Friday, December 5, 2008

The world-wide web of misinformation

So I give a test that asks students something about Ricardian equivalence, which we had studied using two-period framework a la Fisher. We had looked in particular at the difference between the effects on the interest rate of changes in deficit-financed tax cuts (zero) and deficit-financed government spending increases (positive). Well a certain false sentence kept popping up in several exams - say 10%. I immediately suspected the web, and sure enough, in the Wikipedia article on Ricardian equivalence, I read:
Ricardian equivalence states that a deficit-financed increase in government spending will not lead to an increase in aggregate demand. If consumers are 'Ricardian' they will save more now to compensate for the higher taxes they expect to face in the future, as the government has to pay back its debts. The increased government spending is exactly offset by decreased consumption on the part of the public, so aggregate demand does not change.

This is not just wrong, it is flagrantly wrong. I'm sure everyone has their favorite examples of Wikipedia falsehoods, so I'm not saying anything new here. File under: The Wisdom of Crowds: NOT!

6 comments:

ProGrowthLiberal said...

In fact - RE stands for the proposition that transitory changes in government purchases will have little impact on current consumption in the same fashion that an Ando-Modigliani model would say that a one-time tax surcharge would be mostly saved. Now if the increase in government purchases were perceived to be a permanent innovation ...

TheTrucker said...

I think the Wiki article is correct and that Ricardian Equivalence is simply bogus.

1. Ricardian Equivalence takes no account of who (which class will eventually pay -- will the payments come from confiscated economic rent or from wages). And the class that receives the benefit of the borrowing may believe that the other class will be doing the paying. It might also be that the immediate beneficiaries are economically even more illiterate than me.

2. We see no appreciation for the actual future costs of the payments. If the borrowing will have increased productivity then the real cost of future payments will be less.

The bottom line is that roads and bridges are probably less "equivalent" and benefit from borrowing while (fake) wars don't.

kevin quinn said...

Trucker,

Your second point is interesting. It's true that RE treats government spending as consumption. If it is investment - say in infrastructure- then the the tax increase to pay for it in the future (to repay the bonds issued to run a bigger deficit in the present) is offset by the future benefits of the public investment. If the rate of return on public investment is the same as the private return then there is no effect on lifetime income and so no change in present consumption: the increase in Demand is equal to the increase in Government spending. If the rate of return on public investment is greater than he private return, then consumption will actually increase and the increase in demand will exceed the increase in government spending! Moral: chez RE, tax cuts won't, while government spending increases will, increase demand. And best of all are increase in public investment.

jsalvati said...

You'd better have fixed it.

Anonymous said...

Here's a silly story that illustrates the point.

Aliens in flying saucers come and offer us a deal. They will sell us the secrets of antigravity, unlimited conversion of mass to energy, impenetrable force fields, and some related technologies for $20 trillion. Or they will agree to $1 trillion a year forever.

Which is the better deal? We probably can't afford $20 trillion this year at all. But if we get those technologies and pay $1 trillion for ten years, by the time we've applied the new technologies for ten years we could pay a trillion a year without noticing it.

On the other hand, what if the US government didn't survive? Who would pay the debt to the aliens who surely have more advanced technology than they sell. Usually you can figure that a perpetual debt will eventually disappear. Like, at some point some of it will be held by iraqis, and we get mad at iraq and freeze their assets, and pretty soon we've spent the money to finance another war. Next year it's iranians, or syrians, or cubans, or whoever. Old debts go away in the dance of diplomacy, but not when the lenders are far more powerful than us. We'd better be real sure the aliens agree with our concepts of ownership of technology, or we might find we've sold them a lot more than manhattan island.

OK, here's a second story. The aliens come by and give us all this great technology for free. The US government happens to be fighting a couple of wars, and bailing out the banking system, and propping up relic industries, and paying for a new health care system. All this will cost only $5 trillion. Is it better to collect the money now, or gradually over the next 30 years? Gradually, definitely. Even if the money is entirely wasted and has no investment value, it's better to take a ilttle bit out of a bigger economy than a lot out of an obsolescent collapsing one.

There's something wrong with this logic. Can you spot it?








Don't look at the money flimflam game. Look at the resources. The resources that get used to keep obsolete US auto manufacturers going are resources that don't get used for anything else. The bankers who use hotel roooms in the caribbean, sipping their margaritas on the beach and watching the sunset, are taking up hotel space, margaritas and sunsets that somebody else could be using. The bombs we drop on afghanistan are bombs we could be selling on the open market for somebody else to drop on something more important. All of those resources are getting used up and are unavailable to use for investing in the new technology. Regardless how we pay for them, they're unavailable for better uses. It isn't the money that's at issue, it's the opportunity cost for the resources the money buys.

YouNotSneaky! said...

What jsalvati said. Well, I did anyway.