Greg Mankiw is a widely and justifiably respected economist. But sometimes in his blog he prattles on just like the rest of us, absent the empirics that he should be bringing to the party.
So here. He ridicules the CBO for even bothering to analyze a temporary increase in food stamp benefits to effect short-term fiscal stimulus. His reason? Vehicles for heating and cooling the economy should be symmetrical. You wouldn't cut food stamps to cool off an economy, so why would any ninny think of increasing food stamps to stimulate an economy?
I know—doesn't really pass the sniff test, no matter how much it’s couched in language from a particular corner of “standard macroeconomic theory.”
Credit where due: in an update he posts a reply from Jason Furman that makes a damned good case for the idea.
- It’s way easy to implement. (Push a button and recipients have more credit on their cards.)
- There's basically no dead weight loss/friction—the money goes instantly to people who spend it.
- They will spend it immediately, resulting in quick stimulus.
- Furman doesn't mention this triviality, but the money would also go to those who suffer most from recession.
Mankiw can’t really contest such cogent arguments, so he makes a quick shift from WMD to spreading democracy, in the form of compassion for the poor folks whose food stamp benefits would be increased. He proposes instead a variable corporate investment tax, adjusted up and down to heat and cool the economy based on economic conditions.
Accountants and tax lawyers, are you rubbing your hands at that prospect? ("Let's see: should we accrue that investment on March 3rd or March 4th...") Can you say "audit nightmare"? Does Mankiw think this is going to be implemented in the next month, or six?
I'm thinking that the real basis of Mankiw’s objection—what's causing him to be less than reasonable here—is the bit about how tax cuts for the rich result in so much better incentive-to-cost ratio. I'm putting thoughts in his head without his permission, but this is my blog so I'll continue.
Arthur Laffer explains the idea here, in a December interview that's well worth reading for its surprisingly conspicuous quantity of centric sensibleness:
The top tax bracket he [Kennedy] cut from 91% to 70%. That's a 23% cut in tax rates with a 233% increase in incentives. That's a 10-to-1 benefit-to-revenue-loss number.
Now you take the guy in the bottom bracket. The guy in the bottom bracket had a 30% cut in tax rates, from 20% to 14%, and he had a 7.5% increase in incentives. So that's a 1-to-4 benefit-to-revenue-loss calculation.
That's very persuasive theory. But it doesn't have much to do with using food stamp benefits for immediate fiscal stimulus.
And Mankiw’s argument from compassion:
I can more easily imagine, when the economy starts to overheat, telling corporations that their investment credit has shrunk or disappeared than telling poor families that their food budget has been cut.
Doesn’t have that ring of empirical authority (or authenticity) that he’s so well-positioned to provide.
Update 1/20/08: The only thing that’s changed is my own thinking. While it’s tempting to delete this post out of mild embarrassment, instead I’ll add.
First, the imputation I made re: Mankiw’s true reasoning doesn’t make sense. I advise (beg) readers to ignore it.
Second, here’s the real flaw in his "symmetry" argument: the two situations aren't symmetrical. An overheated economy is banging against the bounds of its own capacity. An underheated economy is not. Since the situations are not symmetrical, the responses should (or need) not be.
I continue to maintain that his arguments about the poor abused food-stamp recipients don't make sense. They'll be delighted by the extra money, and disappointed that it doesn't continue indefinitely. But that type of disappointment will always be with us.