Accountability: Passenger Railway Governance
How it used to be done
This post is about accountability: it doesn’t review other areas of the GBR consultation.
It argues that GBR needs a broader-based accountability model than one based on TfL, as the Mayor is directly accountable to Londoners in a way that the Secretary of State is not directly accountable to voters.
There have only been three models of passenger railway governance in the last century.
Until the Second World War, the railway was run by “the big four” railway companies (LNER, GWR, LMR and Southern): private firms but subject to state regulation.
Then from 1948 to 1994, the railway was run by the British Railways Board: politically led but independently operated.
Then from 1994 to Covid, the railway was run by contract, with specifications defined in tenders issued by the Department for Transport, and codified in franchise agreements.
As we move into a new era, will the accountability plans for GBR deliver for users?
Back to the Future
The accountability model for GBR is similar to that of British Railways: an operationally independent organisation subject to a Board appointed by the Government. Here’s what the consultation says:
GBR will, therefore, be held to account first and foremost by the Secretary of State through its Chair and Board. The Secretary of State will appoint the Chair and have a clear role in the appointment of Board members...
This is also very similar to the accountability model for TfL. Indeed, Great British Railways has its genesis in the Williams-Shapps Plan for Rail, in which Transport for London was explicitly referenced as a model. It feels like distant history (it was four years ago!) but here’s what it said:
Under these reforms, Ministers will hold Great British Railways to account through a structured framework underpinned by legislation. Ministers will take key funding decisions and have strong levers to set direction and pursue government policy. The Secretary of State will be responsible for the appointment of the Chair and agreeing the framework for pay, including any performance-related pay. They will also be given statutory powers to set long-term strategy and have powers to issue guidance and mandatory directions to Great British Railways on any matter at any time, creating a relationship between Ministers and Great British Railways akin to the one shared by the Mayor of London and Transport for London.
This is dangerous.
The TfL model looks superficially appropriate for GBR but there are very important differences.
The London system works on the basis that TfL‘s customers and the Mayor’s voters are more or less the same people. The Mayor is scared of the voters and TfL is scared of the Mayor.
Mayoral elections are more or less single-issue votes on transport (as the mayor has such limited wider powers), meaning there is direct accountability from users’ experience to TfL‘s actions through the Mayor.
None of these things are true at the national level.
GBR customers and Transport Secretary voters are completely different groups of people. GBR customers are millions of people spread nationwide; the Transport Secretary is elected by 70,000 people in their own constituency. The fear that the Mayor has of being held to account at the ballot box for their transport decisions simply won’t apply: general elections are fought on many issues but, sadly, transport is never one of them. Anyway, Transport Secretaries know that by the time the election comes around, they’ll probably be doing a different job. Since 2000 we’ve had three Mayors of London and 14 transport secretaries. We’re already onto our second since last July’s election! Being honest, the key ‘customer’ of the Transport Secretary is the Prime Minister (who controls their promotion prospects), not rail users (who do not). This is a very different dynamic from the Mayor of London.
This is important because the Mayor is TfL’s only source of accountability. There are very few constraints on the Mayor. If he wanted to, the Mayor could double fares and halve the service. What the Mayor wants, the Mayor gets; what the Mayor doesn’t want, nobody gets. TfL is more like a medieval court than a governance structure. To deliver for customers, the whole system relies on the Mayor being guided by a desire to deliver for voters.
Would GBR provide the best possible service with such a governance model? And how could the Transport Secretary be the engine of accountability for rail users everywhere from Cornwall to Castleford? Even if they were certain that they’d be held accountable by the voters for their decisions as transport secretary (which they won’t be), is it even credible?
Relying on the TfL model of political accountability is setting GBR up for failure.
Now, you may ask why I say this, given that the TfL model is not hugely different to the old BR model, with an “independent” Board and a Chair appointed by the Secretary of State. The answer is that ministers (and the DfT) have got used to a far greater level of control in recent years. When the British Railways was created, it replaced the Big Four (the British Railways Board itself came a few decades later). The working assumption was that railways were run by independent railway companies: professionals were in charge. Ministers and the department meddled (a lot!) but not to the extent they do now.
It’s very hard to give the DfT total control and then expect them to let the professionals get on with the job. I know the consultation says that ministers won’t meddle, but if ministers control appointments to the Board and the Board is the accountability model then.. they will. Trust me, they will.
Indeed, only last week, the Government abolished NHS England precisely because the Prime Minister believes that ministers should make decisions “at the heart of Government”, and that an arms-length body like NHS England reduces accountability. If that is the Prime Minister’s expectation, any future Secretary of State will expect phone calls from Downing Street on operational detail. And we can’t always assume we’ll have a Secretary of State with the knowledge and experience of transport that we do now.
Moreover, it’s hard to argue that pre-sectorisation British Rail was a model of efficiency and innovation. If the GBR governance model looks like the old BR governance model, we’re in danger of a combination of stagnation and meddling.
“Strong” Passenger Body
Luckily, the consultation pre-empts this problem. There are 52 references to the “powerful” new passenger watchdog that will be created. This watchdog will do lots of great stuff to champion the interests of users. However, the document is silent on how the watchdog will be appointed. We can therefore assume that, like Transport Focus, the new watchdog will be appointed by the Secretary of State. This is a big issue. Why do we need a new watchdog if we already have one? The reason must be that Transport Focus isn’t actually seen to be standing up for rail users. That’s not a criticism of Anthony Smith, the CEO of Transport Focus for most of its lifetime. To quote Matthew Engel in his book Eleven Minutes Late, being appointed by the Secretary of State
…puts Smith roughly in the position of a court-appointed defence lawyer in a Soviet show trial. He is allowed to put in the odd word on behalf of his clients - and indeed is frequently quoted in the news media - provided he does not do so too vigorusly.
It is probably no coincidence that the only time Transport Focus made the news ‘vigorously’ was when it shot down the entire booking office closure programme in 2023: after Anthony Smith had announced his retirement but before he had actually left.
The “strong” passenger body - if appointed by the Secretary of State - will be in the position of being asked to hold their boss to account. Ever tried doing that?
Moreover, there is no independent budget for the passenger watchdog. DfT would sometimes use budget as a source of (strictly unofficial!) leverage over Transport Focus, and that was when Transport Focus was holding to account private train operators. When the new watchdog is threatening to put out a report heavily criticising GBR or DfT, it would be superhuman of the Secretary of State not to use the leverage of the Transport Focus budget as a way of gently nudging the watchdog to reconsider.
The consultation may use the word powerful repeatedly: but power doesn’t come from adjectives: it comes from budget and independence. Without it, the watchdog will have neither bark nor bite.
What Do The Europeans Do?
Britain is unusual in having undertaken a wholesale privatisation of its railways: in Europe, there have been experiments with open access but there are still dominant national train operators with semi-detached infrastructure operators. So the situation is not dissimilar to the position the UK will find itself in post-nationalisation.
So how does governance work across the channel?
Well, the model varies across different countries, but there are some important similarities.
In most European countries, there is a contractual relationship between a client (typically a Government agency) and the operator (often the national operator). This is different to the model adopted for British Rail, in which BR was an arms-length agency responsible to a nominally-independent Board. The contractual model creates a specific client (national or local Government) and a provider (the railway). In the model envisioned, these decisions all happen within the walls of GBR.
Typically, in most major European countries, the national DfT-equivalent specifies service standards for long-distance trains with local regions contracting local and regional trains.
Long distance
Across much of Europe, the high-speed network (what I’ll define here as the ‘pointy nose trains’) is operated commercially. ICE, TGV or Frecciarossa trains have timetables that are largely defined by the national operator in response to customer demand, sometimes with some high-level service standards determining that a certain number of trains must call at certain smaller stations or be extended onto branch lines that would otherwise not be served. Pricing is often commercial. Brits complain about Avanti pricing, but try getting a TGV from Paris to Annecy in the ski season and then grumble about Manchester!
There is more rigid specification for what I’ll call the 'flat-fronted’ long-distance trains (branded as InterCity in almost every mainland European country). These are trains similar to the British CrossCountry network that connect up multiple cities with fastish trains that stop comparatively frequently. Equivalents to CrossCountry’s Scotland to Cornwall trains include the Italian Intercity trains linking Lecce (in the heel of Italy) to Milan (in the north) and the two-hourly French Intercités trains between Marseille and Toulouse.
A German InterCity train - this is contracted by Government to DB to operate
If Britain were to adopt this approach, then the core LNER, GWR (high speed) and Avanti trains would be run commercially by Great British Railways without much Government interference.
The slower trains run by Cross Country, East Midlands Railway, South Western Railway and GWR trains would be covered by a more rigid InterCity contract, that looked rather like a franchise passenger service requirement.
Prices would stay high on the long-distance network at peak times (and be largely unregulated) but this would generate revenue that would help pay for the railway. Flexible walk-up tickets would be at risk, though the ‘cliff-edge’ crowding impact of the end of the peak period would also end.
The slower InterCity trains would probably be pretty similar to those today, both in timetable and price. It’s hard to imagine the DfT moving away from the current spec for these trains, which is a pretty regular hourly clockface timetable for a core, national network.
Local and regional
In most European countries, central Government doesn’t get too involved in local and regional services. Most European countries have strong regional Governments. Germany and Austria have Lander (states), Switzerland has Cantons while France and Italy have Regions. These are regions with developed local identities, well-established Governments and a lot of political influence. In most of continental Europe, these regional Governments determine the service they want and then contract with a railway company to deliver them. While there are smatterings of commercial tendering, the vast majority of these contracts are awarded to the national railway operator.
Replicating this creates an immediate problem for the UK, as we don’t have regional Governments.
The average population of French Regions is 5.1 million; of German Lander 5.3 million. Britain just doesn’t have credible governance bodies at this scale: British counties are far too small. Sub-National Transport Bodies (STBs) have been created specifically to try to take on this role but they’ve not got the capability or the political strength to perform the role that the regions perform in most of Europe. Sub National Transport Bodies would need much bigger, well-resourced transport teams (which means bigger budgets) and the political credibility to represent their region. Both seem unlikely to emerge any time soon.
However, there is a potential solution and, as is so often the case, Switzerland looks like being the model.
The average Swiss Canton has a population of a bit less than 500,000. This makes them equivalent to the Strategic Authorities being set up through the Devolution White Paper. As a reminder, the White Paper asks counties to develop proposals for local authority mergers to create new entities with a population of around 500,000. The big difference between the new Strategic Authorities and the STBs is that once the Strategic Authorities exist, they will be politically powerful and well-resourced: combining all the resources of the existing local authorities. If the Cantons have a role in Switzerland, it’s one that the Strategic Authorities could take on in GBR.
And it turns out that the Cantons have a pretty big role.
The Cantons are responsible for regional and local transport planning and funding. Even though they are far smaller than French Regions, they still negotiate contracts with Swiss operators for regional rail services, buses and other modes of transport - and provide cash for doing so. Indeed, one of the reasons why Switzerland has such a comprehensive local network (Switzerland has the highest density of rail lines in Europe) is because there’s always a local Canton buying (and therefore protecting) local services.
Now you may ask how on earth it can work. After all, there are 26 Cantons. England will have more than 100 Strategic Authorities. Is it even credible that all these organisations can specify services? After all, trains don’t operate in neat borders. Endless trains cross between Cantons: what if different Cantons want different things?
I asked this question when I visited SBB, and the answer is basically that they talk and make long-term decisions. (When I visited last year, I met the team working on the 2050 timetable! The 2037 timetable was already set). There’s plenty of time for the various Cantons to get together and figure it all out. The sense I get is that while it all results in a set of contracts, the creation of those contracts is a process of human-to-human discussion and collaboration.
That’s a great model but it’s counter-cultural for the UK.
Britain works on the basis of short-term muddling through. If we’re to adopt the Canton model, we’re going to need to become better at long-term planning and dialogue.
But that doesn’t seem a bad ambition…
Small is beautiful
One other lesson from Switzerland: there are multiple local railways.
The largest Swiss Canton, Graubünden, buys its service almost exclusively from its local operator Rhätische Bahn (RhB), 51% owned by the Canton of Graubünden, 43% by the federal Government and the rest by small municipalities. Other vertically integrated railways include Bern Lötschberg Simplon Bahn (BLS) and the Matterhorn Gotthard Bahn (MGB). These railways are also generally primarily owned by their local cantons.
A Swiss train operated by BLS
In this post last year, I argued that private v public isn’t a core determinant of good service for users but small v big is. Maybe the fact that 40% of the Swiss rail network is operated by dozens of small, independent operators is part of the reason why it’s so good. Those firms benefit from the ‘small is beautiful’ principle, and it also means that the national operator SBB isn’t quite the behemoth it would otherwise be. Indeed, the Swiss Government recently obliged SBB - as part of its long-distance network licence - to start contracting some inter-regional services to the local operators. The DfT in Switzerland does not want a single, large dominant national operator as - even in a fully-integrated network - it recognises that competing operators have the potential to inspire innovation.
Are there lessons for the UK here?
When the Cornish branch lines no longer have a local GWR franchise keeping an eye on their interests, could they become a small, vertically integrated railway, part-owned by Cornwall Council? Would all the various branch lines in the East Midlands be better off as an independent operator? Would the campaign to reopen the Ivanhoe Line have a greater chance of success if there were a small, agile operator headquartered locally, trying to find cost-effective ways to get the job done and working closely with the local community of which they were part? That railway could be owned by the East Midlands Combined Authority.
Penryn station: could this be part of a new Cornish branch lines rail company?
There is a real danger that if branch lines become part of a much broader GBR funding settlement, they will always be the losers. Realistically, how big will Penryn be, when seen from London (or Derby)? Or Bottesford.
If they have their own bespoke railways, it becomes much harder for them to be chipped away.
Now, I can hear many of my readers complaining that this would be absurdly inefficient: with endless duplication. As I wrote about earlier this year, I’m not so convinced that gigantic organisations are necessarily more efficient. The reduced duplication can be more than offset by the increased inefficiency of how long it takes to get things done.
Dangers of decentralisation
The Swiss experience illustrates one of the benefits of decentralisation: France shows the risks.
Brits have a tendency to be impressed by the French rail network as a result of their experience of TGVs. But the downside of Regions playing a big role in buying services is that they can choose to spend their money elsewhere.
Take Pont-Saint-Esprit, for example. It’s a town of 10,000 people (about the same size as Ludlow) in Occitanie in southern France. The nearest city is Avignon, about 30 minutes away. But make sure you don’t miss the train: there are three in the morning peak and two in the afternoon peak, and that’s it. No nipping into town for an hour. This is pretty typical of rural branch lines, though many French stations have far less service than this: indeed, many railway stations have SNCF services provided exclusively by diesel buses.
Now, fair enough, presumably Occitanie feel they have better things to spend their money on than train services. As a taxpayer, you may agree with them.
But from the narrow perspective of public transport, this may not be a good thing.
Local accountability runs the risk that the railway is at the bottom of the funding pile for cash-strapped Strategic Authorities.
However, it’s important to note that this risk also applies if the whole thing is run centrally.
The reason local rail services in the UK have been protected better than those in France in recent years largely results from the contractual tangle of the privatisation model that makes cuts hard.
Because DfT has to formally specify the services it wants and only does so every seven years, there’s an obvious ‘moment’ when cuts can happen. As a result, journalists are observant and the political danger (even for a transport secretary likely to change jobs) is high.
Whereas a GBR operated with the governance envisaged in the consultation would be very vulnerable to local cuts through salami slicing the service down without anyone really noticing. Penryn currently has a train every thirty minutes. Who’d notice if, for reasons of operational convenience, a few half-hourly slots got missed out? And then it because hourly? And then a few gaps emerged so a few of the hours there was a two hour gap? And then it became two-hourly? And then it becomes like Pont-Saint-Esprit.
Accountability
One of the lessons of the European experience is that the decision-making process is separate from operations.
Key decisions are made through a formal negotiation between the railway and its client (national, regional or local government) and written down in a contract. This is important as it maintains clarity on what the specification actually is, and prevents salami slicing.
The TfL model uses the Mayor as a proxy for the customer, but even then, the fact that decision-making and operations are centralised in TfL for some modes (like the Underground) has an impact, and not necessarily a positive one.
My former colleagues at TfL won’t like me saying this, but when a problem occurs in one of the contracted bits of TfL (e.g. the bus network or DLR), the atmosphere is one of accountability: the operator must fix the problem. No excuses. When, however, it is a directly operated service like the Underground, there is much more awareness of the challenges that have caused the problem and why it is genuinely tough to fix. As a result, TfL can let themselves get away with things they would not let their subcontractors get away with. And that’s in an environment with the Mayor breathing down their necks as a genuine voice of the customer. If a centralised GBR faced a problem with the Weymouth to Westbury line, would anyone really care - unless the Transport Secretary happened to be the MP for Yeovil.
Whatever model is created needs to be clear what the method of accountability is going to be.
In London, it is the Mayor. In the UK privatised railway, it is a contract. In Europe’s nationalised railway, it is a contract. We need one, and my argument is that it can’t be the Transport Secretary.
Fare Box
There is, of course, one more potential source of accountability. It’s the one used by virtually every private business from McDonalds to Microsoft: the paying customer.
The best possible form of accountability is to have to offer your services to someone and ask them to pay for it. That’s one of the reasons I enjoy running Freewheeling (and previously enjoyed running my startup Sn-ap). There’s a discipline and a focus that comes with being a purely commercial business.
The problem is, of course, that railways have a much broader social purpose and - as a result - we wish to have more railway than the fare box will pay for.
However, it’s not a bad idea to create as much commercial discipline as possible.
Indeed, one of the more intelligent acts of old British Rail was to recognise that while the railway as a whole consumes subsidy, there are bits of it that could be run as commercial businesses. InterCity was profitable and Network SouthEast was given an explicit goal of becoming profitable. The remainder of the railway was hived off into a rather amorphous mass called Regional Railways (as writing “The loss-making bit” on the sides of trains wouldn’t have been so attractive).
This wasn’t a bad model for Network SouthEast and InterCity - and it worked, with quality and ridership accelerating. It wasn’t so good for Regional Railways but then, to an extent, it was only an error of omission: they had been somewhat neglected before and they were somewhat neglected afterwards.
If sectorisation provides some important lessons, then the advent of open access provides even more. The East Coast shows the benefits of competition between operators. So does Italy. Travel on the regional network in Italy and you’d think Trenitalia’s paint scheme was a hyper-cool urban grunge style modelled on graffiti - until you realised it’s just that all the trains are covered in graffiti. But the Frecciarossa network is fantastic: with service offers driven up by competition from Italo. Countries that have opened up their long-distance networks to competition (like Italy and Spain) have seen quality go up and prices come down. France operates its network commercially but the only meaningful competition to the SNCF services comes from a funky startup with vividly-coloured trains called OuiGo. It’s owned by… ah… SNCF and uses old SNCF trains, deliberately made uncomfortable.
Some ideas on Improvements
I’ve poked some holes in the consultation but there’s lots of good stuff to build on. It’s thoughtful, intelligent and pragmatic. But, inevitably, humans being what they are, a consultation issued by the Government proposes that the accountability mechanism for a business owned by the Government should be the Government, with the assistance of a watchdog appointed by the Government.
Here’s how we can keep all the good stuff in the consultation but tighten it up a bit.
1. Passenger Rail Should be Run on a Commercial Basis Where Feasible
One of the successes of sectorisation was that Network SouthEast and InterCity were treated as businesses. Railcards, now seen as discount schemes, were commercial tools. Privatisation was a bit like Tesco Clubcard being seen as so valuable it should be protected by statute.
Some people criticised the accounting that enabled these bits to be treated as commercial but I would argue that was something of a masterstroke. It’s also the same model that’s adopted in most European countries, where the ‘pointy nose trains’ are run commercially, in competition with cars, coaches and planes. In some countries like Italy and Spain, they also run in competition with Open Access operators. We should replicate this model for our pointy nose trains.
Our ‘flat-fronted trains’ should replicate the model of European equivalents, and have a direct contract with Government, so it’s explicit what the DfT is buying. In many ways, the paper describes a bit of this, by describing a continuation of the High Level Output Statement (HLOS) and Statement of Funds Available (SoFA) process. It’s just that for the long-distance passenger railway, this should culminate in an explicit, written contract so everyone can see what has been bought.
2. Regional and Local Rail Should be Specified and Funded Locally
Local rail services - such as commuter and rural lines - should be overseen by regional transport authorities, not Whitehall.
New Combined Authorities and Strategic Authorities should take control of regional rail contracts, ensuring they are integrated with buses, trams, and local needs.
Local rail companies should be encouraged where viable. The Swiss model, where regions own small local railways, could be replicated in Cornwall, the North East, East Anglia, the East Midlands and in many other places I’ve not thought of.
As you’re considering governance models, just ponder the Cornish branch lines or the re-opening of the Ivanhoe line. If these feel distant from the place where decisions are made, we’ve not got it right yet. And don’t be fobbed off with regional structures or local HQs: the key question is who’s the customer? For services for which there isn’t enough farebox revenue to ‘buy’ the service, who’s paying?
Now, I know what you’re thinking: what about the interfaces? How do you allocate capacity between a regional rail contract (which wants to prioritise the city services) and a long-distance service? The answer is that it requires discussion and compromise. But if these are contractual with GBR, then it’s explicit. Putting everything into GBR doesn’t make those trade-offs go away: it just means that the decisions will be taken behind closed doors.
3. A Truly Independent Passenger Authority
Instead of the government appointing a passenger watchdog that reports to the Transport Secretary, it should be:
Statutorily independent (like the National Audit Office)
Funded through a levy on GBR and operators, rather than government grants
Given legal standing to challenge GBR in court
This would give rail users a powerful voice, independent of ministerial priorities.
4. A Contractual, Not Political, Relationship Between Government and GBR
The fundamental problem with the current GBR plan is that it gives ministers both operational and strategic control, leading to inevitable micromanagement. (Again, I know the consultation says that GBR is operationally independent, but it won’t be. Not when it does something that the minister doesn’t like).
There needs to be a clearly defined contract between government and GBR, specifying outcomes (e.g., punctuality, reliability). Again, we’re close to this with the HLOS and SoFA - we just need to complete the circle.
Conclusion
There are multiple ways to skin a cat. (So they say. Is that actually true? I’ve never tried to skin a cat but I feel sure there must be a best way…).
So there are multiple ways of doing railway governance.
But the conclusions I take from this little canter round Europe are:
It’s all about accountability. In a natural monopoly, who’s holding the railway to account?
One size doesn’t fit all: different parts of the network can have different answers to this question
Intercity networks can be operated commercially, with open access as one of the key drivers of accountability, and the farebox as the other
Central government should specify the high-level intercity network it wants for the economic success of the country: including the additions that wouldn’t otherwise be included in a commercial network (Newark, for example)
The new strategic authorities will have a key role in defining the services they want from their areas - but it’s crucial they have the budgets to deliver them
It may be that the previous success of Network SouthEast can be recreated: maybe with a joint DfT/GLA specification.
The trade-offs between urban, rural, intercity and freight services are real: but putting them all into GBR doesn’t make the trade-offs go away, it just makes the decisions less acountable
This is difficult and complex, and needs to be got right.
Don’t just create a single GBR bureaucracy reporting into DfT and expect anything other than gradual salami-sliced cuts. A natural monopoly, meddling and a lack of accountability will not be a good combination
This is one set of conclusions but others are possible. This stuff is complex and there will be more than one right answer.
Final thought
I really do want to emphasise that these are difficult, complex and important questions.
DfT, GBR, the Combined Authorities and the Strategic Authorities are going to have to work hard to get this right. They will make mistakes. Be gentle with them and give them the space to experiment, learn, refine and improve. It’s in all our interests.
👋 I'm 𝗧𝗵𝗼𝗺𝗮𝘀. I help organisations like yours drive 𝗶𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻, deliver 𝗰𝗵𝗮𝗻𝗴𝗲, and achieve 𝗳𝗮𝘀𝘁𝗲𝗿 results, drawing on 20 years of leadership across public and private sectors.
🚀 I offer 𝘀𝗽𝗲𝗮𝗸𝗶𝗻𝗴, 𝗺𝗲𝗻𝘁𝗼𝗿𝗶𝗻𝗴, and 𝗰𝗼𝗻𝘀𝘂𝗹𝘁𝗶𝗻𝗴 that energise teams, shape strategies and remove barriers to change. Whether you aim to accelerate innovation, drive change, or inspire your people, I’m here to help. Let’s talk!
