Michael Hiltzik's column advanced the idea that deciding what government should do ought to be based on the overall savings to society. I agree, but there are complicating factors.
It's much easier to develop an estimate of the cost of the government action than the offsetting savings. New Jersey Gov. Chris Christie's estimate of the cost to develop the railroad tunnel was straightforward, but the positive impacts would have to be based on vague factors and speculation.
Furthermore, sometimes it's better to pay a higher price for a more satisfactory outcome. It would probably be cheaper in the near term, for example, for the government to take over all energy development and production, but this wouldn't spawn truly innovative energy solutions. The impacts of these not-yet-invented energy sources cannot be estimated, and thus it is in society's interest to maintain private enterprise in this realm (though better regulated).