Liberalisation, our neighbours and the sad reality
Spain entered the high-speed rail era fifteen years after Italy and eleven after France, and it was no pioneer in liberalisation either: it had examples, technology and time to get it right. It eventually became Europe’s leader in network length, yet the accidents shown in the chart account for 195 deaths in Spain, compared with 124 across Germany, France and Italy combined. Neither liberalisation nor traffic growth justifies the failure of a system that should have protected those lives.
Spain was not the first country to develop high-speed rail. Nor was it the first to liberalise its railways. We arrived later, with examples to study, other countries’ mistakes to learn from and technology that others had already tested. We had an advantage: we did not need to invent the model. We needed to use the available experience and get it right. The sad reality is that this advantage has not delivered the safety outcomes we were entitled to expect.
We arrived fifteen years after Italy and eleven after France
Italy opened the first section of the Direttissima in 1977. France brought its first high-speed line into service in 1981. Germany began commercial ICE operations in 1991. Spain opened the Madrid–Seville line in 1992.
Fifteen years after Italy’s first milestone. Eleven after France’s. One after Germany’s.
Those years matter. They represented accumulated experience in infrastructure, rolling stock, signalling, maintenance and operational management. Spain could study what had worked, what problems had emerged and what solutions had been adopted.
There was no need to discover that tracks deteriorate, that materials suffer fatigue or that people can make mistakes. That knowledge needed to be incorporated into the system’s design, maintenance and day-to-day management.
Arriving later should have allowed us to learn before repeating mistakes.
When it came to kilometres, we did want to be first
Although we started later, Spain eventually became Europe’s leader in the length of its high-speed rail network. The UIC table used in the chart, showing the network at the end of 2024, records:[1]
- Spain: 3,993 kilometres.
- France: 2,760 kilometres.
- Germany: 1,941 kilometres.
- Italy: 1,097 kilometres.
Under this classification, Spain has more than twice Germany’s network length and more than three times Italy’s.
To put that scale into perspective, 3,993 kilometres is almost the road distance from Madrid to Moscow, a journey of approximately 4,100 kilometres, depending on the route. Imagine infrastructure stretching across that distance: rails, sleepers, points, bridges, tunnels, electrification and signalling that must be inspected, maintained and renewed for decades.[2]
Building the equivalent of a route from Madrid to Moscow is not enough. It must be kept safe every day.
Every kilometre opened creates a lasting obligation. Construction ends; maintenance begins and cannot stop while the infrastructure remains in service.
We were not the first to start. We did become Europe’s leader in kilometres. That position demanded maintenance capabilities to match.
A record network length does not compensate for a safety failure. The larger the network, the greater the responsibility to sustain it.
Liberalisation was not new either
Germany began opening its railway market in 1994: twenty-six years before Spain’s liberalisation milestone in 2020. The chart places Italy’s milestone in 2001: nineteen years earlier.[3]
Looking at effective competition between high-speed operators, Italo began competing in Italy in 2012. New competitors entered Spain’s domestic market in 2021. Here, too, we had nine years of Italian experience to study.[4][5]
Nine years to observe how services increased, how demand responded and what operational requirements competition introduced.
For Italy, the chart records an approximate 45% increase in services on the Turin–Naples corridor and a 52% increase in passenger numbers over the periods indicated. For Spain, it shows approximately 55% more services and 48% more passengers.
The corridors and periods differ, but the scale of these increases shows that Spain’s growth was not unprecedented. There were already examples of substantial expansion following the arrival of competitors.
The sad reality is that we had reference points to help us anticipate that growth and prepare for its consequences.
Germany showed that opening the market does not impose a single model
Liberalisation means allowing market access under transparent, non-discriminatory rules. Companies that meet the requirements must be able to apply for capacity, access the necessary services and compete without favouritism.[6]
That does not require every company to have the same market share, receive identical timetable slots or be authorised to run more services than safe operation allows.
Germany opened its market long before Spain, yet Deutsche Bahn still accounted for approximately 93% of long-distance passenger-kilometres in 2024. Legal market opening coexisted with limited effective competition.[7]
There were therefore different European experiences to draw on. Italy offered an example of strong competition in high-speed rail; Germany, one of an open market in which the incumbent remained dominant.
Spain could study both and adapt its planning to its own network. The German example did not permit the arbitrary exclusion of competitors or non-compliance with European law. It showed that liberalisation did not impose a particular distribution of market shares or make rapid growth an obligation detached from operating conditions.
We had examples of how to compete. We also had examples to inform prudent planning. Safety had to be ensured in either case.
ADIF allocates capacity: traffic does not appear by surprise
Operators apply to run trains. The infrastructure manager allocates the capacity.
On their respective networks, ADIF and ADIF Alta Velocidad allocate train paths: where and when a train may run. They manage available capacity, coordinate applications and must reconcile operations with maintenance requirements and infrastructure restrictions.[8]
Those powers are subject to rules, oversight and non-discrimination obligations. But the essential point remains: companies cannot unilaterally put as many trains as they want onto the network, at any time and under any conditions.
The infrastructure manager knows the activity it schedules and allocates capacity for. It must assess whether that activity can be sustained, reserve capacity for infrastructure maintenance and impose the necessary restrictions when conditions change.
If there is congestion, it must manage it. If work is required, it must arrange time for it. If a failure reduces the protection available, it must adjust operating conditions.
Authorised traffic is not something beyond the infrastructure manager’s control. It is central to its responsibility.
That is why presenting increased train movements as an explanation that closes the debate is unacceptable. Planning, capacity allocation and safety management systems exist precisely to manage that increase.
Our neighbours also have large public infrastructure managers
France has SNCF Réseau, part of the publicly owned SNCF group. Italy has Rete Ferroviaria Italiana, part of the publicly owned Ferrovie dello Stato Italiane group. Germany has DB InfraGO, part of Deutsche Bahn, which is owned by the federal government.[9][10][11]
Their structures are not identical to Spain’s, but they share a fundamental characteristic: the infrastructure is managed by large organisations under public control.
Our neighbours have also suffered serious accidents, but the toll from the cases included in this comparison is devastating: Spain accounts for 195 deaths, compared with 124 across the other three countries combined. Public ownership and organisational complexity do not justify that outcome.
The toll: 195 deaths compared with 124
The accidents shown in the chart have the following death tolls:
- Spain: 195 deaths. Angrois, 80; Adamuz, 46; Urda, 69.[12][13][14]
- Germany: 101 deaths. Eschede.[15]
- France: 13 deaths. Voiron, 2; Eckwersheim, 11, during a test run.[16][17]
- Italy: 10 deaths. Piacenza, 8; Livraga, 2, both train drivers.[18][19]
Spain accounts for 195 deaths, compared with 124 in Germany, France and Italy combined, in the accidents included in this comparison.
The German case has a distinctive feature: Eschede resulted from a failed innovation, involving the adaptation to the ICE of a resilient-wheel concept used on trams and commuter trains. The outer steel tyre fractured through fatigue; subsequent studies identified deficiencies in its design and dimensions relative to service loads. It was a catastrophic engineering error that, from 1998 onwards, provided a lesson for the entire sector: twenty-eight years before Spain’s accidents in 2026.[20][21]
The chart includes high-speed trains operating on conventional lines and a test train. It counts the accidents shown, rather than all railway fatalities in each country.
Yet the scale of Spain’s toll is unmistakable: Adamuz and Urda alone account for 115 lives lost in 2026.
These outcomes demand an examination of how the system functions: its infrastructure, maintenance, safeguards, procedures and operating decisions. Neither traffic growth nor liberalisation explains, on its own, why the barriers intended to protect those lives failed.
We had examples. We had time. We had to get it right.
We arrived fifteen years after Italy’s first high-speed milestone and eleven after France’s. We liberalised after Germany and Italy. We knew of experiences involving strong competition and others involving market opening with limited penetration by new operators. We had technology, rules and decades of railway knowledge.
There was no shortage of precedents. What was required was to translate them into rigorous management.
More kilometres mean more infrastructure to maintain. More trains require maintenance and operations to adapt. More operators require better coordination. None of those circumstances reduces the obligation to protect lives.
The specific causes of each accident must be explained through evidence. But the judgement on the outcome cannot be hidden behind liberalisation, commercial success or the size of the network: the system failed to protect those who depended on it. That is the failure before us.
The sad reality is that we had examples, time and knowledge that should have enabled a very different outcome. Being first in kilometres should never have mattered more than meeting our responsibilities for safety.
Methodology
Original analysis based on data from UIC, CNMC, ART, OECD/ITF and Bundesnetzagentur, infrastructure managers’ documentation and public sources on the accidents. The timeline extends from the introduction of high-speed rail in each country to November 2026 and distinguishes legal market opening from effective competition. Network lengths refer to the end of 2024, according to UIC.
Fatalities are added together for the accidents shown, each involving at least two deaths on board, including railway staff, accidents on conventional sections and the Eckwersheim test run. These are absolute counts, without adjustment for traffic volume. Increases in train services and passenger numbers are calculated as (final value − initial value) / initial value × 100, using the periods and corridors indicated. The Madrid–Moscow comparison is an approximate reference to road distance.
Limitations
The data cannot establish that liberalisation caused the accidents or calculate an individual passenger’s risk per journey. The selection does not cover all railway accidents and combines different periods, types of infrastructure and categories of people on board, as well as a test train. Traffic growth figures also cover different geographical areas and periods. Network length alone does not measure utilisation or maintenance requirements. The comparison describes outcomes and historical context but does not, by itself, establish the causes of individual accidents or determine specific responsibilities.
Sources
- UIC, High-speed Lines — Extracto del High-Speed Atlas 2025, actualización de mayo de 2026 con datos a cierre de 2024 uic.org/media-files/spip/documents/2025_extract_uic_hs_atlas_hsl_pdf.…
- Travelmath, Driving Distance from Madrid, Spain to Moscow, Russia, sin fecha; consulta en 2026 travelmath.com/drive-distance/from/Madrid%2C%2BSpain/to/Moscow%2C%2BR…
- OCDE/ECMT, Competitive Tendering of Rail Services — Experience with Competitive Tendering in Germany, 2007 oecd.org/content/dam/oecd/en/publications/reports/2007/09/competitive…
- OCDE/ITF, High Speed Rail Competition in Italy: A Major Railway Reform with a “Win-Win Game”?, 2016 oecd.org/en/publications/high-speed-rail-competition-in-italy_6da6e5c…
- CNMC, La liberalización del transporte ferroviario de viajeros generó un impacto positivo de 578 millones en 2023, 2024 cnmc.es/prensa/inf-balance-liberalizacion-ferroviario-20240514
- Parlamento Europeo y Consejo, Directiva 2012/34/UE por la que se establece un espacio ferroviario europeo único, 2012 eur-lex.europa.eu/legal-content/ES/ALL/?uri=celex%3A32012L0034
- Bundesnetzagentur, Railway Market Analysis 2025 — Datos de 2024, 2025 data.bundesnetzagentur.de/Bundesnetzagentur/SharedDocs/Downloads/EN/A…
- Adif Alta Velocidad, Declaración sobre la Red, 2026 adifaltavelocidad.es/sobre-adif-av/declaracion-red
- SNCF Réseau, Finances — Un statut particulier, sin fecha; consulta en 2026 sncf-reseau.com/fr/finances
- Camera dei Deputati, Società a partecipazione pubblica e ricognizione degli assetti organizzativi — Monitoraggio e controllo, 2025 documenti.camera.it/leg19/dossier/testi/CP0725S.htm
- Deutsche Bahn, Annual Report 2024 — Organizational structure, 2025 ibir.deutschebahn.com/2024/en/combined-management-report/db-group/org…
- RTVE, Se cumplen 10 años del accidente ferroviario de Alvia en la curva de Angrois, 2023 rtve.es/play/videos/telediario-1/10-anos-accidente-alvia-angrois/6940…
- Junta de Andalucía, Decreto-ley 8/2026, de 12 de mayo, de ayudas económicas y fiscales a las víctimas del accidente ferroviario de Adamuz, 2026 boe.es/buscar/doc.php?id=BOJA-b-2026-90117&lang=es
- RTVE, Concluye la identificación de las 69 víctimas mortales de la tragedia ferroviaria en Urda, 2026 rtve.es/noticias/20261102/accidente-urda-identificados-66-69-muertos/…
- Bayerischer Landtag, Landtagspräsidentin Ilse Aigner gedenkt am 25. Jahrestag der Opfer des ICE-Unglücks von Eschede, 2023 bayern.landtag.de/aktuelles/veranstaltungen/gaeste-und-begegnungen/gu…
- Le Monde, Un deuxième mort dans l’accident du TGV — Accidente de Voiron, 1988 lemonde.fr/archives/article/1988/09/27/faits-divers-dans-l-isere-un-d…
- BEA-TT, Rapport d’enquête sur le déraillement de la rame d’essai de la LGV Est-Européenne le 14 novembre 2015 à Eckwersheim, 2017 bea-tt.developpement-durable.gouv.fr/rapport-d-enquete-sur-le-deraill…
- Senato della Repubblica, Resoconto stenografico della seduta n. 112 del 15 gennaio 1997 — Accidente de Piacenza, 1997 senato.it/show-doc?id=4765&idoggetto=0&leg=13&part=doc_dc&tipodoc=Res…
- Ministero dell’Interno, Deraglia un Frecciarossa nel lodigiano: centinaia i soccorritori sul posto — Accidente de Livraga, 2020 interno.gov.it/it/notizie/deraglia-frecciarossa-nel-lodigiano-centina…
- Engineering Failure Analysis, The railway accident of Eschede — Technical background, 2004 sciencedirect.com/science/article/pii/S1350630704000214
- Materialwissenschaft und Werkstofftechnik, Praxisrelevante Bewertung des Radbruchs vom ICE 884 in Eschede, 2007 onlinelibrary.wiley.com/doi/abs/10.1002/mawe.200700151
- ART, Le marché français du transport ferroviaire en 2024 — L’essentiel, 2025 autorite-transports.fr/wp-content/uploads/2025/12/bilan-ferroviaire-2…
How to cite this analysis
Dr. Ing. Francisco Badea (2026). Liberalisation, our neighbours and the sad reality. Vía Segura. https://via-segura.com/en/analysis/?a=la-liberalizacion-nuestros-vecinos-y-la-triste-realidad