Long-Run Real GDP Forecasts: The Hopeless Task of Trying to Pierce the Veil of Time and Ignorance Weblogging
Re-Reading My Weblog: August 2005

Afternoon Must-Read: Paul Krugman: Phantom Phiscal Crises

Afternoon Must-Read: Paul Krugman: Phantom Phiscal Crises: "Matthew Klein takes on...

...the mysterious, persistent fear that Japan and other countries that borrow in their own currencies could suddenly face a Greek-style fiscal crisis.... It’s not just BowlesSimpsonGreenspan who believe in the threat; Taka Ito... is a good and sensible economist; so is Olivier Blanchard.... The answer I seem to get is fear of a dramatic flip in circumstances--that Japan, say, could engage in a sort of macroeconomic quantum tunneling, suddenly transitioning from deflation to crashing currency and runaway inflation.... It does seem an odd thing to be worrying about right now.. does... obsessing about the state of Social Security in 2030, really make that much difference to the prospect of such an abrupt transition? I don’t see the plausibility--and it seems really strange for that concern to loom so large in the face of everything else going wrong.


As best as I can see, the thinking seems to be in an inchoate and half-formed implicit model. I cannot write it down coherently, because I do not think it is coherent. But in it there seems to be a full-employment future period f in which the real value of the debt will be determined by the fiscal theory of the price level with future primary surpluses at their maximum Sf, so that the future price level Pf, the future debt Df, and future real interest rates rf are related by:

Pf = (Df)(rf)/(Sf)

And there is a present period p in which prices are sticky, the current-period real interest rate rp = 0, and the economy is depressed with a low level of output Yp below potential because of an outsized preference for liquidity and as a consequence a low velocity of money:

Yp = (V(rp + π)(M)

Where the velocity of money depends on the current real interest rate, the inflation rate π, and the shift term φ from abnormally-high liquidity preference:

π = Pf/Pp - 1

Add that the current and the future debt are related by:

Df = (1 + rp)(Dp)

Now suppose that shift term φ goes away, and the central bank responds to the rise in velocity by shrinking the money stock M in order to keep the economy from overheating. This reduction in the money stock raises the current interest rate rp. And a problem develops because the higher the current interest rate, the higher the future debt Df, the higher the future price level Pf, the higher inflation π, and the higher the central bank must raise the current interest rate rp in order to avoid overheating.

In order for this to work quantitatively, I think it has to be the case that expectations of future real interest rates on the debt have to be closely (and irrationally) connected to the current real interest rate: something like:

rf = rp

I would like to see evidence for that. I have, so far, seen none...

Without something like that, it simply does not work. Events happening now or in the near future--i.e., an end to extraordinary liquidity preference--have to somehow shake the intertemporal price system far into the future in such a way as to greatly raise the far-future primary surpluses the government must raise in order to amortize its debt, and so turn a debt that was sustainable yesterday into one that is unsustainable tomorrow.