The basic notion is that the way non-experts think about the work has changed in ways that have drastically increased system inefficiency.
I know zip about construction megaprojects. But this is definitely true in software. Many executives want to pretend that projects are perfectly predictable, which a) diverts a lot of effort to non-value-creating outputs (fantasy plans, fantasy schedules), b) creates huge inefficiency by favoring fantasy plans over what's actually happening, and c) creates plenty of rework and chaos trying to solve the real-world problems when the gaps between fantasy and reality become too big to ignore.
To me, this problem is mainly rooted in executive culture, and is related to the phenomenon some call managerialism: https://en.wikipedia.org/wiki/Managerialism
But few have the skills to estimate what any part of nuke should cost.
https://www.orfonline.org/expert-speak/chinas-high-speed-rai...
Note that South Korea and France have nuclear cheaper than renewables still because their processes over decades have led to basically producing nuclear plants with horizontal scaling, driving costs down dramatically. In essence, all energy production within the borders of a country is a matter of political will moreso than even technical requirements. If a culture wants to push back on X and wins politically, it becomes more expensive and distorts intrinsic costs. If a culture pushes for Y and eventually subsidizes it, it becomes dramatically cheaper even below intrinsic costs and is over-consumed (see: fossil fuels)
The main differences in costs and time from 150 years ago are:
(A) Accounting for Extrinsic Costs. as the nations have become more dense and more democratic, peoples voices need to be heard. Environmental, wildlife habitat, etc. are complex issues that have only become important in recent decades. People sometimes get tempted by authoritarian rule because they “get shit done cheaply”, but this is a ruse. By not engaging openly, there are often massive extrinsic costs, so the project is “cheap” by cheating and not counting the true costs. Europe seems to have streamlined this, the USA is still stuck in “might over right” political power struggles rather than compromises.
(B) Cheap labor. Most infrastructure doesn’t require large numbers of high skilled labor, it needs a mix of skilled and unskilled labor. The railways in the 1800s we’re built with large quantities of immigrant labor. This is in rarer supply in the West these days, for a variety of reasons.
(C) Labor rights and safety. Construction projects and rail projects were happy to bury dead workers by the wayside (or in the concrete of the Empire State Building). Or work them to unsustainable hours with limited pay. It took many strikes, some violent, to sort this out. Today’s standards don’t put up with that level of labor abuse.
(D) Investment sources. Many large projects were state funded, or were funded by investors with a lot of patience. That kind of patience is in short supply, as is state funding in the USA due to Republicans being Republicans. Europe seems to be able to pull this kind of funding off.
(E) Contract Grift. There is a lot more incentive and skill to slow roll and milk the cash flow of construction projects, given their complexity. Traditional large works were more vertically integrated and managed by those with incentive to see the job done, rather than paid by the hour. The solution to this is more vertical integration but that’s hard to pull off because of labor shortages.
I’d also note that China has a vested interest in publishing propaganda about its engineering and public works prowess while hiding its extrinsic costs.