Add us to your taskbar by dragging this icon RGI logo to the bottom of your screen.

Share |

Reform is imperative, but solutions must be flexible

01 Jul 2005

The last 30 years have witnessed an assault on the mid-century template of state-owned national railways prevalent outside North America. Today, the pressure of globalised transport markets is forcing governments to find structures that allow rail to remain competitive without requiring unsustainable subsidies

Lou Thompson
Principal, Thompson Galenson & Associates

As Railway Advisor to The World Bank from 1986 to 2003, Lou Thompson was close to the reform process that is still progressing across the world. Starting in 1990, his perceptive annual reports and unique database of statistics tracked its progress in Railway Gazette publications up until 2002

ALMOST two decades have passed since Japanese National Railways was 'privatised' on April 1 1987. A year later, Swedish State Railways (SJ) became just a train operator when the track was handed over to Banverket - arguably the first 'infrastructure manager' to be created since the Stockton & Darlington Railway allowed a form of open access in 1825. Since then, restructuring of the state-owned rail networks - which had become almost universal by 1950, the US being a notable exception - has gathered powerful momentum.

What we have not seen to date is consistency in the alternative structures being created. It is perhaps time to take stock, and consider what has been achieved before pressing forward with further reforms.

The background to all this activity is that the more developed economies are shifting from producing industrial goods to services as growing wealth brings more choices in speed and quality of passenger service. But globalised markets are driving transport decisions, putting pressure on all modes and reducing the ability of governments to control the pace of change. Rail reform - changes in railway structure, ownership and management - has followed, not led this process.

The US experience

Paradoxically, I want to start in America, where was no significant government ownership until bankrupt railroads in the northeast were merged to form Conrail in 1976. The new company was successfully sold in a public offering in 1987.

But reform there was, because by the 1970s the industry was failing. Regulators prevented the large Class I railways from withdrawing unprofitable passenger services, and exerted rigid control over freight tariffs, preventing them from responding to customer needs.

The first response was to transfer inter-city passenger services to Amtrak, a new public company, in 1971. Amtrak pays for its access to freight railway tracks, and is supported directly from the Federal and state budgets - although this may change shortly under new legislation proposed by the present government. Next, freight tariffs were deregulated on railways and highways in the early 1980s. Regulators also permitted a wave of Class I railway mergers, from around 70 companies in 1970 to about nine today.

Results have been positive. Though Amtrak experienced policy and funding crises, it at least rescued the freight railways from passenger losses. Their productivity is up sharply, as are profits, although return on capital invested remains below other sectors. In real terms, average tariffs have fallen by more than half.

There have been difficulties with the last few mergers, and the industry has recently faced capacity strains, but the US rail industry is in far better shape than it was 30 years ago.

Latin American turnaround

The state-owned railways of Latin America were financial and operational disasters, collectively consuming billions of dollars in losses and public subsidy. Some, like those in Brazil, were surviving because of strong underlying growth of the national economy. Others, notably in Argentina, had collapsed.

In the 1990s the governments of Argentina, Brazil, Chile, Bolivia, Peru, Guatemala and Mexico broke their railways into separate networks for freight and suburban passengers. Inter-city passenger services were either terminated or transferred to local control and funding. The networks, typically vertically-integrated, were then concessioned to the private sector for maximum payment to government (freight) or minimum payment by government (passengers). Surplus employees were reduced by 50% to 90% under 'safety net' programmes designed to reduce the impact on individuals.

Results were generally positive, though the concessions were affected by the circumstances in the countries around them. Productivity rose rapidly, by factors of three to four for labour, costs fell significantly, traffic grew, and overall rail freight tariffs fell by over 20%. Latin American railways are far stronger at the beginning of the 21st Century than at the end of the 1990s.

EU backs on-rail competition

The European Commission, deeply dissatisfied with the weak performance of EU railways, decided to force reform through structural change. The Commission initially ordered in 1991 that the accounts for railway infrastructure be separated from operations, and that certain competing international freight operators be allowed to use the infrastructure on non-discriminatory terms.

Since then, the Commission has ruled that infrastructure must either be managed by a separate agency, or must have access policies and tariffs controlled by a separate agency. Track access charges must be non-discriminatory. The EC also wants operators separated on a line-of-business (LoB) basis: freight from passenger, and subsidised passenger services from commercial passenger operations. In addition, the Commission's interoperability requirements aim to make cross-border operations technically feasible, although the potential cost is daunting.

Sweden and the UK deserve mention. Sweden created Banverket to equalise public support to rail with that given to highways. SJ's freight operator Green Cargo has been divorced from passenger operations, and parts of the Swedish passenger system have been opened up to competition from private operators.

The UK government initiated total separation of infrastructure (Railtrack, replaced by Network Rail in 2002) from over 25 franchised passenger operations and five commercial freight operators. Rapid and complete privatisation of all sectors was substantially completed in 1996-97.

It is hard to characterise the results in Western Europe as each of the national railways has chosen its own approach. Some have completely separated infrastructure (UK, Sweden, Denmark, Finland) while others (Germany) have formed holding company structures in which infrastructure and operations remain under common control.

The UK, Sweden and Germany have separated freight from passenger operation, but most have kept the two businesses under common control. Very few publish separate results for passenger and freight operations. Some (UK, Denmark, Sweden and Germany) have opened their passenger sector to competition; others (France, Belgium) have not. Freight competition is emerging (UK, Germany), though it is has had little effect on the national operator in the latter case.

Some degree of interoperability is developing, but infrastructure access charges remain a disparate patchwork across Europe and significantly limit rail's ability to compete for freight flows across frontiers.

Privatisation varies. The UK went further than others, but has now retreated with NR's 'private' status often regarded as a charade. Infrastructure maintenance workers formerly working for contractors are now directly employed.

Looming over the picture is the fact that the governments of 15 EU countries spend around €40bn/year, including €10bn in capital investment, on systems that, in total, carry less than 8% of passenger-km and less than 15% of tonne-km.

Results in the UK in particular are controversial. The operators have succeeded in driving up both passenger-km and tonne-km by around 35%, while safety continues to improve despite 20% more train-km being operated. The rolling stock leasing companies (ROSCOs) are delivering a substantial fleet of new equipment, albeit with teething problems, reducing the average age of the passenger fleet from over 20 to just 15 years.

Railtrack failed partly because of conflicting roles, partly because of built-in problems such as faulty access charges and contracts for track maintenance that left contractors to plan and monitor their own work. Another factor was a lack of expertise on the part of Railtrack's management.

Government imposed reforms in a rapid and unco-ordinated way, and it was caught unprepared when traffic grew rather than stagnating as expected. As a result, government has been involved in a catch-up process that is still ongoing, driven primarily by a fourfold rise in taxpayer support in real terms between BR's final year before reform started in 1994 and the current fiscal year.

Soviet Union and Eastern Europe

The roof fell in for railways in Central and Eastern Europe (CEE) and the former Soviet Union (CIS) after the Iron Curtain collapsed in 1989. Even though these economies have started to grow again, few railways carry today even half the freight and passenger traffic they previously enjoyed. They face a daunting challenge of oversized networks and excess manpower along with a need to adjust from a production to a market-driven management mentality. To date, few have been successful.

Those countries that joined the EU last year, or aspire to join soon, have adopted their own versions of the EU directives. Indeed, Poland and Romania have adopted or proposed changes in structure and ownership that would put them in the forefront of EU reforms. In some countries, however, the reforms lack substance and little has really changed.

Two cases - Russia and Estonia - merit attention. Russia has recently adopted significant reforms in which:

  • the former railway ministry was split into a joint stock holding company, OAORZD, and a government function that was transferred to the Ministry of Transport;
  • the railway company under OAORZD has been split into an infrastructure function, a freight operator, and a passenger operator - which is to be further divided into one or more long-distance companies and a series of local passenger operators;
  • freight wagon and locomotive ownership has been opened to private investors;
  • access to the infrastructure has been opened to competing private operators.

Private wagon ownership has grown rapidly, and private freight carriers are already operating. OAORZD is aggressively pursuing the creation of separate passenger carriers, and seeking to transfer a share of the planning and funding responsibility for suburban operations to local entities. The challenge for Russia currently is that the pace of reform, primarily driven by OAORZD, has outstripped policy development and regulation on the government side.

Estonia was the first of the CEE or CIS countries to privatise, creating the new Estonian Railway (EVR), which is 34% owned by government and 66% by private investors. EVR manages the infrastructure and freight operation, with other agencies or companies managing passenger services.

Until recently EVR did well, growing freight by 11% between 2001 and 2004. Unfortunately, government policy on access to infrastructure changed after privatisation, and EVR has recently been ordered to provide access to competing operators on congested lines, a problem made worse by a level of access charges that is far below the costs that EVR itself has to cover. This has eroded the ability of EVR to invest, and could weaken the privatisation process.

Success in Japan

Despite at least four attempts to bring the losses of the Japanese National Railways under control, deficits were around US$15bn/year in the mid-1980s, and JNR's liabilities eventually reached US$337bn.

In 1987 the government split passenger operations into six vertically-integrated networks. A single national freight operator pays for track access. Shares in the three Honshu companies (JR East, JR West and JR Central) were subsequently sold off. Much of the debt and the non-rail assets of the old companies were shifted to a settlements corporation, although the three large passenger operators and the freight company did assume about US$131bn in liabilities.

The three smaller companies on the islands of Hokkaido, Shikoku and Kyushu still lose money, but their deficits are covered by a trust fund that the government established.

The government views the outcome as a success. The annual deficit has been replaced by taxes on the profits of the three large passenger railways. Traffic has grown slowly on the three large systems, but has stagnated on the three smaller islands and the freight railway. Labour productivity more than doubled, and there have been no significant tariff increases since 1987. The world's largest passenger railway system (by journeys) or third largest (by passenger-km) has been successfully and safely operated by the private sector since 1987 - although questions have arisen about competitive pressures at JR West since the accident near Osaka in April this year.

Does any of this work?

A number of successful reform themes do emerge from the results, and these are summarised on p421. In addition, it is important to recognise that the largest group adversely affected by private-sector involvement is the employees, who will resist labour force reductions that can be up to 90% for the same output. Governments that take an open approach to labour and adopt impact reduction measures with reasonable compensation (Argentina, Brazil, Mexico, Poland) have been able to manage the reform process with a minimum of pain and disruption.

The 200th anniversary of Railway Gazette will certainly have an article reporting on railway reform, and/or it will report the death or retreat into expensive irrelevance of a number of major railways. While a future for rail (if not all of today's railways) in urban transport seems assured, and while at least some markets for high speed passenger rail may well survive and prosper, some existing inter-city rail passenger markets are increasingly under attack from cars, buses and planes, and are unlikely to survive without massive and probably unjustified public support.

Railways still enjoy an advantage in long-distance haulage of massive quantities of bulk cargo. Large countries producing high tonnage of coal and agricultural commodities (USA, Canada, China, Russia) meet these criteria, though the future of coal is clouded by the reaction to global warming. European markets stretch from Gibraltar to the Urals, but the ability to generate truly long hauls by rail is limited by national frontiers, operator interchanges, and changes of gauge.

The EC has tried to overcome border effects, but success so far is limited and, of course, only a real visionary would predict a seamless network all the way to the Urals. Railways have no particular advantage in hauling higher value commodities, and the future of this market segment will depend on seamless and reliable service, both in terms of company and national boundaries. Trucks and airlines already offer such service, and their competitive power will only get stronger.

Whether the next centennial issue of Railway Gazette reports success or failure in the freight arena for Europe depends heavily on the railways themselves. If infrastructure is freely and rapidly available to all freight operators on a non-discriminatory basis, with access charges that promote free traffic flow, and if European freight operators can merge and operate effectively across all borders, then a positive picture may emerge.

But if EU rail freight operators remain hunkered down behind national borders, protecting national interests and opposing access by competitors, then the freight section of the July 2105 issue may be shorter and less pleasant. That would be a shame.


Four themes of railway reform

Disentangling railways from government

A universal aspect of railway reform has been separation of the operational and commercial function from social and policy aspects of the government's role. Typically, this has entailed creation of a railway corporation under commercial law, with policy and regulatory roles remaining in ministries.

Unless this is done, it is nearly impossible for the railway to compete against other modes in the private sector. Government railways bear social costs and political inefficiencies that restrict their competitive abilities, but they often twist policies and regulations to make on-rail competition impossible.

Line-of-business (LoB) management

Railways serve at least three distinct market segments: freight, inter-city passengers and suburban or regional passengers. These have such different characteristics of demand, competition, regulation, subsidy and policy that no single management can successfully handle them. LoB appeared in the US (Amtrak) and Canada (VIA) along with state or locally-operated commuter authorities.

Most Latin American concessions were separated between freight and passenger, with little interaction among them. The Japanese creation of separate passenger and freight companies now has its equivalent in the UK, Sweden, Russia and Australia, among many countries, but still without on-rail competition in Japan - although there is competition between passenger companies on parallel routes in the major conurbations. EU directives are also headed in this direction, though they are only now reaching the point of requiring true separation by LoB.

Vertical separation

Separation of infrastructure from operations has proved more controversial, though much of the opposition to the idea is based on a misunderstanding of terms and objectives. There are three broad approaches: the integrated monolith, the owner/tenant approach, and the separated approach.

Each is appropriate under the right circumstances. Few would argue against operating single-commodity railways like the iron ore lines in Western Australia as an integrated business. Where a particular user occupies only a small percentage of line capacity (Amtrak and VIA in North America) the owner/tenant approach can be an effective solution.

In cases where there are multiple, nearly equal claimants on line capacity, and/or in cases where there is a need to promote on-rail competition on the same line, then some form of neutral provision of infrastructure is a useful solution. The case for separation is stronger when government wants to subsidise only some of the operators, or where support to infrastructure must be neutrally provided to all users.

Vertical separation creates specific issues, particularly transaction costs among operators and infrastructure, co-ordination and safety costs of split operation, and the need to develop access charges that raise the desired amounts of user fees without seriously distorting traffic in total or among classes of users. In short, there is no 'one size fits all' approach available.The private sector

The public sector can, and does, build and operate railways successfully (many EU countries, Russia and China). There are equally large and successful freight and passenger railways operated by the private sector (North America and Japan). Latin American concessioning reduced government costs and freight tariffs while increasing demand and productivity, and the same can be said of Japan. There are also favourable but mixed outcomes from the UK and Estonian privatisations.

Private-sector involvement has particular value when the existing railway is inefficient, financially burdensome, or unresponsive to market forces (often all three). It can also have manifest benefits when the commercial activities of the railway need clear separation, and when at least some parts of the railway are forced to compete with other modes.

That said, if an existing publicly-owned railway is well operated and competes effectively and fairly with the rest of the transport sector, then the benefits to be expected from private involvement are reduced. In addition, the transaction costs of establishing the terms of franchises or of asset sales are high, and the outcome can be uncertain, especially where the government has only limited experience with such transactions, or where the ability of the private sector to take on risks and make investments is limited.

Concessioning and privatisation can fail, expensively, and have done so when circumstances are wrong. Governments should play carefully in the rail privatisation game.


Rapid expansion in China

The national Chinese Railways network of 73000 route-km managed directly by the Ministry of Railways in Beijing is now the world's largest in terms of freight traffic, although US Class Is carry more in total. CR is also supporting traffic growth (157% tonne-km and 244% passenger-km) with significant network growth (17% route-km) between 1980 and 2002.

In addition, there are some 9000 km of standard gauge 'local' or 'joint venture' lines that appear to be managed independently from the MoR, with more under construction.

More importantly, the MoR has announced a massive expansion plan costing more than US$200bn up to 2020. This will extend the CR network to about 100000 km, more than doubling the amount of double track and electrified route. In addition to that, some 13000 km of new, dedicated higher speed passenger lines and a network of designated high-capacity freight lines are planned.

The reform component of the plan remains under discussion. The government apparently envisages a separation of railway enterprises from the current MoR functions. MoR has announced that its 14 existing Administrations will eliminate their sub-divisions, removing an entire layer of management that has become outdated with the implementation of the new Transportation Management Information System (TMIS).

Four new Administrations have been created to give a total of 18, apparently to produce geographically more manageable parts. The fate of earlier ideas, including reducing the number of Administrations to four or five, infrastructure separation at the Administration level, and separating passenger operations from freight into distinct lines of business, is unclear.

  • CAPTION: Green Cargo was created as a stand-alone freight operator through the break-up of Swedish State Railways; it later expanded by taking a 45% stake in Norway's CargoNet
  • CAPTION: Sweden's pioneering decision in 1988 to separate the management of infrastructure and train operations triggered a wave of reform and the introduction of on-rail competition around the world
  • CAPTION: The concessioning model first adopted by Argentina in 1991 has proved popular in Africa and other developing regions
  • CAPTION: Railway reform in Russia has enabled RZD to form partnerships with shipping lines and other rail operators to tap the growing flows of landbridge traffic between the Far East and Europe driven by market globalisation
  • CAPTION: With 18 years of experience since the break-up of Japanese National Railways, the three principal JR Group companies have been successfully floated on the stock market


Reform is imperative, but solutions must be flexible


The last 30 years have witnessed a revolution in the near universal template of state-owned national railways outside North America. Today globalised markets are putting pressure on all transport modes, forcing governments to find new structures that allow rail to remain competitive without unsustainable subsidies. Disentangling railways from government, line-of-business management, vertical separation and privatisation are the main components of a complex problem to which there is no universal answer

La réforme est impérative, mais les solutions doivent être souples

Les 30 dernières années ont été le témoin d'une révolution dans le moule quasi-universel des réseaux ferroviaires nationaux détenus par les états, Amérique du Nord exceptée. Aujourd'hui, les marchés mondialisés mettent la pression sur tous les modes de transport, obligeant les gouvernements à trouver de nouvelles structures qui permettent au rail de rester compétitif sans subventions insupportables. Démêler les liens entre chemins de fer et gouvernements, introduire le management par activités, l'intégration verticale et la privatisation sont les principaux composants d'un problème complexe pour lequel il n'y a pas de réponse universelle

Reformen sind zwingend, aber Lösungen müssen flexibel sein

Die letzten 30 Jahre haben eine Revolution in den beinahe universellen Muster von sich in Staatsbesitz befindenden nationalen Bahnen ausserhalb Nordamerika gezeigt. Die heutigen globalisierten Märkte erzeugen einen Druck auf alle Transportmittel und zwingen Regierungen, neue Strukturen zu finden, welche die Schiene ohne untragbare Subventionen konkurrenzfähig erhält. Die Entflechtung der Bahnen vom Staat, Spartenmanagement, vertikale Trennung und Privatisierung sind die Hauptkomponenten eines hochkomplexen Problems, für welches es keine universell gültigen Antworten gibt

Es necesaria una reforma, pero las soluciones deben ser flexibles

Los últimos 30 años han sido testigos de una revolución en el modelo casi universal de los ferrocarriles públicos nacionales fuera de norteamérica. Los mercados globalizados de hoy en día est? n haciendo presión sobre todos los medios de transporte, forzando a los gobiernos a buscar nuevas estructuras que permitan al ferrocarril seguir siendo competitivo sin necesidad de subvenciones insostenibles. Separar el ferrocarril del gobierno, una gestión orientada a la actividad comercial, separación vertical y privatización son los componentes principales de un problema complejo para el que aún no existe una solución universal