Fiddling Numbers

Paul Krugman’s various economic theories and points may all be valid and correct, I don’t really know, but the way in which he goes about proving them varies from weird to bizarre. The latest is his desire to justify omitting food from the calculation of the “core inflation index” calculation because they are “too volatile,” and to debunk the “right wing claim” that food prices are soaring wildy and would be causing major inflation if they were included.

So he includes this chart, where blue is food price inflation,

I guess he could make a small case for the “volatility” bit, with that little dip in year (or month) two, but it’s a pretty big stretch to do that when nine of the ten periods show a remarkably consistent pattern of food cost being about one half percent higher than core inflation, don’t they?

When the pattern is as clear as this one is, a mind which is thinking clearly tends to toss out the one anomaly and see the pattern that remains. Doing that suggests that, while food prices are hardly "soaring," their omission is indeed keeping the core inflation rate a bit lower. But, of course, the mind has to be thinking clearly. Even with the one problematic period, a procedure which would be working properly nine times out of ten would be pretty good. The way we are doing it now seems to be accomplishing its intended purpose only one period out of ten.

Unless its intended purpose is to publish a lower inflation number.

If I were in Paul’s position and trying to illustrate the point that he’s trying to illustrate, I would pull up that chart, take one look at the thing and say to myself, “I need a different chart.”

In any case, all of that beggars the question. What is the purpose of the “core inflation index” that we are calculating, anyway? Given that it is used as sort of a “misery index,” it should be used to measure the rate of change of the cost of living, the rate of change of what it costs to sustain the standard of living that one has in this nation. Is eating food not part of that standard of living? Is buying food not part of the cost of maintaining that standard? How can you omit a significant part of the cost of living and still claim that the inflation index has any meaning at all?

Energy in all its forms is also omitted, also “too volatile,” and between them these two things make up a significant part of the cost of sustaining, not only a standard of living, but life itself. That is particularly true in lower income brackets, and the lower one’s income is the closer it becomes the case that food and energy will constitute the entirety of one’s spending. With those two items omitted from calculation, the core inflation index becomes an utterly meaningless number.

Update: Inflation compounds, so small percentages can have large effects. Inflation of 2% monthly turns into an 26.8% annual rate, while 2.5% monthly becomes a 34.5% annual rate. Inflation is an annual rate, but it goes on year after year, nonetheless, and difference is not as trivial as Krugman paints it.