Tuesday, December 28, 2010

Is "Volatility" the Last Refuge of a Scoundrel? (Or At Least a Keynesian?)

To "prove" his argument that commodity prices and inflation have no relationship, Krugman has a blog post in which he uses the graph below.

The problem is this: Which commodities? We have seen prices of gold skyrocket, although Krugman would regard gold as passe or even evil. But an even more important bellwether would be oil prices, since oil worldwide is denominated in U.S. Dollars. Thus, oil might well be more sensitive to changes in the value of the dollar than might other commodities.

In other words, de-homogenization might be something worthwhile here. Just a thought.

The second problem I see is that once again, Krugman uses the deus ex machina explanation of "volatility" to explain away anything that might be uncomfortable. Are oil and gold prices going up? Well, so what? They are volatile; that explains everything.


Anonymous said...

Krugman puts up charts with intentionally goofy axes and time periods to obfuscate the point. He is doubly guilty of what he accuses the republicans of. And yes, that is all he is doing - playing politics.

Steve said...

As I understand it, since one commodity (or even group of commodities) would account for only part of the CPI calculation, the CPI will by definition smooth out the commodity curve once thousands of other products are factored in.

But interesting to note, if you shrunk the Y-axis for the blue line by a factor of 20, those two charts would trend in the same direction quite nicely.

Not sure how this fits into the whole of economic theory but I can spot a meaningless graph from a mile away.

Bob Roddis said...

Wouldn't the price of scarce metals stated in the form of scarce gold and silver money tend to not rise as opposed to prices for such metals stated in constantly diluted funny money?

Tel said...

To my eye, the blue line in the graph looks a lot like fuel and fuel prices are indeed volatile. They are sensitive to economic activity, so when logistics services are not delivering, when families are staying home, and when the wheels of industry are not turning, fuel prices fall (as they did in 2008).

However, in the bigger picture fuel prices go up more often than they go down, and if you want to compare your experience at the pump with 10 years ago, or 20 years ago, it's pretty easy to see beyond the volatility.

The cracker for Krugman is that he is perfectly happy for FOOD to be excluded from "core inflation" whilst food prices are kept non-volatile by a whole host of mechanisms, and food prices are steadily rising too (a bit faster than "core inflation" from what I can tell). If Krugman can explain why CPI excludes food then I'll start to believe he is on the level.

Anti Vigilante said...

I was about to take my Krugman vudu doll out, but then Tel stated the obvious and it was no longer necessary.