It's Mankiw v. Krugman.
Let me ignore the adage that when elephants fight the mice get trampled and join in. This is from Krugman:
Which then raises the question: how can you believe that [if we could somehow get businesses to spend some of that cash, it would create jobs], and not also believe that if the U.S. government were to borrow some of the cash corporations aren’t spending, and spend it on, say, public works, this would also create jobs? (Brad DeLong has tried to make this argument repeatedly).
First, I have no doubt that a public works project could "create jobs". In an extreme case, the day shift could dig holes and the night shift could fill them in. At union scale, of course.
Now, one might argue that the discipline of actually showing up and working at the team-oriented task of digging or filling holes developed useful job skills for the participants. Then again, it might breed cynicism and skepticism, so let's agree that the social value of the dig/fill jobs program is zero. In that case, rather than "creating jobs" by way of useless public works it would be just as efficient to cut taxes for people likely to spend the money, or hand out money to random passers-by, or extend unemployment benefits.
Yet we won't find Paul Krugman arguing for the stimulative benefits of, for example, a cut in the payroll tax. Presumably he believes there are all sorts of worthy public works projects representing positive value-added that have never been undertaken because We the People have heretofore been too foolish to spot their value. Maybe! Or maybe the sorts of public works likely to be quickly undertaken are simply not that valuable.
From a different perspective, Prof. Krugman would fall down laughing at the notion that temporary tax cuts could induce permanent changes in consumer or business behavior. And most economists would join him - as recent examples, the cash for clunkers credit propped up sales until it lapsed, as did the home-buyers credit.
Yet Prof .Krugman is perfectly comfortable insisting that a temporary increase in spending will produce longer term changes in business and consumer behavior:
Spend now, while the economy remains depressed; save later, once it has recovered. How hard is that to understand?
Could he be right? Maybe - the key would be to disguise the spending, not so easy in this age of transparency.
The problem with temporary tax cuts is that "everyone" knows they are temporary. Stores don't restock after a successful tax-refund sale; factories don't expand, hire new workers, and run more shifts to meet the new orders. Everyone knows it is a one-time event.
Some public works provide the same transient demand and have the same zero-impact. However, certain types of spending are well-disguised. If a government work crew breezes into town to patch all the potholes or expand the library, merchants know the demand created by those workers is temporary. But a guy spending his unemployment check looks a lot like a guy with a regular job. And a policeman or teacher who is getting paid by way of a Federal grant doesn't look any different from a policeman or teacher who has had a regular job for years.
If Krugman is right that people can't figure out that the government-sponsored demand is transient, it may induce permanent changes in hiring or spending decisions. Otherwise, a government program widely understood to be temporary will have about as much impact as a tax cut understood to be temporary.