>>1262893It's one of those ideas that's good in theory but doesn't bear out in practice. The main issue is that unlike airports and motorways, railway infrastructure and operations are highly dependent on each other. For example, on a highway, if you're in a fast car and get stuck behind someone slow, you can just pass them, no planning neccessary. On a railway, you'll need to plan that out in advance so that the blocks are clear, a passing loop is nearby, and that the switches are set correctly. If the other train driver, signallers, or schedulers screw up, the whole thing grinds to a screeching halt.
The papers I've read argue that traffic density is the main factor on whether or not it's a good idea to split infrastructure and operations of railways. On more lightly travelled lines, the increased competition will lower prices more than administration costs increase, so there's a net reduction in costs (assuming the line is profitable). However, on dense, highly travelled lines, any gains from competition are eaten up by the the bureaucracy necessary to coordinate with all the other companies.
One of the biggest problems is that Britain fractured their railways into WAY too many pieces. It used to be that the company that owned the tracks and the company that performed maintenance on the tracks were two different things. After Railtrack imploded, they reintegrated most of the infrastructure operations into Network Rail, but the train-side operations are still really fragmented, with the most egregious example being that the franchisees that operate the trains and the companies that own the trains aren't the same. They're looking at ways to roll this back, but the structure of Britain's privatization model (with it's time-limited franchises) has made it difficult, to say the least.
Here's a picture of the clusterfuck things were when Railtrack was still kicking. It's not as bad now, but it's still horrendously complicated.
