luckily it does solve it, because most of the cost is due to low volume.
construction in the US is very labor intensive and has low-productivity.
plus the permitting process is also very low-throughput. (too many possibilities to stop the process by appeal/suing, and the whole jurisprudence of what is considered a sufficiently detailed environmental assessment is fucked up. the law needs a bit of tweaking to focus more on cost-benefit, include cost of rejecting a project, etc.)
and this is also why nuclear power plants are super expensive.
I mean, there is correlation -- whatever we do spend on is extremely expensive, so we get less, whereas other nations that spend more efficiently get more, and therefore there is a correlation between efficiency of spending and quantity obtained, but that's not an argument that if we spent a lot more, we would get better value. A case needs to be made beyond this correlation.
A good example is bus service. One can't plausibly argue that US cities have no experience with bus service. Or that it's a difficult engineering problem (unless maybe, you are talking about trolleylines in the Swiss Alps) - but let's see how efficient they are. San Francisco spends 1.3 Billion on SFMTA, with 5700 employees and average weekday ridership (pre-pandemic) of 800K. That's 1 employee per 140 weekday riders.
Let's compare to the canton of Zurich, which has 1.5 million people, and 670 square miles (of Swiss mountains). VBZ (Verkehrsbetriebe Zürich) is responsible for all of the transit in this canton - bus, trolley, light rail. It provides rides to 900K daily riders (post-pandemic) that travel 1.7 million kilometers per day, but this is achieved with only 2300 employees, or 1 employee per 390 daily riders, and a budget of $560 million dollars (converted to USD).
Now maybe it's "like a nuclear power plant" and if San Francisco were to spend a lot more on bus service, they'd eventually figure out how to deliver it efficiently.
But my guess is that if San Francisco were to spend a lot more on bus service, they would deliver it even less efficiently than they are now.
So you still have a problem of inefficient delivery in the U.S. Then we can talk about why it cost the city over $300 million to create a single BRT line along Van Ness. Perhaps painting a lane red is like building a nuclear power plant, but I suspect that's not the issue here. https://www.sfchronicle.com/sf/article/S-F-s-Van-Ness-transi...
But it's very much like that. The approval process is extremely inefficient in the US. As you said, "building" the red line is not hard, but getting it through the permitting gauntlet is.
After a certain level of saturation these things just get easier. See road construction, expansion, etc. It's the default, yet it's clear that "just one more lane" won't solve traffic.
Regarding buses (and public transport in general): what's the utilization of the existing capacity for SF and Zürich? It would be interesting to compare number of buses, total distance driven, and other parameters for these rider numbers. Also do you happen to know anything about cost breakdown? (My guess is that Zürich is simply a lot more dense, and public transport is just the default there, whereas it's not in SF.)
construction in the US is very labor intensive and has low-productivity.
plus the permitting process is also very low-throughput. (too many possibilities to stop the process by appeal/suing, and the whole jurisprudence of what is considered a sufficiently detailed environmental assessment is fucked up. the law needs a bit of tweaking to focus more on cost-benefit, include cost of rejecting a project, etc.)
and this is also why nuclear power plants are super expensive.