company would take; the requirement for
competition and open access was never seriously tackled; and the
question of continued public subsidy was largely overlooked. The desire
to complete the privatisation before the 1997 General Election,
combined with Labour being so far ahead in the polls and committed to
renationalisation, led to confusion and issues with methods of sale of
the various parts of the industry. 98
Privatisation itself involved breaking up BR into around 100 component
companies, introducing complexity, and the powers available to the
regulator were often unwieldy. The relations between the component
parts of the industry were also perhaps over legalistic to the point of
engendering confrontation and antagonism rather than co-ordination
and co-operation.
The privatised railway itself lacked a long term planning body with the
sort of strategic duties BR had and Railtrack arguably lacked knowledge
of its asset base, contributing to later problems (e.g. dealing with the
aftermath of the Hatfield crash). At the beginning there was a lack of
97
this is the rail regulator, it changed its name on 1 April 2015 to include reference to
its new role regarding Highways England; for further information on the history of
ORR see HC Library briefing paper SN2071
98
Labour’s May 1996 policy paper Consensus for Change: Labour’s Transport Policy for
the 21st Century, said: “Labour is committed to a publicly owned and publicly
accountable railway […] An incoming Labour government will use all the levers at its
disposal to halt the damage of privatisation, reintegrate the network and generate
higher levels of investment. On coming to office Labour will set in place a structured
programme to return the railways to an integrated whole. We will build on what is
left of British Rail to create a renewed publicly owned company which will be
charged with reintegrating the network, protecting the public interest and
organising public private partnerships to increase investment […] We will end the
franchising process [and] give the powers of the franchising director to British Rail,
who will supervise the franchises which have already been let and resume control of
these lines when the contracts expire”. This was significantly watered down in the
Labour Manifesto for 1997 and the promise to renationalise was removed
clarity about the roles of Franchising Director and Rail Regulator, leading
to confusion and there were shortcomings in the range of ‘network
benefits (e.g. ticket sales, National Rail Enquiries and passenger
compensation). More fundamentally there was evidence of poor
management particularly at Railtrack, which appeared to be inept at
estimating the costs of large projects and managing its subcontractors
while failing to reconcile its public interest objectives with the interest of
its shareholders to maximise profits. 99
Railtrack
Railtrack was set up on 1 April 1994 under the 1993 Act to manage the rail infrastructure
(track, stations, etc.). It was sold to the private sector in May 1996. Railtrack's main sources of
revenue were the charges it levied on train operators for track access and the lease income it
received for stations and depots. Until 2001 Railtrack did not receive direct revenue subsidy
from the Government although it was indirectly dependent on the significant amount of
public sector support received by the train and freight operating companies. Railtrack plc was
put into administration on 7 October 2001 and came out of it on 1 October 2002. Network
Rail took over many of its responsibilities on 3 October. 100
Railtrack’s difficulties came about for a variety of reasons. Those offered at the time included:
lack of awareness about poor asset condition; privatisation was overly ambitious and
financially risky; the complex contractual base of the privatisation; the extent and nature of
Government and regulatory intervention; and poor management. 101
All of this meant that by October 2001 Railtrack was insolvent, even if it was not bankrupt,
and plans for some sort of restructure of Railtrack had been around for some months. The
four most widely talked about at the time were:
• renationalisation (consistently ruled out by Ministers as too expensive); 102
• restructuring, whereby the Government would take an equity stake in the company,
suspend the regulator and fund the company directly for a few years until it was in a
better position (this was the favoured option of Railtrack’s directors, referred to as
‘Project Rainbow’);
• converting the company into a not-for-profit entity of some sort (‘Project Ariel’); or
• breaking up Railtrack into regional businesses.
In the event the Government chose a variant of the third option; a decision that was not
without controversy, largely to do with the way in which it chose to do it. 103
99
The Government's Response to the Environment, Transport and Regional Affairs
Committee's Report on the Proposed Strategic Rail Authority and Railway
Regulation, Cm 4024, July 1998
100
for more information on Railtrack and its administration see HC Library briefing
papers SN1224 and SN1076
101
taken from contemporary comment, articles and reports, e.g. Christian Wolmar,
Broken Rails: how privatisation wrecked Britain's railways, 2001; Lord Cullen/HSC,
The Ladbroke Grove rail inquiry: Part 2 report, September 2001; "Cabinet simply
watched as 'poll tax on wheels' went off the rails", The Times, 8 October 2001;
"Disastrous round trip was marked by failure to maintain lines", The Times, 8
October 2001; "Watchdog's severe targets skewed company's judgement", The
Times, 8 October 2001; and "Signal failure", The Sunday Times, 14 October 2001
102
because they would have had to buy the company shares at the full market value
103
for the purposes of understanding the administration, please note that Railtrack
Group plc was the company in which shareholders had shares; its main operating
subsidiary was Railtrack plc, which was responsible for the management of the
national rail infrastructure and it was this company that was placed in administration
That said, the railways in the immediate period after privatisation were
not all bad. Until the Hatfield crash in October 2000, there was strong
growth in both passenger and freight traffic; punctuality and reliability
were slightly better than in the final years of BR and safety standards
were gradually improving. More trains were running (around 1,700
more each day than before privatisation); 30 per cent more passengers
were being carried in 2000/01 compared with 1994/95; the amount of
freight lifted had increased 40 per cent in the same period; and both
punctuality and reliability improved, although performance had declined
since the high point of 1996/97. 104 Railtrack's initial share price of 380
pence rose to 1,700 pence two years after privatisation, but by 2001
the position was very different and on 1 October 2001 the share price
was 265 pence.
From 1997/98 to 1999/2000 performance was roughly stable, but it
experienced a sharp decline following the Hatfield accident in 2000. 105
Fragmentation and industry costs

A number of stakeholders have argued that the key flaw in the privatisation of the
railways was the fragmentation of the previously integrated railway (McCartney, S, and
Stittle, J, 2017). Some have argued that a degree of fragmentation is not necessarily bad
but the extent and speed of fragmentation in Britain was excessive and was something
altogether new in the history of utility privatisation (Bartle, 2004).
It is also argued that fragmentation in Britain led to a high interface costs due to a more
complex supply chain in which each actor requires a profit. For example, rolling stock is
supplied by an independent manufacturer to a rolling stock operating company, which
then leases it to a train operator. Some academics have suggested that each interface
costs 5% of payments to the industry’s suppliers at each level in the supply chain. There
are also ‘cash leakages’ as interest payments and dividends were extracted from industry.
It was estimated that between 1995/95 and 2002/03, these leakages amounted to
£5.45bn, just over half the £10.7bn subsidy paid to TOCs in that period (McCartney, S,
and Stittle, J, 2017).
Smith et al. (2009) showed that the total of infrastructure costs and passenger train
operating costs per train kilometre rose by 34% in Britain over the period 1996/7 to
2005/6, suggesting that the reforms in Britain failed to control costs. Taylor and Sloman
(2012) separately estimated “fragmentation costs” at £581 million in 2009, made up of
interface costs between the TOCs and NR (£290m), costs of NR outsourcing (£200m) and
operating margins of TOC and ROSCO sub-contractors (£91m).
McCartney and Stittle (2017) concluded that “…costs and subsidy have both increased.
Privatisation has inflated costs: more exactly, given the volume of passenger traffic, the
privatised industry has incurred higher costs than would have been incurred by British
Rail”. They also estimated that the additional aggregate costs of privatisation amounted
to more than £50bn (at 2013/14 prices) between 1997/98 and 2013/14.
The IPPR think tank wrote in 2002 that “privatisation arguably

introduced some useful innovations, notably competition in the delivery
of passenger and freight services, performance incentives linked to the
passenger's charter, transparent regulation, security of funding and
access to private finance for investment”. 106
104
Railway Forum fact sheet, Britain’s growing railways (factsheet no. 1), 30 January
2001; and DfT, Rail usage, infrastructure and performance (RAI0101), 11 December
2014
105
DfT, Rail usage, infrastructure and performance (RAI0105), 11 December 2014
106
IPPR, Getting back on track, June 2002, executive summary
Following privatisation there was a great deal of criticism of the

performance and reliability of the train companies, of Railtrack and the
amount it was investing in the infrastructure and in the safety regime.
Labour was particularly concerned about whether the Regulator had
sufficient powers to deal with the private sector companies who ran the
railways and about the supervision of public funds. With that in mind,
its 1998 transport White Paper stated that the rail industry needed an
element of stability and certainty if it was to plan its activities
effectively. 107
Particularly, the Government was concerned that there was “no focus
within the privatised industry for long term strategic planning”; that the
Franchising Director's remit was “too narrowly focused on the
passenger railway”; that too many key policy decisions on the future of
the rail industry lay in the hands of a statutorily independent regulator;
that the sanctions available to the regulatory authorities were
‘unwieldy’; and that investment was not sufficiently prioritised. 108 It
therefore proposed setting up a Strategic Rail Authority (SRA) to address
these problems, asserting that this would provide the rail industry with
the strategic leadership lacking since privatisation and the
fragmentation of the sector. It would ensure the railways were run in
the public interest, would promote their use and would ensure they
were properly integrated with other forms of transport.
The SRA would tell the TOCs what services and network benefits the
Government wanted to buy. It would ensure that the railway was
properly integrated with other forms of transport and that the railway
system was run as a network, not merely a collection of different
businesses, particularly when franchises were re-let or re-negotiated. It
would also ensure that the plans of freight operators were taken into
account in the planning of the network as a whole.
Part IV of the Transport Act 2000 gave statutory backing to the SRA
from 1 February 2001. It transferred the functions, rights and liabilities
of the Franchising Director and the residual functions, rights and
liabilities of the British Railways Board (including responsibility for the
British Transport Police) to the SRA. It set out its objectives and
functions and established its structure and procedures and the terms
and conditions of its members. The legislation also transferred the
Regulator’s responsibilities for consumer protection to the SRA and the
responsibility for railway closures to the Secretary of State.
The SRA was set up to provide leadership for the rail industry but it
lacked the power to bring together the various elements of the industry.
It did have some successes. For example, it helped to get a grip on
major projects which were running out of control, such as the West
Coast Main Line route modernisation project; it developed and
107
DETR, A new deal for transport: better for everyone, Cm 3950, July 1998, para 4.22
108
The government's response to the environment, transport and regional affairs
committee's report on the proposed strategic rail authority and rail regulation, Cm
4024, July 1998, paras 6-11; responding to: ETRA Committee, The proposed strategic
rail authority and rail regulation (third report of session 1997-98), HC 286, 18 March
1998
implemented a strategy to upgrade the electricity supply for rail services

south of the Thames; it began the process of reforming the franchising
regime; and it promoted better use of the network through its Route
Utilisation Strategies (RUS). The SRA also developed a stronger regional
focus, working to promote community rail partnerships and to promote
joint working between the different parts of the industry and to drive
up performance. 109
The problems in the rail industry, however, turned out to be greater
than realised in 2000 – there was the aftermath of the Hatfield
accident; increasing infrastructure costs; Railtrack’s entry into
administration; and the subsequent transfer of its responsibilities to
Network Rail. Consequently, the Regulator played a more pivotal role
during these years and the SRA found itself in an increasingly difficult
position. A leadership model based on influence and persuasion was not
strong enough.
As a result, the 2004 rail White Paper proposed that the SRA’s strategic
and financial responsibilities would pass to the Department for
Transport, as would its responsibility for awarding the passenger
franchises. The Secretary of State would take responsibility for setting
the national-level strategic outputs for the railway industry, in terms of
capacity and performance. Operational matters would remain the clear
responsibility of the industry. The SRA’s role in monitoring the
performance of the train companies and drawing up timetables would
be taken over by Network Rail. 110 Section 1 and Schedule 1 of the
Railways Act 2005 abolished the SRA and transferred its powers to the
relevant bodies. This took place on 8 June 2005. 111
In July 2004 the Regulator was replaced with an Office of Rail
Regulation (ORR) and in its 2004 rail White Paper the Labour
Government announced its intention to transfer safety responsibility to
the ORR. This was legislated for in the 2005 Act. The ORR became both
the economic and safety regulator for the rail industry on 1 April 2006
when it took over the safety regulation responsibilities of the HSE. These
include the power to authorise a person to investigate and make a
special report on a major incident.
2.4 McNulty Rail Value for Money Study,

2011
The Labour government established a review group in 2009, chaired by
Sir Roy McNulty, to report on rail’s costs in Britain and identify options
for improving value-for-money. In preparing the final report, the
McNulty Review commissioned L.E.K Consulting to produce the
Alternative railway structures study, 112 which identified four possible
structural options for improving value for money in the rail industry:
109
SRA press notice, “Britain’s railway ‘rehabilitated’”, 15 July 2004
110
DfT, The future of rail, Cm 6233, July 2004
111
Railways Act 2005 (Commencement No. 1) Order 2005 (SI 2005/1444)
112
It included Volumes 1, 2 and 3
• Base package: Radical franchise reform, regional accounting

separation of Network Rail (NR), and cost and revenue sharing;
• Horizontal separation: NR split into regional units, some or all
of which could be sold to alternative providers;
• Vertical integration: Competitively tendered concessions for
integrated train operators and infrastructure managers on a
regional basis; and
• Vertical alignment: Some functions would be provided by some
form of an alliance between the regional infrastructure manager
(“Regional IM”) and the dominant train operator in that region. 113
In making an assessment of these various options, L.E.K concluded that
vertical integration was the structure that achieves the best alignment of
incentives between infrastructure managers and train operators.
However, it caveated this finding, asserting that this varies on a case by
case basis and “there are many other parts of the network where
vertical integration should not be implemented”. 114 It concluded that
where part of the network was relatively self-contained, 115 had a
dominant train operator and where there was limited prospect for “in
the market” competition, “it is very likely that there would be
incremental benefit in implementing vertical integration in these
regions”. 116 L.E.K recommended “a phased roll-out whereby vertical
integration is implemented in a single region to start with and the
learnings from that region are used to inform the decision as to where
else to implement vertical integration”. 117
McNulty himself concluded that key objectives of privatisation, including
subsidy reduction, had not materialised, and highlighted key structural
barriers to efficiency. Despite accepting that privatisation had not
brought the expected benefits, McNulty dismissed the case for major
structural reform. 118
2.5 Developments post-McNulty

The Coalition Government published a rail Command Paper in 2012. It
set out the different reform options to integrate the railway, but did not
commit to any deep levels of reintegration, commenting:
Government welcomes the direction of the study’s
recommendation and considers greater alignment of incentives
for efficiency between Network Rail and train operators to be the
most pressing reform necessary to drive down costs for rail
industry. Government is committed to exploring the full menu of
options for promoting greater alignment, including options to
113
L.E.K. Consulting, Alternative Railway Structures: Final Report – Volume 1, 7 March
2011, p4
114
Ibid., p21
115
Five of NR’s existing nine operating routes are already relatively self-contained and
have a single dominant TOC, e.g. Greater Anglia, South West Trains, Kent,
Merseyrail and Scotland
116
Op cit., Alternative Railway Structures: Final Report – Volume 1, p11
117
Ibid., p11
118
Op cit., Realising the Potential of GB Rail: Final Independent Report of the Rail Value
for Money Study, pp283–286
place responsibility for train operations and infrastructure

management in an area in the same hands.
We also believe that decentralisation of certain functions within
Network Rail and the formation of alliances with train operators is
essential if we are to make Network Rail more accountable to its
train operator customers. 119
Though it did acknowledge that it would explore the potential for
vertical reintegration:
We agree that vertical integration could offer promising benefits
in the longer term. So building on alliancing and concessions, and
subject to EU legislation, we will explore how full integration on
discrete parts of the network might potentially drive further
efficiency benefits and investment, helping both track and train to
align around service delivery. This is most likely to be an option on
Network Rail routes such as Wessex and Anglia where the
majority of passenger train services are run by a single train
operator. 120
The structure of the railway in Great Britain has remained broadly the
same in the five years since the Command Paper was published, though
franchises and Network Rail have experimented with alliancing, with
mixed success. This was highlighted in the Transport Select Committee’s
February 17 report on rail franchising:
It is clear that the alliancing programme, while a step toward
greater alignment between the operator and Network Rail, has
not achieved the desired benefits that were initially envisaged for
this programme. We conclude that the programme is ultimately
limited by the misalignment between franchises and Network Rail
routes that prevents the establishment of deeper commercial
arrangements. 121
Dr Andrew Smith of the Leeds Institute for Transport Studies also
concluded that “there is limited evidence that alliances between
monopoly operators and Network Rail on individual routes have yet
brought about cost reductions”. 122
Shaw review into Network Rail, 2015-16
Further discussion of structural reform was evident in the Shaw Report
into the future shape and financing of Network Rail published in March
2016. 123 It made seven key recommendations, based in three areas:
customers, devolution and growth. In terms of devolution, one of her
two recommendations was to focus on the customer through deeper
route devolution, supported by independent regulation. This has been
on the rail industry agenda for a long time. 124 Those on the left and the
119
Department for Transport, Reforming our Railways: Putting the Customer First,
March 2012, p.41
120
Op cit., Reforming our Railways: Putting the Customer First, p.45
121
Transport Select Committee, Rail franchising (Ninth Report of Session 2016–17), HC
66, 5 February 2017, para 90
122
Op cit., Liberalisation of passenger services, Case Study – Britain, p.33
123
DfT, The future shape and financing of Network Rail: The recommendations, 16
March 2016; led by Nicola Shaw, the Chief Executive of High Speed 1
124
e.g. McNulty (op cit., Realising the Potential of GB Rail: Final Independent Report of
the Rail Value for Money Study, sections 23.1.2 and 25.2); ATOC (A new structure
for success on Britain’s railway: An ATOC position paper on industry structural
right – whatever their thoughts on whether the public or private sector

should be running the railway – support greater local accountability
and/or some sort of reintegration between infrastructure operations and
train services, split at privatisation in the 1990s. 125
NR has already taken steps in this direction. The company currently
operates an organisational structure that combines devolution to
geographical operating units known as ‘routes’ with centralised delivery
of key support functions, as shown below: 126
As Shaw explained in her scoping report:

… the routes are accountable for aspects of a number of
functions including train operations, asset management, and
maintenance as well as aspects of enhancements and renewals. It
is worth noting, however, that at this level of analysis the routes
are not solely accountable for any function, but share
accountability with the centre – where the centre is generally
responsible for overall strategy, professional standards and
conducting performance benchmarking. 127
In the 2015 Summer Budget the Government stated that devolution
within NR was “a key priority for the company’s current leadership”,
and the company’s stated ambition remained that of reducing the role
reform, March 2011, pp3-4) and the TUC (Transport for Quality of Life, Rebuilding
Rail, June 2012, pp66-67 & 69
125
“Network Rail 'too big', says Sir Richard Branson”, BBC News, 26 June 2015; and
“Regulator proposes splitting Network Rail into eight businesses”, Financial Times,
15 July 2015
126
from DfT, The future shape and financing of Network Rail: The scope, November
2015, p28
127
ibid., p29
of the centre and devolving greater responsibility to the route-level. The

Government asked NR to continue with this process of
decentralisation. 128
To support this, NR began implementing plans for a new route-based
operating model, which would provide routes with more scope in
choosing what services they need from the centre and what services are
best provided at a route level or by the wider market. 129
Shaw’s final report saw progress in this area in two ways. Firstly, on
deeper route devolution, she proposed a new model of operation for
NR, clearly showing the routes closer to their customers and more
autonomous from the direct input of the centre: 130
The second part of Shaw’s recommendation in this area was to redraw

the geographical scope of these routes to better map them onto
devolved areas. She acknowledged that an ‘ideal’ geography was
unlikely but argued that a new structure could and should:
• enable political accountability and support economic growth by
directly aligning with one body or a small number of bodies that
are responsible for transport and/or wider economic planning in
the region;
• enable effective coordination between the route and train
operators by aligning with a small number of train operators; and
• remain of a manageable size and scale for the Route CEO and top
team to effectively and efficiently operate the network. 131
128
HMT, Summer Budget 2015, HC 264, 8 July 2015, para 1.255, para 57
129
op cit., The future shape and financing of Network Rail: The scope, pp29-30
130
op cit., The future shape and financing of Network Rail: The recommendations, p50
131
ibid., p72
She also said that better-aligned local routes could leverage more
sources of local funding for enhancements (e.g. developer
contributions, local government borrowing, Business Rate Supplement,
and Business Rate retention). 132
Current Government view, 2016-
In December 2016 the Secretary of State for Transport, Chris Grayling,
announced the Government’s support for the East-West Rail scheme. 133
He said that it would be a separate regulated entity from Network Rail,
and a standalone vertically integrated railway, with private sector
finance playing a key role in funding. 134
He also indicated a desire for new franchises to implement joint
management teams, which would include representatives from both the
private train operator and Network Rail; though both entities would
continue to exist independently from one another. As indicated above,
this sort of model is better suited where there is one predominant
operator on a network and fewer interfaces with other train companies.
It is envisaged that the South Eastern and East Midlands franchises will
be the first to implement these joint management teams when they are
re-let in 2018. At present, legal and regulatory constraints mean that a
joint venture model is not possible to implement in the short-term. The
DfT is therefore likely to fall back on some form of an alliancing model,
but it is unclear how this model would differ from the Network Rail
alliance currently in place with Scotrail (and the previous alliance on
South West Trains that has now been dissolved). 135
2.6 Labour’s alternative: public ownership

The Labour Party went into the 2017 General Election on a platform to:
Bring […] our railways back into public ownership, as franchises
expire or, in other cases, with franchise reviews or break clauses.
We will introduce a Public Ownership of the Railways Bill to repeal
the Railways Act 1993 under which the Conservatives privatised
our railways […] A publicly owned railway system can be the
backbone of our plans for integrated transport. It will be built on
the platform of Network Rail, which we will retain whole, working
with the devolved administrations. We will ensure new rolling
stock is publicly owned and will encourage expansion of public
freight services in a publicly owned railway… 136
This on the face of it is a proposal to reform the rail system along the
lines set out in section 1.1, as a fully integrated, publicly owned
monopoly.
132
ibid., p117
133
HCWS322, 6 December 2016; in 2007 when shadow transport spokesman, Mr
Grayling had said that in hindsight it had been a mistake to separate track and train
in the way that was done in the 1990s, see: “Network Rail to lose sole control of rail
maintenance”, The Guardian, 3 December 2016
134
“Grayling to reveal plans for fully privatised railway line”, Rail Technology Magazine,
6 December 2016
135
“On track towards an integrated railway?”, Rail Professional Magazine, 2016
136
Labour Party Manifesto 2017, pp90-91
In an article for the New Statesman, published on 11 May 2017 the

Shadow Secretary of State for Transport, Andy McDonald, put the policy
in context. He said that: “public transport has increasingly become
detached from the concept of public service. Too often it is seen as a
series of opportunities to profit from an essential service that no
government can let fail”. 137 He went on:
The network of companies who operate passenger services on
Britain’s railways – 75 per cent of which are foreign companies or
foreign-owned state companies that extract profits from British
taxpayers and commuters in order to reduce fares back home –
come together in a jumbled network that drives up the cost of
improvement works, complicates ticketing structures, slows
ticketing reform and extracts eye-watering profits that could
instead go on improvements or keeping fares down from the
system. The hit to the pockets of commuters stands in stark
contrast to the hundreds of millions of pounds in dividends paid
to shareholders of private train companies each year.
The case for nationalisation is good economics, too. Last year,
TUC research showed that the costs saved from bringing
franchises which expire from 2016 to 2020 back in house could
save up to £604 million a year by 2020, enough to lower
regulated fares by up to 10 per cent. 138
Labour’s policy builds on the work of a number of groups, notably the
trade unions and their Action for Rail campaign. One of the most
cogent cases for bringing the railways back into public ownership is set
out in the June 2012 report, Rebuilding Rail, commissioned by Aslef, the
RMT, TSSA and Unite. It argued that the privatised railway is too
expensive and too complex, leading to high ticket prices, inefficiencies
and a service run for the benefit of shareholders rather than passengers.
In summary, it said:
There is a widespread concern – shared across the political
spectrum – that we are not getting good value from the
substantial sums of public money that are invested in the railways
every year. Since privatisation, the cost to the public purse of
running the railways has risen by a factor of between two and
three times … Over the same period, the money going into the
railways from passenger fares has also increased in real terms.
Much of the increase in cost may be attributed to fundamental
problems with the complex privatised railway structure created by
the Conservative Government in 1994. Key reasons for the
increase in cost include higher interest payments in order to keep
Network Rail’s debts off the government balance sheet; debt
write-offs; costs arising as a result of fragmentation of the rail
system into many organisations; profit margins of complex tiers of
contractors and sub-contractors; and dividend payments to private
investors.
[…] The current structure of the railways affects passengers in
several ways. Britain has Europe’s highest commuter fares for
both day returns and season tickets. Ticket purchase is excessively
complex. When things go wrong, there is a lack of clear
accountability.
137
“Passengers not profit: the case for public ownership”, New Statesman, 11 May
2017
138
Ibid.
[…] [I]nnovation is discouraged by the complex and fragmented

structure of the privatised railway, and is more difficult now than
it was before the railways were privatised. The hoped-for
innovation has not materialised. Genuine at-risk private
investment (as opposed to private capital expenditure that is
underwritten by the Government) makes an insignificant
contribution to the railways, representing of the order of one per
cent of the total money that goes into the railway each year. This
is substantially less than the additional costs posed by the
privatised structure. Privatisation has also failed to increase the
efficiency of the railways. Notably, this appears to be due to
increased numbers of administrators and managers …
Fragmentation has produced duplication of functions in the
different private companies and new staff to deal with all the
interfaces between those companies. 139
The report set out a possible governance structure for a future
nationalised and integrated ‘GB Rail’, derived from the Deutsche Bahn
governance structure in Germany (see above) but modified to give
broader accountability. While we do not know what structure Labour
would advocate, this gives us a reasonable example of the sort of thing
it could look at: 140
In effect, the system they proposed was one where the existing
infrastructure manager, Network Rail (NR), would become GB Rail
139
Op cit., Rebuilding Rail, pp7-8
140
Ibid., p72
Network & Operations, with its ‘essential functions’ of allocating and

charging for network capacity hived off to GB Rail Access, but otherwise
its current network operations could continue largely intact. Thereafter,
the organisational changes to NR would centre around building its
capacity to run train operations. It would also involve changes to rolling
stock (train) procurement, whereby GB Rail would be able to procure
new trains directly, using either government grant or government-
backed debt. The report argued that procurement of rolling stock by GB
Rail could be carried out in such a way as to support UK train
manufacturing industry. 141
A slightly different model was proposed by the Centre for Research on
Socio-Cultural Change (CRESC) in 2013. They envisaged a more defuse,
regional model of rail provision to “provide a framework for political
accountability and financial cross-subsidy as long as the railway system
has to provide both an integrated national network and intra-regional
services”. 142 Some sort of regionally or locally owned and run public
railway has received support from other sources. 143 The Co-operative
Party and Co-operatives UK have championed a different sort of
mutually-owned and run railway. 144
141
ibid., pp73-4
142
CRESC, The Great Train Robbery: privatisation and after, June 2013, p162
143
E.g. Compass, see All on Board: A publicly owned railway for an interconnected
world, July 2014 and Progress, see “Make rail regional”, Progress Online, 9 July
2014
144
Co-operatives UK, Co-operative rail: a radical solution, New Insight 6, 2011, p21
[written by Christian Wolmar, the noted transport commentator who stood to be
the Labour candidate for Mayor of London in 2016]
3. Structural reform in the EU

The separation between the infrastructure manager and the rail service
It is not yet clear
provider has been a key feature of EU rail policy for many years. where the UK will
Underpinning this was the desire to end the idea of rail companies as be as regards EU-
natural monopolies. As discussed by the European Parliamentary wide rail policy after
Research Service: Brexit. What, if any,
Built upon the successful reform of other natural monopolies, EU law continues to
such as the telephone and electricity utilities markets, the vertical be applied may
separation principle aims to ensure the independence of depend a great deal
infrastructure management from any rail company, to perform on the final
essential functions such as the allocation of rail capacity or agreement reached.
charging for use of infrastructure. 145 For more
This principle was first enshrined in Directive 91/440/EEC, which information on
these issues see: HC
required a compulsory separation of accounts between infrastructure
Library briefing
and transport entities, leaving institutional separation as an option. This
paper CBP 7633.
implied that transport and infrastructure management services could be
provided by a single legal entity but that at the same time, the accounts
relating to these activities would have to be kept separate to avoid
potential transfers of aid received from one area to the other.
3.1 The Railway Packages

The EU has pursued a policy of gradually opening up railway services to
competition through its four Railway Packages, with the aim of
achieving full liberalisation. 146 However, as discussed by the OECD:
Although progress has clearly been made in formulating and
implementing the EU Commission’s Directives aimed at creating
an open access rail market across the boundaries of the Union,
the current status of the system lags significantly in developing
effective competition between commercially driven enterprises in
national or international freight markets and, even more so, in
passenger markets. 147
Enabling fair and non-discriminatory access to rail infrastructure and
opening up rail services to competition would have had little effect
without specific provisions governing effective access conditions. All
conditions to access the rail market were specified in the Third Railway
Package. Independent regulatory bodies were also established as part of
the EU reform process. The precise structural arrangements vary from
country to country but various regulatory functions (licensing, path
allocation, infrastructure charging and monitoring of public service
obligations) must be entrusted to bodies or firms that do not themselves
provide any rail transport services.
In January 2013, the European Commission presented its Fourth Railway
Package, consisting of six legislative proposals, namely three Regulations
and three Directives, accompanied by an overarching Communication.
One of the key aims of the package was to enhance governance
145
European Parliamentary Research Service, The Fourth Railway Package: Another step
towards a Single Railway Area, March 2016
146
For more information see HC Library briefing paper SN184
147
Op cit., Recent developments in rail transportation services, p32
structures so as to ensure equal access to the infrastructure. Initially, the

Commission wanted to introduce a strict separation between the
infrastructure manager and the service operator. However, under
pressure from some European countries and operators (particularly
German DB and French SNCF operators), it stopped short of requiring all
railways to implement what amounted to the Swedish model. A number
of academics also criticised the approach of the Commission in seeking
to legislate for full separation. Nash et. al (2014) commented that it did
“not seem appropriate to adopt a policy of requiring all railways to be
vertically separated as this will increase costs”. 148 Mizutani at. al (2014)
found that the cost of imposing vertical separation, where it was not
already in place, would increase EU-wide rail costs by around €6 billion
per year. 149
As a result, the final version of the proposal made no provision for
mandatory institutional separation. Vertically integrated businesses are
accepted (for instance in a holding company), provided that the
infrastructure manager is fully independent and has effective decision-
making rights.
The Commission had also sought to grant access rights to all EU rail
operators on all domestic passenger markets, whereby rail companies
would either offer competitive commercial services or participate in
tendering procedures from December 2019. Reaching an agreement on
this ‘Market Pillar’ of the Fourth Railway Package proved difficult.
According to the European Parliamentary Research Service:
The long, ongoing, legislative process is a clear indication of the
different viewpoints and of the debates between the Commission,
the co-legislators, and other stakeholders, on how far
harmonisation of rail governance structures and rail liberalisation
should go. The abandon of the flagship project of mandatory
unbundling initially proposed by the Commission is a sufficiently
telling example in this regard. 150
The European Parliament voted on 14 December 2016 to adopt the
final wording for the Market Pillar of the Fourth Railway Package,
concluding “almost five years of intense negotiations”. 151 Despite
agreement, its actual effect remains unknown to a great extent.
Essentially it means that anyone would be able to bid to compete on a
commercially viable EU rail network from 2020, as already happens in
GB (e.g. open access operators bidding to run services on the East and
West Coast Main Lines to compete with the services offered by the
franchised operator).
From 2026 private companies would also be able to bid for public
service contracts that are awarded by governments on lines that are not
148
Op cit., “Structural reforms in the railways: incentive misalignment and cost
implications”, pp16-23
149
Op cit., “Comparing the costs of vertical separation, integration, and intermediate
organisational structures in European and East Asian railways”, p32
150
Op cit., The Fourth Railway Package: Another step towards a Single Railway Area,
p27
151
“The Fourth Railway Package market pillar - a level playing field or more of the
same?”, International Railway Journal, 17 January 2017
as profitable. At the moment, the majority of domestic rail lines across

the EU are operated under public service contracts. This involves
countries often directly awarding rail contracts to the local incumbent,
which is either compensated or granted exclusive rights on the line.
The Commission’s original proposal would have introduced mandatory
competitive tendering for such rail contracts (i.e. the GB system).
However, opposition from Member States resulted in changes which
mean that governments will be able to directly award contracts where
the geographical characteristics are such that it would result in service
improvements, or where they do not receive enough bids.
In announcing agreement on the Market Pillar in October 2016 the
Council said that competitive tendering would “become the norm for
public service contracts, with some exceptions. Direct award will still be
possible where it leads to better quality of service or cost efficiency”. To
ensure ‘continuous and well-functioning services’, Member States
would also be able to limit a new operator's right of access if the
proposed new service “would compromise the 'economic equilibrium'
of an existing public service contract”. 152
The continued relevance of these provisions in the UK depends entirely
on whether the UK negotiates some sort of EU exit agreement that
includes an exemption from the Market Pillar (like e.g. Switzerland). This
in turn would then mean that a future government that was so inclined
could renationalise the railways under a unified structure (i.e. merge
track and trains) along the lines set out in section 1.1, above. It would
also allow privately-run companies to run ‘vertically aligned’ rail services
across the country, again if a future government were so inclined.
3.2 Structural models in the EU

With the exception of Ireland and Northern Ireland, all EU Member
States have implemented railway reforms. The separation between
infrastructure managers and rail operators, intended to ensure equal
treatment of all rail service providers, has been heterogeneously
transposed in the different States. From these reforms, three broad
models with individual variations have emerged:
Structural models in the EU

Vertically separated Independent Integrated, with
Infrastructure legal and accounting
Manager, One separation
Dominant Operator
Sweden France Germany
United Kingdom Estonia Austria
Bulgaria Belgium Italy
Czech Republic Hungary Poland
Denmark Latvia
Finland Luxembourg
Greece Slovenia
152
European Council press notice, “Better rail services: Council adopts 4th railway
package market pillar”, 17 October 2016
Lithuania
Portugal
Romania
Slovakia
Spain
Netherlands
Source: European Commission: SWD accompanying the 2012 report from the
Commission on monitoring development of the rail market, Annex 2, SWD(2012)
246 final/2.
The most recent summary of the infrastructure management of and

private sector involvement with EU rail systems can be found in the Fifth
report on monitoring development in the rail market, COM(2016) 780
final, 8 December 2016.
About the Library
The House of Commons Library research service provides MPs and their staff
with the impartial briefing and evidence base they need to do their work in
scrutinising Government, proposing legislation, and supporting constituents.
As well as providing MPs with a confidential service we publish open briefing
papers, which are available on the Parliament website.
Every effort is made to ensure that the information contained in these publicly
available research briefings is correct at the time of publication. Readers should
be aware however that briefings are not necessarily updated or otherwise
amended to reflect subsequent changes.
If you have any comments on our briefings please email papers@parliament.uk.
Authors are available to discuss the content of this briefing only with Members
and their staff.
If you have any general questions about the work of the House of Commons
you can email hcenquiries@parliament.uk.
Disclaimer
This information is provided to Members of Parliament in support of their
parliamentary duties. It is a general briefing only and should not be relied on as
a substitute for specific advice. The House of Commons or the author(s) shall
not be liable for any errors or omissions, or for any loss or damage of any kind
arising from its use, and may remove, vary or amend any information at any
time without prior notice.
The House of Commons accepts no responsibility for any references or links to,
BRIEFING PAPER or the content of, information maintained by third parties. This information is
Number CBP 7992
provided subject to the conditions of the Open Parliament Licence.
27 July 2017

Rail Structures, Ownership and reform-UK

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Rail Structures, Ownership and reform-UK

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Rail Structures, Ownership and reform-UK

Uploaded by

Copyright:

Available Formats

BRIEFING PAPER

Number CBP 7992, 27 July 2017

Rail structures, ownership By Andrew Haylen

www.parliament.uk/commons-library | intranet.parliament.uk/commons-library | papers@parliament.uk | @commonslibrary

1. Structural options and case

competition for the market by means of a franchising or concessions

1.1 Vertically integrated

These railways can also be bound by political pressures that lead to

The principle argument in favour of having a fully integrated, publicly-

Another objection to this type of model is that there is no intra-modal

Case Study: Japan

As observed by the World Bank, “almost all private operation of

or concessions have proven more daunting and less attractive as a

In-market competition between integrated rail

Restructuring to promote competition between vertically-integrated rail

Case Study: North America

Firstly, competition over parallel track requires the existence of two or

integrated railways competing among themselves in parallel fashion and

1.2 Intermediate models

2 separate companies within a state-owned group of companies

between infrastructure and operations in Italy, but both parts are

Case Study: Germany

Sources: Kirchner, C. (2005). Germany. In Reforming Europe’s railways. An assessment of

1.3 Vertical separation

market share of rail, which is a declared goal of European and

From the point of view of technical efficiency, 55 vertical separation

Case Study: Sweden

It should be noted that cost increases depend on the circumstances of

by separation”. 68 Additional realignment mechanisms have been

Great Britain: Aligning the incentives

However, it is acknowledged that as “near-impossible” to design a track

and timetables, as well as managing delays and disruption. These issues

1.4 Other factors at play

differences between these two countries in terms economic growth, 76

regulatory reform measures only as other connected measures

2.1 What did the rail market look like before

entirely vertically integrated, that is to say it owned its own trains,

structure. The reorganisation also led to a marked reduction in the PSO

there was no mention anywhere in the Bill of Railtrack – an indication of

Railways Act 1993

The legislation radically changed the structure of the railway industry by

2.3 What happened after privatisation?

Fragmentation and industry costs

The IPPR think tank wrote in 2002 that “privatisation arguably

Following privatisation there was a great deal of criticism of the

implemented a strategy to upgrade the electricity supply for rail services

2.4 McNulty Rail Value for Money Study,

• Base package: Radical franchise reform, regional accounting

2.5 Developments post-McNulty

place responsibility for train operations and infrastructure

right – whatever their thoughts on whether the public or private sector

As Shaw explained in her scoping report:

of the centre and devolving greater responsibility to the route-level. The

The second part of Shaw’s recommendation in this area was to redraw

2.6 Labour’s alternative: public ownership

In an article for the New Statesman, published on 11 May 2017 the

[…] [I]nnovation is discouraged by the complex and fragmented