Bryan Caplan  

Inflation: Why We Should Probably Count Food and Fuel After All

View from the Rear-View Mirror... Learn to Love the Liberal Welf...

When I was a strongly Austrian-influence undergraduate, I scoffed at people who downplayed inflation by saying, "Well, if you ignore food and fuel..." It seemed like a sleazy effort to cover up a government-created problem by refusing to count a big part of the problem.

Once I got to grad school, though, I started to rethink my position. Food and fuel really are much more volatile than other goods for reasons (like drought and hurricanes) that have nothing to do with monetary policy. If you want to isolate the portion of inflation to pin on the Fed, dropping food and fuel seemed like a reasonable procedure.

Lately, though, I've started to think that I was right the first time around. True, food and fuel are unusually subject to supply shocks. But as commodities, sticky-price models tell us that food and fuel are the very sectors where the effects of loose monetary policy will become visible first. Think about it this way: Suppose there are ten goods. Nine have big menu costs; one has no menu cost. If you helicopter drop money on this economy, the short-run effect will be a sharp spike in the price of the good without a menu cost. Dropping the zero-menu-cost food from your price index masks an inflationary effect that should have been obvious.

Are there any data series out there that try to deal with my concern? How successful are they?

Comments and Sharing

COMMENTS (14 to date)
John V writes:


You know what would make a great post that I would love to read...

An answer to This by Krugman and this by Waldman

What gets me about the Krugman post is that it flies in the face everything (not much) that I actually understand about Monetary dangers and historical examples explained from a Austrian/Friedman POV. I mean, is Krugman serious?? Come on. I count on economists to sense of all the BS with some level of dispassion and I feel like I'm not seeing it...unless a mixed libertarian-based conceptualization of all this is dead wrong and off the mark.

On the Waldman piece, the remarks over the performance of pro-investment and capital policies seem a bit silly without considering or controlling for, somehow, the role of Monetary policy and tripled energy prices...never mind food.

Is armchair knowledge totally and utterly off the mark or are economists like Krugman and Thoma getting so wrapped up in "advanced post-graduate" economic theory and exceptions that they've lost sight of where they started from?

I'd really like to know...

John writes:

There's also research, I believe a working paper at one of the Fed's (just came out), that core inflation measures aren't actually less volatile than the headline figures. And that core inflation isn't a better predictor of future inflation than headline inflation. Can't remember the link, but I think I got it from a post within the past two weeks from BigPicture.

Eli writes:

Menu costs shouldn't really affect the measured rate of inflation in anything but the very short run. Caplin and Spulber (1987) show that as long as price revisions are imperfectly synchronized, (s,S) pricing and a monotonically increasing money supply gets you monetary neutrality.

Even if price revisions are synchronized, menu costs are a short-run phenomenon. Surely long-run inflation trends are sufficient to demonstrate the extent to which the Fed is responsible for inflation.

Bill Conerly writes:

The relationship between money supply and prices has some slop (sorry for the technical jargon). In quantity theory terms, velocity is not perfectly stable. (In Austrian terms, no numerical relationship has meaning.) So temporary changes in total inflation do not necessarily result from monetary policy. But persistent changes in total inflation do require expansive monetary policy. So focus on core inflation until you have a prolonged (24 months or more) disparity between total and core, and then focus on total.

Grant writes:

When designing any sort of control unit for a complex system, engineers don't toss out a signal because it contains a lot of noise, they filter it and use what they can. I can't see how totally ignoring fuel and food makes any sense? Were I designing some sort of "black box" to estimate future inflation, I would absolutely make some use of fuel and food costs provided I had the spare clock cycles to do so (which the Fed most assuredly does).

oh minseok writes:


I'm a korean living in seoul.(name: Oh Minseok)
Korean special investigative team investigated samsung
But they did not work right.
Suspicious to have been bought off.
Samsung corporation had many crimes.
And the team investigated samsung corporation.
It contained korean companies samsung,huyndai,sk CEOs' illegal issuing
stocks or bonds. ( previous CEOs or present CEOs )
The quantity were plenty.
(Three company CEOs did(and are doing) many crimes to me.
Many koreans are knowing it.
But many koreans are bought off by illegal issuing stocks or bonds.
Korean prosecutors are also guilty.)
The team knew it.
Korean special investigative team must investigated this.
But they concealed it.
I ask for asking for this criminal investigation to prosecutors in any
And help the shareholders and me.
Three companies are hacking me and trying to kill me.
And are suspicious to use my name and email illegally.
If you receive another message that I dictated above are not true,
it is not from me, but from three companies.
The things I dictated above are true.

B.H. writes:

The Cleveland Fed produces a monthly median CPI inflation rate: the idea is that insteading of removing so-called volatile sectors from the CPI, it simply produces a distribution of price changes for all of the parts of the CPI. The median inflation rate is the price change of the middle of the distribution.

The Dallas Fed does the same thing as Cleveland but uses the personal consumer expenditures price index instead of the CPI.

Both are publicly available free of charge, available at the respective websites, and are available from data vendors.

It turns out that the "core" CPI is highly correlated with the median CPI inflation.

fundamentalist writes:

Not sure what kind of data sets that would be. But wouldn't elasticity of demand play some role in price volatility? I understand the menu concept, but it seems to me that the prices of items such as gasoline and food in which people are very price sensitive will adjust more quickly to changes in supply than other items, like sterios. The real problem is determining if the price change resulted from supply changes, demand changes, or monetary changes.

Paulidan writes:

When Keynes said "when the facts change, I change my mind" his followers covered their ears and pretended not to hear.

Nick Gregory writes:

Isn't energy and food the base of an economy right along side labor? They are fundamentals that touch everything and are necessary for the function of society.

So wouldn't it simply be exposing the loss of faith and value in the dollar? Or is there also supply and demand issue such as government paying for biofuels? Or is there a combination of factors, a perfect storm.

We are also at a turning point in globalization and all roads point to China. The investors seem to believe the largest consumer nation of tomorrow will be China probably based on their population numbers. And China does seem to have an inflation problem.

Even worse, can the Fed expand credit any further? A credit expansion while American wages are under price contraction forces seems unwise as you are basically asking for more defaults.

We seem to be at a point where most average American's have to take a pay cut to balance out with supply and demand of wage costs over the globe. Are we seeing that pay cut in action? So how can you have inflation when you have massive forces pushing down on American wages and thus justifying the reasons 80% of Americans feel they are in a recession. Now Americans are forced to take pay cut to balance with the cost of world labor while big business is global, Americans are not. And as long as Americans are considered to be over priced labor so inflationary forces generated in the US would flow outside the United States as investments in foreign labor markets. The major inflationary force of housing is sinking to the level where American wages can support it without credit expansion and it is getting harder to find another inflationary place to expand in beyond outside the United States.

I really don't think you will have major inflation in the US for awhile, you will have inflation outside the US until things balance. A credit expansion may create a few jobs here but most likely will bypass Americans and go outside the country until Americans become cheap enough to employ for global trade. So maybe the question is, what is the global inflation rate but how do you measure it with fiat currencies that stand on top of a sinking resever currency, the dollar?

If your house is based on the dollar and it is sinking, everyone sinks. So you can't really compare currencies as most of them pretty much stand on top of the dollar, especially as long as fundaments such as oil is traded in dollars.

It's a big mess but that is the price of a system based on political money instead of real value. It's also the price of believing you can receive something for notion as Friedman believed. Now we have to pay the price in defaults and relative global wage balance.

Nick Gregory writes:

Let me simplify it

There is the real possiblity what we are seeing is an American pay cut. What is going on is a balance of value in labor. Americans will see it as inflation, but really it is simply their value falling to a global wage balance point. You can't have free trade without wages attempting to balance at some point.

If you want you can call it the "Global Wage Rebalancing Act caused by Free Trade" In my opinion this is the core of what is going on that explains the inflation and the reasons why Americans feel they are in a recession. Actually, Americans are in a recession as they are losing the value they once wielded in the world and now are being balanced with the rest of the world.

Russ Wood writes:

If only there were a better, more accurate, simpler signal of inflation.
Oh well, I wonder what the price of my gold bar is today...

Michael Kolczynski writes:


Maybe I am reading this wrong, but your post seems to ignore the very nature of inflation.

We know inflation is not prices rising. Supply and demand is not inflation, regardless of how many items see an increase in price due to an increase in factor costs.

If the only change in the system is moderate inflation, then supply and demand curves should move in harmony -- so essentially there is no change except the number of currency.

If there is no inflation but prices are rising, then supply curves move upward (generally), supply moving upward (oil demand increasing, causing anything that uses oil as a factor of production to have a supply curve that moves upward). Of course you could see demand curves fall due to budget constraints that would maybe bring the price back down to its former level, but then you'd still see few items purchased, which is a real change, unlike moderate inflation where there is no change other than price scale.

Those are two major distictions. We don't measure inflation by prices because it's optimal, we do it because it's the (supposedly) best "estimate" we have. But if we include things we know are based on the latter (supply curves moving upward) then we completely lose our ability to guage the former (actual inflation).

Michael Kolczynski writes:

Yea, I read your post wrong, ignore everything I just wrote.

Comments for this entry have been closed
Return to top