A hard-to-capture cost factor is keeping highly used services operational or replaced while doing work on it. A super-fancy new subway with lots of stations in Shenzhen where people weren't already taking trains and expecting to use stations for their daily commute will be much cheaper than disrupting a line service to open a new station in London or New York City. Replacement bus bridge services, network-wide service advisories, staff and enforcement training, are all costs that are not easy to quantify or justify because it doesn't directly manifest in the construction of the infrastructure.
Of course there are other factors that vary a lot across cultures:
* China's heavy-handed governmental decision-making leaves no room for inefficiencies via neighborhood antagonism towards the constructions or zoning;
* Car culture differences create different magnitudes of popular roadblocks to investment in non-automobile infrastructure, which can result in large expenditures on lobbying and campaigning;
* Significantly more arbitrary middlemen making their way into the process because of "let the market figure it out" does, in fact, create some process inefficiencies in supply chains and contract pipelines. Few countries are as aggressively market-based for things meant for public good as the US.