How should government officials decide on whether to fund big projects such as fighter aircraft, highways, bridges, and other types of infrastructure?

First, they should check the Constitution to see whether they are legally allowed to spend on the object in consideration.

Second, they should assume that the item will cost at least twice as much as initial estimates indicate. There should be a 2‑to‑1 hurdle when the price tag of a project is being considered.

Government purchases of military hardware, highways, energy projects, space equipment, and other items often cost 50% or 100%, or more (see here and here), above what politicians originally promise.

Let’s be conservative and say that a 50% cost overrun is typical, such that we can expect a new $1 billion project to actually cost taxpayers $1.5 billion. But as economists often point out, paying for $1.5 billion in government spending will cost taxpayers much more than $1.5 billion because of the “deadweight losses” or inefficiency costs created by extracting taxes from the private sector with a complex and high-rate system.

How much more? Harvard’s Martin Feldstein thinks deadweight losses might be $1 for each added dollar of taxes. But let’s be conservative and say it’s only 50 cents on the dollar. So government projects impose deadweight losses of 50% on costs that are likely to balloon at least 50%. 

The bottom line is that when America’s taxpayers hear that politicians want to spend, say, $10 billion on a new scheme, they should assume that they will face an ultimate financial hit of $22.5 billion. And that’s conservative!