David Leeder's article on reform

In 2011, the McNulty Review was published to broad acceptance, whilst conspicuously avoiding the cost incentive issues that would have been obvious in any other industry. Such easy acclaim should have been a warning sign, and four years on, railway costs have grown to the point where Network Rail is now being micro-managed by DfT appointees.

In retrospect it was obvious that crucial reforms that one would have expected McNulty to address had been diluted or evaded, even before publication. But the industry then got to work, deploying the tried and tested strategies of producer interests everywhere – watering down proposals, delaying reforms and promising efficiency miracles in the distant future.

So who killed McNulty's plans for efficiencies? As in Agatha Christie's famous railway novel, 'Murder on the Orient Express' all of the suspects appear to be guilty, because all of the economic actors benefit strongly from the status quo.  

The TOC-owners in general have neither the balance sheets nor the project management experience to countenance significant investment in infrastructure. The ROSCOs have quietly benefited from the structural shortage of rolling stock created by rapid demand growth, and their then private-equity owners were much more focused on cashing out than taking part in risky infrastructure plays. Importantly the industry's executives and frontline staff continue to benefit significantly from the current arrangements.

Most importantly, Network Rail remains structured as the economic equivalent to the Doomsday Machine in Dr Strangelove – the fiendish nuclear device that, once activated, cannot be controlled and cannot be neutralised.

Railtrack plc as created by the Major government was already a strange hybrid – a private monopoly that depended on subsidised customers for its revenues, and which was politically unloved even by a "Blairite" Labour Party that had been campaigning for rail nationalisation since before the First World War. Crucially, the Tories’ model of Railtrack earned its income from track access fees and commercial income, such as property development.  It was not intended to receive government grant. Indeed, it was expected at the time of its creation that an increasing proportion of total train mileage would be provided by open access operators, or unsubsidised franchisees. Therefore Railtrack may have taken on a more normal commercial character over time, perhaps comparable to that between the regulated airports owned by the privatised BAA plc and commercial airlines. But this was not to be.

The incoming Labour government's first decision was to politicise Railtrack's regulators, creating not one, but two hostile bureaucracies in the Tom Winsor-era ORR and the expansionist Strategic Rail Authority, who then proceeded to fight a three way battle between Railtrack and each other. 

Labour next introduced direct grants to Railtrack. This was a fundamental change, as much for its political impact as its economic one. In practice, the company probably had little choice but to accept the King's Shilling (the monarch in this case being John Prescott MP), but from this moment, the company was probably doomed as a private entity.

The sequence of fatal derailments and collisions that followed in the early 2000s escalated the war of words still further, until the combination of political and regulatory hostility was simply unsustainable.  Railtrack plc became a ward of the state and was replaced by Network Rail. 

At this point in the evolution of the Blair government, the idea of a "Company Limited by Guarantee" (CLG) seemed to be an answer to a much broader question about how a post-Marxist Labour party would rethink how public utilities could be structured. A CLG would be simultaneously public and private, and would be “off-balance sheet” for HM Treasury. And there would be no vulgar shareholders, grasping for dividend, which solved a political problem and seemed to solve a financial one. Like most apparently new ideas, it was actually a very old one – a revival of the 1930s concept of a 'company owned by no one and accountable to everyone' that had led to the creation of both the BBC and the London Passenger Transport Board, as the first iteration of TfL, in 1933. 

No doubt had NR been a more immediate success, many other privatized utilities would have suffered the same fate as Railtrack. As it was, the financial problems of NR were evident from an early stage, even to those ideologically disposed towards corporatist solutions.  In 2003 the smart question was "who owns Network Rail?' It took just over ten years before the question received a simple answer – Her Majesty's Government. By 2015 the fiction that NR was in any way independent of government was no longer sustainable. 

What we have now is therefore the product of both a century of political hostility to private investment in railways, combined with a series of short-term structural expedients since 1996. 

If the cleverest people in Britain had been specifically charged with the task of designing a machine to waste money, it is just possible that they would have come up with the current incentive structure. NR’s tortuous history has left it with a combination of economic incentives, all of which exist in other organisations, but which have rarely been combined to such damaging effect. 

Any serious attempt at improving cost efficiency will have to move beyond producing another well-written headmaster’s report on the theme of “Why You Must Try Harder”.

The starting point must be recognition that no fundamental reform is possible without tackling the very flawed economic incentives currently in place. I will summarise just six (but there are others):

1. NR’s customers are indifferent to price

The track access charging regime agreed in 1996, and only incrementally modified since, leaves the TOCs largely indifferent to the level of track access fees. If costs increase, subsidies go up (or premia go down), and vice versa. 

This has encouraged a sense that track access costs are "fixed", whereas the 120 preceding years of railway history have shown that operational as well as infrastructure changes have a direct impact on operating costs. This suits the TOC owners, and leaves DfT and ORR advocating efficiency – a far weaker pressure than would exist if the TOCs’ commercial lives were on the line.

2. Network Rail has a RAB

A Regulated Asset Base is a common device in privatised monopolies. A RAB is intended to allow transparent agreement about what assets the entity requires, and what they cost to build and maintain. It also acts as a direct incentive towards capital investment, because the higher the RAB, the higher the allowable profits that will derive from the regulated “X percent” profit margin set by the regulator.

Any utility with a RAB (eg the water companies, or the larger airports) will spend a great deal of time "talking up the RAB" – ie trying to persuade a skeptical regulator that they really need that Richard Rogers-designed Terminal, and not an off the peg shed of the kind that works perfectly well for Tesco (or an unregulated airport like East Midlands). 

But they also have greedy private investors who have a strong incentive to beat the agreed plan, and take the savings as profit in the relevant regulatory control period. A RAB without shareholders is like a bank without a locked safe.

3. NR is deficit financed

Now imagine the bank with the doors wide open and a sign saying “out to lunch – back in 5 minutes”. We have returned to a situation where the difference between NR’s actual spend and its income is the subsidy – the kind of ‘deficit funding’ used to prop up British Leyland in the 1970s. And performance fines become useless – the government in effect, fining itself.

4. Network Rail has no private shareholders

To believe in NR, you really have to believe that a government department is better at managing costs than someone like Brian Souter whose personal wealth is at stake. Since the 1960s, very few people have justified state ownership on the basis that it will lead to lower costs. 

5. Network Rail can't be benchmarked

There's only one NR. And although there other track authorities do exist in the UK (the London Underground, Northern Ireland Railways, the urban subways and light rail schemes) all of them are to some extent special cases, and none seem keen to fall under the economic wing of what was the ORR. 

This makes benchmarking extremely difficult. Without real-world comparators, cost benchmarking – intrinsically complicated – becomes an interesting academic exercise that NR is almost guaranteed to win.

6. Network Rail is "too big to fail"

Railtrack conclusively proved that a single-network operator was “too big too fail”. But it may also have shown that such a large entity was too large to be investible. Many investors around the world are seeking infrastructure opportunities, but NR as a combined entity is both too large and too politically risky to access this kind of capital.

The current reviews do not need to solve all of these problems simultaneously, but any review that fails to tackle the fundamental economic factors is unlikely to deliver significant medium-term improvements. A reform package that does not provoke opposition from the beneficiaries of the current structures is probably too cautious to deliver the major change that is needed.