However, this is not a problem of separating infrastructure and operations. It’s a problem of running trains on the territory of another company, which prioritizes its own operations. The point of separate ownership is to avoid problems like this, so that tracks will be owned by a neutral company.

‘Come to think of it, they separated network infrastructure from operating companies using that infrastructure in the California electricity market, in order to worship at the cult of the abstract market, and it all went horribly wrong with Enron.’
They did it in NYS, too, and it’s working just fine….

The key difference is that they did it right. The NYS market regulator didn’t just abandon regulation of electricity prices when they separated infrastructure costs from generation costs — they watch like a hawk for potentially manipulated prices, and they will step in and order fixed prices whenever they see a market failure.

They’ve also been implementing a very slow rate transition — formerly the per-kilowatt charges were subsidizing the infrastructure *very* heavily, making a hidden subsidy for extending electric service to tiny rural hamlets, and a hidden penalty for efficiently located houses and industries. The shift to a larger fixed charge for connection and a smaller variable charge for generation is part of a slow scheme for redressing this balance, and should *eventually* lead to it being cheaper to put your house near a power plant than far away from it. (Reasonable, right?)

“The argument in this piece is dead on target – the problem with the real world version of market being artificially constructed is that the financial success of the infrastructure company becomes walled off from the success of the operating companies in delivering transport benefits to passengers.”
Well, if the appropriately-subsidized operating companies are paying the infrastructure company to provide appropriate levels of infrastructure — and the monopoly infrastructure company is both setting prices correctly *and* keeping its accounting honest — both big ifs — then it should be possible to make the success linked to the success in delivering transport benefits.

The thing to avoid is a lunatic scheme like Railtrack. Network Rail in the UK appears to be performing tolerably.

I’m actually not sure about that — the record shows that private “vertically integrated” railroads routinely played bookkeeping games where they underfunded their infrastructure. Separating it out may be a way to guarantee full funding for infrastructure, or at least to make it clear and obvious when it’s being neglected.

The infrastructure *must* be publicly owned, and the carriers should probably remain publicly owned *too*. Administratively, however, it may make real sense to have them as separately accounted subdivisions.

Note that Amtrak already operates on infrastructure owned by others through most of the country. Where this infrastructure is owned by state or local government agencies, it is usually maintained fairly well; where it is owned by private companies, they have traditionally done their best to mistreat Amtrak, though this attitude has changed at Norfolk Southern and BNSF.

Proven to be a bad idea in commuter rail operations. DB continued to use them after their consultants advised them there might be a problem and a commuter agency using them warned them they failed earlier than expected.

“DB has skimped on maintenance for decades”

That’s a rather sweeping statement. Afaik, the skimping started just a few years ago when politicians decided they wanted to float shares in DB on the stock market. The exception was the Berlin S-Bahn, which was hemorraghing money. Still, skimping on safety is always a braindead idea.

“leading to the Eschede disaster”

Incorrect. Eschede was caused by a decision to retrofit perfectly good but stiff monoblock wheels on first-generation ICE trains with a design that featured a steel core, a rubber ring and a steel rim around that. The idea was to improve passenger comfort while the ICE ran on DB’s legacy tracks at 125mph. The damping ring was a proven concept in commuter rail operations, but it failed catastrophically in this high speed application. Since then, no-one anywhere in the world has been crazy enough to use anything other than monoblock wheels on high-speed trains.

The national rail companies are already pseudo-private in France and Germany, in the sense that they’re expected to run a profit; SNCF is also politically independent, the hallmark of good private sector service. Conversely, both SNCF and DB have hurt their service by trying to minimize costs, all while enjoying a monopoly on rail service. DB has skimped on maintenance for decades, leading to the Eschede disaster and this year’s recall of the Berlin S-Bahn’s rolling stock. SNCF took shortcuts with TGV station placements, creating unusable beet field stations.

Splitting infrastructure from service won’t worsen the existing problems. If anything it could force DB to build better trains, since passengers will be able to choose the safer TGV.

The argument in this piece is dead on target – the problem with the real world version of market being artificially constructed is that the financial success of the infrastructure company becomes walled off from the success of the operating companies in delivering transport benefits to passengers.