Stephen Williamson Commits Himself to Cargo-Cult Macroeconomics

by: Brad DeLong

The cargo cults of the South Pacific noted a strong correlation during World War II between (a) the existence of airstrips, and (b) cargo airplanes landing and disgorging huge amounts of consumer goods. They thought it was worth trying to see whether this relationship was causal: would building airstrips cause cargo planes to appear and land? They experimented... and found that causation ran the other way: it was the expectation on the part of the Seabees of the United Nations forces fighting the Japanese Empire that cargo planes would want to land on some island that prompted them to construct the airstrips.

Stephen Williamson reveals that New Monetarist Economics is isomorphic to cargo-cult science. From New Monetarist Economics: How to Get Worked Up Over Nothing:

Suppose a cash-in-advance model with a representative consumer, period utility u(c), discount factor b, constant aggregate endowment y. c is consumption. The consumer needs cash to buy c each period. Suppose y is a fixed quantity of output received by a firm, which is sold for cash within the period, and then the cash is paid as a dividend to the consumer at the end of the period. Have the money stock grow at a constant rate m. The real interest rate is constant at 1/b - 1. The nominal interest rate is (1+m)/b - 1, and the inflation rate is m. Constant m implies a constant nominal interest rate and a constant inflation rate. If m < 0, there is deflation, and the nominal interest rate is sufficiently low to support the deflation. I can think of the instrument the central bank sets as either the money growth rate or the nominal interest rate - that part is irrelevant.... What's the problem?

As Jed Harris pointed out, if Stephen Williamson had been in Paul Volcker's place in 1979-1983 he would have lowered rather than raised interest rates in order to reduce inflation. Williamson would then have been much surprised when inflation further accelerated. He would have been significantly more surprised than the cargo-cult airstrip diggers were when no planes arrived. They, after all, were simply testing a hypothesis about the direction of causation in an empirical regularity. They were not claiming a gold-plated theoretical warrant that the direction ran in his particular way.

As Milton Friedman liked to say, high nominal interest rates are a sign that money growth has been high in the past and is expected to be high in the future. But you can't lower expected future money growth by lowering interest rates. Let me look... ah, Scott Sumner has, I see, beaten me to it.

From Friedman:

Low interest rates are generally a sign that money has been tight, as in Japan; high interest rates, that money has been easy. [...] After the U.S. experience during the Great Depression, and after inflation and rising interest rates in the 1970s and disinflation and falling interest rates in the 1980s, I thought the fallacy of identifying tight money with high interest rates and easy money with low interest rates was dead. Apparently, old fallacies never die...

But where Williamson really makes his play for the Stupidest Man AliveTM Runner-Up Prize is that he is not only wrong, but uses his wrongness to mock the people who are right:

Have at look at this: 1. Krugman. 2. Rowe 3.Harless 4. Thoma Is there something in the water, or did these guys sit through some funny macro classes? What they are objecting to in Kocherlakota's speech is one of the most innocuous things he said...

The problem, of course, is that Williamson does not have a model, but rather a set of conditions required for a steady-state equilibrium path. But in order to figure out whether an economy is or can be expected to be evolving along a steady-state equilibrium path, you need a model. And he ain't got one.

The standard, conventional model can be drawn on a graph with the nominal interest rate i on the vertical axis and the inflation rate π on the horizontal axis. This model has a "Wicksellian balance" condition--Williamson's r* = i - π --that does not hold at all times but, rather, is a steady-state equilibrium condition: if the economy is to the left of and above the Wicksellian balance line so that the real interest rate will be high, inflation is falling; if the economy is to the right of and below the Wicksellian balance line so that the real interest rate is low, inflation will be rising. And the Federal Reserve will not make its target interest rate all the time either: if the interest rate i above the red policy rule line, the nominal interest rate will be falling as the Federal Reserve pumps out the money; if the economy is below the interest rate will be falling as the Fed sells bonds for cash.

Skitch.com > braddelong > System-7-1-3Click to enlarge

In this model the point where the blue and the red curves cross--the point where Williamson says inflation and nominal interest rates are constant--ain't a point where inflation and nominal interest rates are constant: we do not expect the economy to be there. What we expect to happen is that if the economy starts to the left of the green line it heads off toward deflationary Valhalla at point C, and if it starts to the right of the green line it heads off to inflationary Valhalla.

Of course, nobody believes that when inflation reaches 120% per year the Fed will still maintain its policy of pegging the nominal interest rate at 2% and so pumping out the money at 120% per year. Instead, we all believe that at some point--when inflation is too high--the Fed will abandon its nominal interest peg and cut back on money growth, and so its policy rule line will have a kink in it, like so:

Skitch.com > braddelong > System-7-1-3Click to enlarge

Then we expect that if the economy starts to the right of the green line it will ultimately head for point A--and that if it starts to the left it will still head for deflationary Valhalla at point C. In this model, lowering the leftmost constant-nominal-interest rate peg arm of the policy rule is not a way to reduce inflation. Instead, it is a way of moving the green line to the left and increasing the set of initial conditions for which the economy converges to point A.

I did talk about this before...