Very very few very big projects come in even near the original estimates, whether public or private. That's not the estimators' fault, usually, though there are exceptions when the estimates are politically driven. Large tunneling projects are among the worst, historically. The problem in real life is that large public works projects need large budgets, and involve public funding sources, which are subject to politics, so there's a large incentive to take the low end estimate rather than the high to get it approved because the higher estimate doesn't look good. When (not if) the overrun happens, well, it's not the politicians' fault (probably long gone by then) is it?

The only ways to avoid the overrun are to either never do a big project (which many people prefer, regardless of what its usefulness is), or to use the high-end estimate which may be wrong (sometimes it's too high, resulting in a tiny bit of money back) but is probably realistic. Privately-funded projects are seldom so big and complex as to be truly wicked, and can be more realistically estimated and controlled - even so, some are abandoned or otherwise don't turn out well, which can affect stock price ...

Contractors for large public works projects are talented at using every dime available and more besides (you called to check on progress - CHANGE ORDER!). That's not, usually, against the law. Expect and budget for it.