On rail the largest costs are building and maintaining the rail until you get to very high frequency. For most rail in the US the largest factor in maintenance is weather and so you could run a lot more trains without changing the costs much. You do need to buy more trains, and they will need to be cleaned, but it wouldn't be hard to get enough to people on board to pay those incremental costs. (in the US the bottleneck is often an expensive tunnel that is shared between several not busy lines, each line could itself handle many more trains all day than they have at the peak without changing maintenance costs - but the tunnel is full and it costs too much to build a new one - this is why so many in transit are focusing on construction costs - if we can build a short tunnel we unlock a lot of better transit)

▲

reaperducer 3 days ago | parent [-]

On rail the largest costs are building and maintaining the rail until you get to very high frequency.

In the cities I've lived, it's not quite that.

Building rail is a lot of dollars, but politicians are often happy to throw money at that problem. It's good for a dozen industries, like construction.

But that money cannot then be used to operate the rail long-term. That burden is on the city entirely.

I've lived in two cities that turned down millions of dollars in federal transit grants because they didn't have the money for maintenance and operation.

	▲	bluGill 3 days ago \| parent [-]
		Capital costs are paid differently and so we can often safely ignore construction costs as they are a different budget. Putting maintenance into that capital bucket though is accounting fraud and illegal. If you have a rail line the largest cost is regular maintenance which is based on time not wear until you have a very large number of trains running. So my point stands even if you separate the buckets.