The nature of the beast. Guys like me prepare project cost estimates before they go to bid. The bigger the project, the longer it will take to construct, the more room for uncertainty there is, and the higher the likelihood the estimate will be significantly off. Far from a rail only problem, brutally expensive highway projects often run drastically over budget too (and there's no better way to get poor value for money than try to move everyone long distances in single occupant cars). Over multiple year (or several year) projects, prices escalate, unforeseen situations result in expensive change orders, a lot happens. My 2020 $4.5M estimated road cost over $6M and was completed nearly a year late. A city parking lot redo I designed was estimated and bid at $350k, and the contractor has since secured over $200k in change orders and counting. Shit does indeed happen. A recession should help material price wars f**k off, though.