As a big critic of those who claim that tax cuts pay for themselves, I have been hesitant to embrace the "government spending pays for itself" defense of using government expenditures to stimulate the economy. With tax cuts, there does appear to be some offset, but it is not large, i.e. on net, tax cuts increase the deficit by quite a bit (revenues of 10-15 cents on the dollar is, I think, a generous estimate of the dynamic effects). It's a theoretical possibility that a cut in taxes will pay for itself, but the empirical evidence just isn't there.
The people making the claim about government spending are careful to note that it only applies to very depressed economies, and government spending paying for itself in the long-run is a certainly a theoretical possibility in that case. And events in Europe are suggestive that a careful empirical investigation will support the claim that the opposite, austerity in a depressed economy, can worsen the long-run budget picture. Some things do pay for themselves, and then some. A business investment had better do that, or there will be no profits (and infrastructure spending can have similar properties). So I don't mean to imply the claim is outlandish per se, not at all. It may well be true. I just don't have the evidence I need to make a strong statement.
One more note. I also don't really like the framing. It's not what's intended by those making the argument, and they are careful on this point. But careful or not it makes it seem as though the spending is only worthwhile if it does, in fact, lower the long-run deficit. But there are benefits to, say, putting people back to work that are not captured by this sort of expenditure/tax-return analysis. The question for me is whether, on net, the spending generates more benefits than costs (including all economic costs, not just the accounting costs). On that score, I have no doubt that more spending would be beneficial. That's Paul Krugman's conclusion too -- see below -- and more to the point with respect to what follows, I have no doubt that the opposite, austerity, would be harmful under present economic conditions:
The Future Is Another Country, by Paul Krugman: Simon Wren-Lewis has another very good blog post, this time on the very weak case for demanding fiscal austerity right now, even if you believe fiscal consolidation is needed in the long run. I want to tie this together with Brad DeLong’s preview of DeLong-Summers on fiscal policy in a liquidity trap.
What Wren-Lewis says, and Delong-Summers argue at greater length, is that there is very little direct connection between spending over the next couple of years and long-run fiscal prospects. Between low borrowing costs and high multipliers, spending now would do little to worsen the debt picture a decade from now; indeed, as DeLong-Summers argue (and some of us have been arguing, less formally, for a while now), there’s a plausible case that spending more now actually improves the long-run fiscal picture — and that austerity worsens it.
So what’s the counter-argument? Aside from deeply flawed arguments about crowding out even with zero interest rates, what you usually encounter are assertions that there is no such thing as a temporary spending increase — and the related claim is that spending cuts now will signal spending restraint over the long term too. Such statements are invariably made with an air of worldly wisdom — hey, we all know that stimulus spending never goes away when it’s supposed to.
But do we know that? Actually, all the evidence says the reverse. Did the Works Progress Administration survive into the 1950s? No. Have the provisions of the Recovery Act been made permanent? Not a chance.
Nor, as Wren-Lewis says, is there any good reason to believe that fiscal austerity now will persist when times are better. A prime example is what happened to the de facto austerity of the 1990s, when spending restraint in the face of a booming economy finally moved the US government into surplus. Without a moment’s hesitation, George W. Bush rushed to squander that hard-won surplus on tax cuts and an unfunded war.
To a first approximation, in other words, the effect of current fiscal policy — whether stimulus or austerity — an the actions of future governments is zero. And what that means is that stimulus and/or austerity should be considered on their direct merits, a comparison that makes a strong case for stimulus while revealing austerity as a (literally) unmitigated disaster.