Correlation does not mean causation. It's also possible that splitting and competition improves one railway but would make another worse. Maybe a railway is just organized well enough that splitting is not an improvement. If the timetable gets designed by the director's nephew, something needs to change. But if multiple competing teams produce proposals and the best one gets chosen, you already have the benefits of competing companies without the disadvantages.

Or there may be cultural differences between the railway companies. Maybe one company is staffed by hard workers who are motivated by working at a state-owned railway rather than for some capitalist that isn't them and would be offended if you introduced competition hoping to get them to work harder. The other could be staffed by lazy bums who find working for the state distasteful and do need a kick in the arse.

Forcing a split could also cause other reforms that outweigh the damage the split causes.