The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Asian Central Banks: Working Hard To Stave Off Dollar Depreciation? | Main | Consumer Sentiment Throws A Funeral -- And Nobody Comes »

September 15, 2005

Where's The Inflation?

Certainly not in the core measures.  Although the headline number reported by the Bureau of Labor Statistics today was pretty ugly -- the Consumer Price Index advanced at a 6.3 percent annual annualized rate in August -- the core measures are still holding steady.  That has been true for the median CPI all year, and it has been true for the CPI excluding food and energy components for the past six months.  Here's the summary, from the Federal Reserve Bank of Cleveland:

Over the last 12 months, the median CPI rose 2.3%, the CPI 3.6%, and the CPI less food and energy 2.1%.                                                                                             

  Percent Change From Previous             Month
Mar. Apr. May June July Aug.
CPI 0.6 0.5 -0.1 0.0 0.5 0.5
CPI less food & energy 0.4 0.0 0.1 0.1 0.1 0.1
MEDIAN CPI 0.2 0.2 0.3 0.2 0.2 0.2
  Percent Change, Last 12             Months
Mar. Apr. May June July Aug.
CPI 3.1 3.5 2.8 2.5 3.2 3.6
CPI less food & energy 2.3 2.2 2.2 2.0 2.1 2.1
MEDIAN CPI* 2.4 2.3 2.4 2.3 2.3 2.3

The large discrepancy between the core measures and the full CPI measure of inflation reflects, of course, the impact of last month's acceleration in energy-related prices. Given the persistent gap between overall CPI inflation and the rate of change in the core measures, it may be helpful to remember why we tend to focus on the latter. For monetary policy purposes, what we are interested in is the drift in average prices over time.  For that reason, a situation in which all prices rise by 6.5% would seem a lot more worrisome than a situation in which a few prices increase a lot -- for identifiable reasons that will presumably not repeat themselves indefinitely.

If the core inflation measures didn't convince you that the second scenario is the one we are confronting now, a look at the distribution of price changes in the August data should do it:


That picture, courtesy of Mike Bryan, makes the point.  When weighted by their expenditure shares in the market basket used to construct the CPI, the majority of price increases were far less than the average increase.  Over 50 percent of the weighted price gains were less than 3 percent (when annualized).  Nearly twenty percent actually fell.

You might say that this is all fine and good, but you happen to actually purchase the average market basket, and the price of that went up by a full 1/2 percent in one month.  In other words, your cost of living rose, and stripping out those prices that increased a lot does not make you feel one bit better.

Hey -- that's a good point, and one that moves us in the direction of discussing what monetary policy ought to be trying to accomplish.  Should we be content with managing the rate of inflation going forward?  If so, the core measures seem exactly the thing to be focused on, as they likely provide a more accurate picture of the inflation trend.  But there is no guarantee that such a let-bygones-be-bygones approach will undo the effects of large one-off increases in some price or subset of prices, even in the medium term.  In other words, there is no guarantee that focusing on core inflation will stabilize the cost-of-living over horizons that people may care about.

Just something to think about.

UPDATE: This piece at Econbrowser is a must-read complement to this post.  General Glut appears to be on the same page,

September 15, 2005 in Data Releases , Inflation | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Where's The Inflation? :

» Who cares about core inflation? from Econbrowser
This is another one of those months when you could report pretty much any number you like to summarize the current inflation rate, and, as William Polley noted, newspapers did. At times like these, the concept of "core inflation" can be very helpful. [Read More]

Tracked on Sep 16, 2005 5:22:54 PM

» If your only tool is a hammer, every problem... from William J. Polley
Via macroblog: When weighted by their expenditure shares in the market basket used to construct the CPI, the majority of price increases were far less than the average increase. Over 50 percent of the weighted price gains were less than... [Read More]

Tracked on Sep 18, 2005 7:41:35 PM


"...the Consumer Price Index advanced at a 6.3 percent annual rate in August..."

This is a typo, right?

Posted by: terrible economist | September 16, 2005 at 12:15 AM

"This is a typo, right?"

Well it is - since the August index level of 196.4 was 3.6% higher than in August 2004 - but it's a kind of 'freudian' typo, since the August monthly rate (0.5%) does come to 6.0% on an annualised basis.

One of the big difficulties we have in comparing data is the fact that different countries compile and release it in different ways.

"there is no guarantee that focusing on core inflation will stabilize the cost-of-living over horizons that people may care about."

Obviously. Food and energy are supposed to be 'volatile' elements. This implies that they rise *and* fall in a rather unpredicatble fashion (ie not open to things like seasonal adjustments). As long as they remain volatile I see no reason to change focus, but, if say oil should be up and stay up, then this would no longer be simply volatile, and you would need to account for this. However, the jury still seems to be out on oil.....

The bigger issue seems to be why the pass-through rate is so low.

Posted by: Edward Hugh | September 16, 2005 at 04:12 AM

I could use some help here. When time allows, how about differentiating one more time between "cost of living" (which the CPI tries to track) & inflation (which it doesn't)?
I excerpted this response to DeLong this a.m. for your consideration. Seems relevant to a lot of Economic argument these days.
"... your analyses tend to accept the routinely gathered data as "reality," and I've never seen you ask, "what sort of data are we missing?" To my mind, that's where the significant advances in economics will be found-- in data sets gathered by people who want more than what routinely drops into their laps from the Treasury or the Fed or the Census Bureau or the other Federally Authorized Truth Collectors.
(Yes, I'm also suggesting that we don't yet have an accepted Unified Theory of Economics. Call me a nutcase, if you must.)
Posted by: Tom Markham | Sep 16, 2005 1:26:34 AM. THANKS.

Posted by: bailey | September 16, 2005 at 07:23 AM

Not exactly a typo -- just a late-night brain freeze. I was referring to the annualized monthly increase, not the 12-month increase. I have mended my error.

Posted by: Dave Altig | September 16, 2005 at 07:27 AM

Well, I've bought a nice freezer to keep food in so I can avoid the food price spikes and I drive a lot less than I used to so avoid the energy spikes. The new local Costco has been great in terms of both.

But hey, it's not like most of us have to eat or get anywhere, right? So those little inflationary spikes don't affect things much... oh, and of course, paying people less, that helps keep those inflation numbers down!

Keep those charts looking good, guys - that's what's important after all, isn't it?

Posted by: donna | September 16, 2005 at 04:05 PM

So we don't have to worry about inflation as long as the economy is growing? Once the next recession hits, those energy prices are going to drop like a rock and we won't have to worry about deflation as long as the core rate stays positive. We just need to average inflation over the business cycle. Buckle your seat belt, we are in for a bumpy ride.

Posted by: Lord | September 16, 2005 at 04:41 PM

bailey -- Here's a quick jump at your question. I like to think of the cost of living as the amount of resources that have to be expended to maintain a fixed basket of consumption. We can imagine defining this in world where money is completely nonexistent -- how many hours you have to work, for example to keep your consumption constant. The main thing is that if we define the cost of living this way, it will change whenever relative prices change. When I talk about inflation, I am usually thinking about changes in the purchasing power of fiat money, a concept that does not fundamentally involve changes in the relative prices of real goods and services.

That said, inflation and cost-of-living are often used as synonyms, and I really was doing the same in this post. My point here was that, from the monetary policy perspective, one might be favorably disposed to a world in which the Fed literally targets a price *level*, which means offsetting the effects of any temporary price spikes sometime in the future. Note, however, that this does not do do one bit of good in helping a recovery in the cost-of-living concept I defined above. Which is to say:

donna-- I feel your pain, but I'm not sure there is a whole lot monetary policy can do about it (beyond not becoming part of the problem).

Posted by: Dave Altig | September 17, 2005 at 05:56 PM

As regards the evil tendency of looking at core price measures so as to be able to declare victory against inflation and go home - what alternative do the gripers suggest? I think the good Mr Altig has it just right. The Fed could do serious harm by responding to headline price measures, without thinking about what is going on among the components. More than 50% of CPI component rose less than the mean rise among all components in September, which is to say that the distribution of price increases was skewed. If inflation is a generalized rise in prices, then we certainly need to notice when headline price measures are rising fast, but prices on most goods and services are not.

If food prices rise because of weather, energy prices rise because of weather, while other prices are far tamer, and the Fed cannot specifically target food prices, energy prices or weather, do you really want the Fed responding to headline price measures, while ignoring all else?

Posted by: kharris | October 17, 2005 at 03:39 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad