Not sure that's really true. Look at the low cost high speed carriers like Ouigo.

With rail, the physical rail becomes an economic chokepoint. Who controls the rail came name his price. Now you might think that the same goes for airports, but often there are multiple viable choices. E.g. in the bay area you could use SF, Oakland or San Jose.

I think this is probably the most accurate answer in this particular thread. There's essentially unlimited capacity in airspace, and while some airports are massively oversubscribed (LHR!) most are not. You can just keep putting planes up in the air in 3 dimensions where increasing the number of trains needs laying more excessively expensive physical infrastructure in 2 dimensions

I think it is actually the other way around and also not what GP said. Trains need more scale than planes.

Example: From Amsterdam Airport to Palermo, there are 2 flights per week (Friday and Monday). That is 300 passengers per week. It is a profitable to fly planes between these two cities, but a direct rail route including maintenance of the railway system would be far to expensive. This is partly resolved by stops along the way, but definitely not completely. The busiest direct railway routes (AMS-PARIS, AMS-LONDON, PARIS LONDON) in Europe are far from full, with one train an hour.

So the problem is not insufficient capacity, but rather that the scale is to small. I see electric buses playing a big role in the future to allow for better point-to-point connections.

just 2 directions, beyond that you can (like with airlines) go bigger and more frequent. After that the proverbial landing strips become the bottle neck.

Except "who controls the rail" is often a (private or national) entity (DB Netz / SNCF réseaux) who will have to offer access to the infrastructure to pretty heavily regulated and uniform conditions.

Much like the idea of separating ISPs from the last mile infrastructure by the way.

Even with the network controlled by a "neutral" entity limited capacity means you can pretty quickly reach a situation where additional competition can't happen without actively worsening the schedules of the existing operators.

Timetabled connections are likewise impossible to set up with too much uncoordinated competition, because without making the connection times unattractively long they cannot really be set up between more than one train per relation (maybe sometimes two at a stretch).

Plus connections that aren't ridiculously padded (and therefore result in absolutely uncompetitive journey times) really only work with some sort of through-ticketing so passengers aren't left stranded if they do miss a connection after all.

> Except "who controls the rail" is often a (private or national) entity

In Europe. But not in the US.

https://www.freightwaves.com/news/railroad/us-and-european-f...

http://www.mapattacks.com/2015/01/who-owns-americas-rail-inf...

Ouigo is a play by SNCF to avoid being undercut by a more aggressive low cost operator on the few profitable French lines. If you carefully look at their cost structure you will see that without the significant boost given to them by their parent company they would probably not be profitable.