Comments on: Railtrack attack

By: Matt

Matt — Thu, 07 Apr 2005 15:19:00 +0000

It’ll be interesting if the Adam Smith Institute are as keen on public bailouts in the case of Rover as they are with Railtrack.

By: john b

john b — Wed, 06 Apr 2005 13:05:00 +0000

SBB is state-owned. UK railways were nationalised (alongside most of the other things nationalised in 1945-50) because they were bust, although some pro-privatisation types claim that they were only bust because the government screwed them over during WWII.

By: dsquared

dsquared — Wed, 06 Apr 2005 12:52:00 +0000

they'd (a) make damn sure it happened

I happen to know one hell of a lot about the West Coast Main Line (it is/was a cause celebre for Plaid Cymru) and no they wouldn't. The WCML upgrade is a more or less intrinsically intractable task. The problem is the section between Chester and Holyhead (which is massively important to the economics of the whole project as this is a main goods route to Ireland). There are dozens of small tunnels and the rails run along the North Welsh coast on terraces cut into mountainsides in some very difficult geology (slate). British Rail didn't upgrade this line for a reason, and the reason is that nobody really knows how to.

Switzerland doesn't have "private railways", does it? SBB is owned by the cantons, isn't it? If there are any private railways in Switzerland I would suspect that they are funiculars or two-point lines, which are hardly the same sort of thing as network rail.

By: john b

john b — Wed, 06 Apr 2005 12:51:00 +0000

I think Dan’s point was that since Virgin’s management are less insane than the Railtrack management, they would never have signed up to something as mad as Railtrack’s "let’s assume a new technology is magically invented that justifies our assumptions" WCML modernisation plan.

By: Peter Clay

Peter Clay — Wed, 06 Apr 2005 12:43:00 +0000

d2: You're right about the problems of side-payments and byzantine contracts being no substitute for markets. The bizarre privatisation structure guarantees inefficiency as there isn't enough choice in the micro-markets for it to work. Railways are naturally vertically integrated.

I disagree with you about Virgin basing its business plan on a capital investment: if they were making the investment and doing it themselves, they'd (a) make damn sure it happened and (b) quite happily go to investors basing a business plan on it working.

Private railways used to work in the UK. Why were they nationalised, other than ideology? They seem to work OK in Switzerland, too.

By: Nick

Nick — Wed, 06 Apr 2005 10:54:00 +0000

Not much of significance to add, but I like Gregg's maybe intentional apposite homonym:

historical under-investment is the route cause of many problems

By: dsquared

dsquared — Wed, 06 Apr 2005 10:53:00 +0000

The burden of my rather cryptic comment above is that there is an "information bottleneck" in the planning problem when you have the track owned by a different entity from the one running the trains. The planning problem is one of

a) maximising efficient use of the network
b) minimising the cost of repairs
c) reducing to an acceptable level the risk of catastrophic failure
d) creating a fund to pay for future improvements

It's very difficult to solve all these problems simply by the use of the price mechanism. Side-payments and legal contracts help much less than you'd think, because they introduce huge numbers of opportunities for strategic behaviour; it's all very well for Virgin to base their entire business plan on the WCML improvement, but they wouldn't have done that if it was a project that they were carrying out themselves, rather than one the failure of which they could be compensated for out of bottomless pockets.

It's to solve coordination problems like this that firms exist in the first place. I don't know what Ronald Coase actually thinks about this, but I'd suggest his theory points to either private monopoly or nationalisation.

By: Gregg

Gregg — Wed, 06 Apr 2005 10:38:00 +0000

Of course, historical under-investment is the route cause of many problems. But attempts to reverse the under-investment are negated by the presence of the private sector. The problems in rail were not caused by market failure or working practices or poor management, but, as you say, by decades of under-investment. In those circumstances, all that was needed was greater investment. Involving the private sector was a barrier to that, as the investment needed was bled away.

The same is happening in health - the government's massive spending increase has been blown on PFI, the money barely touching the sides as it was sucked straight through the NHS and into the pockets of big business. Having private companies run cleaning services in hospitals has been a diasaster. Having them build and operate new buildings in the health and education services has already lead to massively inflated capital costs, and will lead to operating costs being much higher than they would otherwise be for the next thirty years.

It is also happening in education, a variation on the theme is happening in the fire-service, in fact across the board of the public sector the government is tying spending increases to "modernisations" of this type which render those increases utterly useless. The only way to solve the problem of under-investment in the public sector is, frankly, to throw money at it, but the private sector, which the Thathcerite dogmatists insist must be involved and must be followed, is just getting in the way, grabbing that money before it can reach the target.

By: john b

john b — Wed, 06 Apr 2005 10:35:00 +0000

Correct, and yes, respectively. Please can you mail me at gmail (john dot band) from whatever account you’re using at the moment…?

By: dsquared

dsquared — Wed, 06 Apr 2005 10:29:00 +0000

On the subject of competitive structure for railways, I think that the fundamental one is a cost accounting and overhead allocation one, and a good textbook example of what’s wrong with marginal economics. There is always pressure to set MC=MR in pricing the network to operators, and the marginal cost of allowing one more train to run over the rails is close to zero as makes no odds. Rail track maintenance doesn’t come in continuous functions; it comes in discrete units, and there is thus a more or less intractable problem of creating a correct "fully allocated" cost of running a train on a track, let alone a differentiable one that would allow classic Marshallian price-setting.