2-node-supercomputer.net 2022-05-08
Most people probably have heard that California is building one swell set of tracks of high-speed rail (HSR) between Madera, CA and Bakersfield, CA, eventually to connect San Francisco and Los Angeles.
Most may also have heard how much it costs… (a lot!)
The shit show (sorry, but using a more polite word didn’t seem polite to reality…) that is California High-Speed Rail cost overrun seems utterly nuts. Initially in the ballpark of $38 billion for the whole network, now Madera to Bakersfield alone is supposed to be $100 billion. Why?
Getting good answers, or any answers at all, is difficult. Some say it is the property acquisition, the lawyering, the delays caused by acquisition delay, feature creep like nicer stations, grade separation everywhere, massive protective walls, more grade separations, just plain cost misestimation, risk mismanagement, politics, or maybe all of the above.
One reason I have heard recently repeated on different scales again and again is much simpler: The US doesn’t have the expertise, yet. This may be true on several levels: from the bureaucracies in government to the companies building it. Expertise must be built up through experience, and the last time the US built a lot of rail is about a hundred years ago. Some of that expertise is now lost, and what isn’t lost needs updating for high-speed rail.
Thus, a big portion of the cost overruns are likely developing the technology and management skills to build HSR. Judging by the fancy grade separations that are being built, I wouldn’t be too surprised if this will lead to significant technological learning. The US also needs to learn to manufacture all the rails, HSR-quality.
Sure, the US isn’t starting from scratch like a third-world country, so the other problems probably contribute significantly as well. However, if this technological learning is actually taking place, then the Texas HSR project, and indeed all HSR projects throughout the country would benefit from California’s investment into HSR.
Is technological learning actually happening? After all, you’d think the US already knows how to build bridges and such. How different can HSR be? Maybe the companies are just learning to extract more money from the government?
So how can we increase the technological learning rate? For one, contractors need to fear their competition. The CA HSR authority needs to be able to quickly cull individual contractors. If you only have very few contractors, then you cannot afford to loose any, so they can demand almost any price and deliver as slowly as they want without much consequence or learning happening. That means, we need more contractors, which likely means smaller ones that have the potential to grow.
I leave it as an exercise to the reader to keep developing the strategy.