In a recent blog post titled, "More on Disinflation" Paul Krugman comments on the headlines trying to interpret the recent CPI announcement:

As Mark Thoma recently suggested...it’s odd that most macroeconomic forecasts show inflation remaining stable or even rising despite continuing very high unemployment. Both history and logic suggests that this is wrong.

And declining inflation is clear using any core measure. The Fed, I happen to know, tends to focus on the core personal consumption expenditure deflator:



There are some strange bobbles in this measure, suggesting that it’s not as good a measure of inertial inflation as advertised, but leaving that aside we see, once again, serious disinflation as a result of the recession.

I still think there’s a real risk we’ll turn Japanese.

If you want to see me tackle the claim that both theory and history support Krugman's belief that high unemployment ==> low price inflation, see my recent Washington Times op ed.

Also of relevance is the fact that Krugman focused on an odd time frame. Here's the full picture:



Now is it really jumping out at you that the Price Index of Personal Consumption Expenditures--stripped of food and energy--go down during recessions?

And even in the periods where it does seem to work, is the mechanism really the Keynesian one about recessions being caused by a fall in aggregate demand? Couldn't you just as easily explain it by saying that people have to buy food and energy, and so when their prices spike--especially if it's during tough times--then the prices of everything else tend to fall?

In other words, my point here is that people say, "Oh we're stripping out volatile food and energy to look at the underlying trend." But since people care a lot more about food and energy than a lot of the remaining purchases, that's closer to throwing out the trend and looking at the residual.