Core of Gold

Using Wolfgang Munchau as an example, Mark Thoma criticizes certain commentators for ignoring “theoretical” arguments for using a core inflation index (rather than headline inflation) to guide monetary policy. I agree with Mark, but in the comments section, I expressed skepticism about our ability to make the point effectively given the complexity of the theories involved. Nonetheless, I’m going try. I won’t go into the actual theory, but I’ll try to make the substance of the case in ordinary English. First I bring Mr. Munchau back to the whipping post:

Some private sector economists made the preposterous suggestion that housing should also be excluded from the core index. If you go down this route, you end up with a core inflation index that no longer bears any relationship to reality.

Though I haven’t made such an explicit suggestion, I might broadly be considered one of those economists, so I guess this is personal. But let’s see. If an index that excludes food, energy, and housing (and maybe a few other things) “no longer bears any relationship to reality,” how about an index that excludes everything except gold bullion? If the former already has no relationship to reality, I suppose the latter must have an inverse relationship to reality. And yet it was just such an index that formed the critical anchor for monetary policy from long before the Common Era until as recently as 35 years ago. (Of course they didn’t call it an index; they just called it “gold.”)

The gold standard certainly had its problems. Along with many other economists, I believe it was partly responsible for the Great Depression. But it would be hard to accuse gold standard advocates of being soft on inflation. In fact, the great advantage of the gold standard was that it gave people a reason to have confidence in the long-term value of money. Even if it had no “relationship to reality,” this gold-only price index was apparently sufficient to solve the main problem that motivates monetary policy’s concern with inflation.

OK, let’s try to solve some of the problems with the gold standard. The overall problem is that a single commodity can have dramatic supply and demand shifts, which produce dramatic price swings relative to other items, and as a result, if money is tied to that commodity, we can experience dramatic swings in the money supply, producing events such as the Great Depression. So instead of just using a single commodity, let’s add some more things to the index: maybe clothing, education, entertainment, communication services, vehicles, recreation, medical care, and a few other things. The resulting index may still have no “relationship to reality,” but if it’s reasonably diverse, it should succeed in minimizing the chance of another Great Depression. The price of gold can fluctuate quite dramatically relative to other items, but the price of gold+clothing+entertainment+communications+etc. – while it may not remain perfectly in sync with the overall price level, and indeed (like gold alone) it may even drift over time – isn’t likely to have dramatic swings over periods of a few years.

Now the question is, where do you stop? Are there certain things you might want to avoid adding to the index? There may be many things, but I would suggest 3 broad categories:

1. Items which experience such dramatic price swings that, even as part of a diverse index, their inclusion might increase the risk of depressions. A certain dark, thick, slippery liquid comes to mind.

2. Items which are over-sensitive to monetary policy. The most extreme example would be short-term bonds. If the Fed were to target an index that included short-term bond prices as a significant component, it would never be able to change interest rates, because any change would immediately produce a dramatic change in the index. This same problem exists, to a lesser degree, with most other assets, such as stocks and (of particular importance) houses.

3. Items which react perversely to monetary policy. If tightening monetary policy casues some items in the index to rise in price, this would make monetary policy more volatile than necessary. This category would generally include items that are extremely capital-intensive, but the particular one that comes to mind is housing services – that is, rent.

I guess it’s about time to make a preposterous suggestion about what to exclude from the core CPI.