Vertical fragmentation on the railways doesn't work
From CroziervisionWiki
Vertical fragmentation on the railways is the separation of infrastructure and operations (or the ownership of infrastructure from its maintenance.) In other words, the running of the track and signalling is separated from the running of the trains.
It doesn't work. Or perhaps I should say it doesn't usually work. There are exceptions. For instance, the London and Greenwich Railway didn't run a single train on its tracks from the 1840s until it was absorbed into the Southern Railway in the 1920s.
There is also the sort of exception of one company allowing another to use its tracks. This happens all over the place, especially in the UK before 1948. However, it's a rather odd example because all companies also ran their own trains. A piece of track was for the most part only used by the trains of the company who owned it.
But the overwhelming evidence is that it doesn't work. The very first proper railway, the Stockton and Darlington began by allowing anyone to use its tracks. Very quickly (within months) it realised this didn't work and decided to take control of both sides of the business.
In the EU, in 1991, the EU enacted Directive 91/440. This demanded fragmentation. Each country implemented this fragmentation in a slightly different way. Nowhere did it work particularly well (see Carlo Pfund.) However, the results varied quite a lot. It seems that it was worst where tracks were most intensively used ie in Britain and the Netherlands.
An example of the sort of problems that vertical fragmentation causes is the Southern Region new rolling stock programme. The TOCs through their ROSCOS ordered new trains. But they were a lot heavier and needed more power. But no one thought to tell Railtrack. Result? An emergency power supply upgrade and (possibly) more expensive maintenance.
Some of the TOCs (such as Firstand South West Trains) have called for an end to vertical fragmentation. As have Network Rail. And one of the politicians who voted for fragmentation in the first place. South West Trains has introduced some degree of pseudo-re-integration in the form of an integrated control centre.
Why doesn't it work? Well, we don't know. After all roads and the airline industry work perfectly well with a fragmented structure. But there's something different about railways. It's probably something to do with their greater stopping distances which means you have to have signalling which (I think) means you have to have timetables which means you have to someone in charge.
It's probably to do with the nature of railways as a system. For instance, if you are thinking of buying some new trains the decisions you make there have all sorts of impacts for the infrastructure. If they are faster or heavier the infrastructure may not be able to power them or to take it. Heavy trains can bash hell out of the track. If you want to run longer trains you may have to think about lengthening platforms. If you want to run faster trains you often have to completely redo the signalling systems (longer stopping distances, see?) In a fragmented system the buck gets passed. In an integrated system it stops with the Fat Controller.
