Friday, May 28, 2010

The Stupid!

If you read Mark Thoma's blog, you will have seen his passionate response, called "Modern Macroeconomic Theory and Fiscal Policy" to some self-styled "modern macroeconomic theorist" who claims that people like Brad Delong and Krugman who think that fiscal policy can help us out of a recession when the interest rate is at the zero bound are not modern macrotheorists and so are obviously wrong.

Thoma and Krugman's response has been to cite papers by DSGEers in which a version of the model does allow fiscal policy to work - papers by Eggertson and Woodford and others of that ilk. Read the two comments by Robert Waldman to be reminded of why he- Robert - is a national treasure. Waldman says the DSGEers cannot be presumed to have explained anything about the modern macroeconomy at all, so the idea that we have to reject the idea that fiscal policy is efficacious if a DSGE model can't show it to be is a nonsensical non sequitur.

I've looked at the latest Eggertson paper. They are all -all the fiscal-policy friendly DSGE papers, I mean- alike in making the mechanism by which fiscal policy works in a liquidity trap this: it increases expected inflation. This is not the mechanism by which it works in Keynes. Of course if it works, it increases Y and that will increase inflation and expected inflation chez Phillips/Friedman/Phelps. The expected inflation is consequence, not cause.

The reason that it works in Keynes, of course, is that it increases demand. Period. Put another way: In a liquidity trap, the problem is that the Wicksellian/Keynesian natural rate is below the expected deflation rate (the latter may be a small negative number, as it is now), which latter is the lower bound on the real rate of interest given that nominal rates can't be negative. Fiscal policy in the trap works not, or not necessarily, by reducing the lower bound, but rather by raising the natural rate above the lower bound. But this can't happen in the DSGE models, because the natural rate - what they call the natural rate, which isn't what Keynes and Wicksell meant by it - can never change. That's a fact about their ridiculous models, not a fact about the world.

2 comments:

Nick Rowe said...

Kevin:

Damn! I can't find the link to the latest Eggertson paper to check!

You sure you are right on this?

The canonical New-Keynesian DSGE model has an IS curve and a Phillips Curve. Define the "natural rate" as where the IS curve intersects the Long Run Phillips Curve.

The IS curve is an Euler equation from infinitely-lived households, with rational expectations. So tax cuts won't work to shift the IS curve, because this model is essentially one in which Ricardian Equivalence holds.

But, change the model so that half the households are hand-to-mouth (they consume their current disposable income), and tax cuts shift the IS right, increasing the natural rate, and increasing output. This is essentially the same reason fiscal policy works in old Keynesian models. By making half the households hand-to-mouth, we raise the mpc out of transitory income from 0 to 0.5.

Add G to the model, in a way that does not interfere with the Euler equation (and the simplest assumption is to assume that G either gives no utility, or else does not affect marginal utility of consumption), and an increase in G shifts the IS right, with a multiplier of 1. Just like in old Keynesian models, if the mpc were 0. And this raises the natural rate.

Add distorting taxes to the model, and fiscal policy can shift the LRPC, and can raise the natural rate. (That's not in old Keynesian models, but is simple to add.)

In addition, you are right, if any of these fiscal policy changes do raise the natural rate, they will increase actual and expected inflation (assuming i=0), and give an extra "induced" increase in demand and output. But this only happens because they initially raise the natural rate.

kevin quinn said...

Nick: I actually may have been too hasty. I don't have the paper in front of me, but what you say makes me think I may have misunderstood. Thanks!