The Big Picture

« Inside the Investor's Brain | Main | ECB on Core Inflation »

Inflation: CPI, Core Rate, Inflation ex-Inflation

Thursday, October 04, 2007 | 06:46 AM

I have been a critic of the official inflation rates for quite some time. Recently, some have taken me to task for this. Brad DeLong chastises me, saying "Barry Ritholtz Does Not Seem to Understand the Purpose of "Core Inflation." Economics and... phrases the issue thusly: What is the core for?

These criticisms miss my main point. But that's probably my own fault, for not explaining the issue better. After you've posted on the same issue a few 100 times, you assume everyone has followed your line of thinking. Today's commentary will attempt to clarify that.

Let's start by framing the issue: Regarding inflation, these are the questions I want to explore:

1) What is the actual, real world, rate of inflation?

2) Why does the BLS model -- the "official" inflation rate -- vary so greatly from the real world? How significant is the spread between the two?

3) Why does the Fed Focus on the Core rate, and not the actual rate? What are the Fed policy repercussions of this?

4) What does this mean to Consumers? Investors? Savers?

(Our prior explorations of those questions are here).

Long time readers know I do not believe that the BLS CPI data actually reflects the true price increases of the real world. Further, the Fed's focus on the core rate has significant repercussions for policy. Additionally, the debasement of the U.S. dollar means that U.S. have seen their purchasing power eroded just as surely as if inflation were even higher.

Let's look at few new charts to see if they can shed any light on these issues. (For you "play along at home" types, here is the inflation data in Excel, via Bloomberg: CPI_vs. Core inflation.xls).

Any good economist will tell you that inflation is primarily a monetary phenomenon, so let's go to the relationship between Money Supply and Inflation, via Now & Futures:

chart courtesy of Now and Futures

>
Note the massive increase in M3 money supply. (Why the Fed decided to cancel the reportage of this key data point, we will leave to another discussion).

What has this increase in money supply meant for CPI? Let's look at the headline versus the core inflation data for the past 5 years:

chart courtesy of Michael J. Panzner

>
(Note: the plummeting inflation rate in the summer of 2006 was caused by a change in the widely followed Goldman Sachs Commodities Index -- now owned by S&P -- which temporarily sent Oil under $60).

Let's look at the spread between the two rates:

chart courtesy of Michael J. Panzner

>
Since late 2002, the core rate dramatically misrepresented actual inflation. This led to the Fed taking rates down too low, and leaving them there too long. The repercussions of this have been very significant, and are still being felt today -- from the credit crunch, to $80 dollar crude oil, to huge increases in food prices, to unaffordable housing.

The bottom line is that ultra low rates, and big increases in money supply, sent the U.S. dollar to 15 year lows, raised the costs of all goods denominated in dollars, and lowered the standard of living for many people -- perhaps most -- living in the U.S.

This is why its is preferable for the BLS and the Fed to offer less spin on inflation, and to report actual, real world data -- not an artificial construct that consistently under-reports inflation.

I'll have more on this later . . .

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c52a953ef00e54f0464948834

Listed below are links to weblogs that reference Inflation: CPI, Core Rate, Inflation ex-Inflation :

Comments

the core-CPI statistic is the biggest fraud on the american public since WMD.

it took awhile for people to wise up to the WMD sham, but eventually they did.

this will be no different...

Posted by: m3 | Oct 4, 2007 8:12:09 AM

"...the debasement of the U.S. dollar means that U.S. have seen their purchasing power eroded just as surely as if inflation were even higher."

This is a pure non-sequitor. Unless you mean "measured inflation" rather than "inflation" you have said that inflation is higher than inflation. If you really mean "measured inflation" then you are using you premise as an argument for your premise. That is not a good way to make your case.

A further weakness to the statement is that "debasement" is not what one would call a neutral, objective term. Much of your 100 or so efforts to make this case have equally relied as much on rhetorical flourishes like that as on fact.

You claim to be after the "real" rate of inflation, but fall back on M3, which you assume is an iron-clad determinant of inflation. That is generally known as "begging the question" I believe.

The problem is not that DeLong has missed your core argument. The problem is that you think you have made a strong argument when you have not. Your religious zeal has eroded your objectivity. Go back. Answer the criticism offered by DeLong and others objectively, without resort to rhetorical special pleading or circular arguments and maybe you'll will learn and become more convincing in the same effort.

~~~

BR: I plead guilty to the usage of the word debasement.

My assumption is that a 30% decline in the US dollar versus other currencies would qualify. How much does the US dollar have to fall before the word debasement is acceptable -- if ever?

Or asked via Stephen Colbert, "How can I be objective when the facts themselves are biased?"

Posted by: kharris | Oct 4, 2007 8:13:12 AM

Perhaps someone more knowledgable than I can answer this: If the core rate is a dramatic misrepresentation of actual inflation, what can the informed investor do to leverage that "disconnect" to their own advantage?

If there is no real value to be gained in knowing "the truth" then the knowledge just becomes an intelligent conversation that parallels discussing politics -- in the end, politicians are all liars but what are we going to do about it?

Now I understand why Plato, fortunately for us all, gave up politics for philosophy...

Posted by: The Financial Philosopher | Oct 4, 2007 8:22:53 AM

The point that the incorrect inflation number causes the Fed to apply incorrect policy is telling. Using the actual rate (ie: wages vs cpi have declined since 2000) would have identified the massive house price bubble earlier. Clearly, that large deviation from long-term normal ratio of housing price to income would have indicated a problem.

Posted by: wally | Oct 4, 2007 8:38:44 AM

Barry, I'm curious as to why you would use charts created by someone who I believe is predicting financial armageddon rather than using the government's own charts of CPI and Core CPI? Essentially, they should both be the same, so I'm curious as to why you chose Panzer's rather than the official versions.

Secondly, the chart you posted shows that there is currently no spread between the Core and All-items CPI. Bullish, no?

The BLS uses the Core statistic for monetary policy as it is not as volatile as the All-items data. If the All-items data was instead used, we might expect to see huge swings in monetary policy, as the gov't would be reacting a very volatile index. I doubt that the gov't would be any better than most market timers, and this approach would likely be disasterous. Instead, the Core is used for policy making as it is obviously less volatile and allows for carefully measured changes over time.

I haven't read all your posts on this topic, but what I have read does not suggest a better solution for guiding monetary policy.

Finally, why has not anyone addressed the likelihood of falling inflation due to falling home prices?

~~~

BR: I asked Mike to create the last two charts for me, as I was traveling w/o access to all of my work in the office, he has access to Bloomberg, and he is a wizard with Excel.

Posted by: Woodshedder | Oct 4, 2007 8:47:22 AM

Barry - excellent charts and they help clarify your argument but kharris makes some good points, albeit rhetorically. Being curious took my own pass at digging into those questions which is at: http://tinyurl.com/327bdt .
If you look back CPI is more volatile but tends to converge on CPIx, as historically do both the PPI and PPIx. Where the argument hangs together is if a) we're seeing a l.t. structural shift in food & energy because of the BRIC impact on world energy supply/demand balance. Much more worrisome, to me, is a recent, unprecedented and rising gap between cumulative PPI and CPI. The 2nd related question is whether or not, after 25 years of secular deflation, we're enterring a new world of rising expectations.
Which when you stop to think about is pretty much what the Fed has been trying to tell us. The second hidden thought is that the economy has adjusted incredibly well and resiliently to the upward pressures. Did we get lucky, or is it a downturn in the economy or are we just finishing an adjustment process to higher energy prices and moving to a new plateau ? Or some combination ?

Posted by: dblwyo | Oct 4, 2007 8:49:53 AM

How do you counter the position that using M3 is useless as it much more reflects demand for money. On past Kudlow shows you have raised M3 to have it quickly countered that it's a false measure of money supply, rather is a better reflection of demand. In fact, all your measures, the typical guests on his show would claim are useless without considering demand side of the equation. It has been pointed out that Russia has enormous money supply growth, but only because demand is even higher, therefore do they have inflation? Can you discuss the expansion of money supply without including figures to discussing concurrent money demand?

Posted by: Stuart | Oct 4, 2007 8:52:57 AM

Russia has no inflation? Da, because the ruble has appreciated from 34 to 25 to the dollar. Bought a flat in Moscow for $114,000 in 02. Sold it in 04 for $185,000. I understand it is now worth $300,000. Russia used to be the biggest user of U.S. cash in the world besides the U.S... Now it is Euros. Even Russians understand inflation better than our politicos. We need more glasnost here.

Posted by: Ross | Oct 4, 2007 9:12:28 AM

I wish people would stop using "Begging the question", for an oldtimer like me who remembers the original meaning, I would prefer "Asks the question", if that is what they mean.

Posted by: Big Al | Oct 4, 2007 9:20:11 AM

Unless you mean "measured inflation" rather than "inflation" you have said that inflation is higher than inflation. If you really mean "measured inflation" then you are using you premise as an argument for your premise. That is not a good way to make your case.

Um, I don't see him doing that.

BR
(1)compare _prices_ (various CPI) with money supply
(2)makes the case that some price measurements track money supply, some don't.
(3)The one the Fed uses (core) is the one that does not.
(4)using the core as a proxy measurement of monetary inflation is bad
(5)using the core as a measurement of cost of living is bad.

Actual money supply, CPI-U, track well. If one were to look for a means to track monetary inflation, why would one use core, rather than CPI-U?

BR is making the case that use of core would be to understate monetary inflation.

You claim to be after the "real" rate of inflation, but fall back on M3, which you assume is an iron-clad determinant of inflation. That is generally known as "begging the question" I believe.

Use price of dollars in dollars and have a perpetual inflation of zero. By changing the things we measure, we can have a result of measured inflation of zero, deflation, or inflation. Who knows the real rate? But if you were to use a price index to measure the effects, Why would you use one that eliminates most of the things where the effects would show up?

Posted by: cranky | Oct 4, 2007 9:23:27 AM

Using the definition of the St. Louis Fed that inflation is a rise in general prices caused by excessive monetary expansion, the rate of monetary expansion becomes critical.

Gary North has these comments: "....in April, 2007, I compared the four major monetary aggregates as predictors of price inflation. Only M1 was remotely accurate over the last four decades.

Mish has come up with a new aggregate, which he calls M-prime. He symbolizes it as M'. Using Shostak’s article as a guide, he argues that M' is superior theoretically because it does not include any credit transactions, i.e., "sell this asset and get money." I agree with his assessment."

The full article: http://seekingalpha.com/article/47419-is-the-us-printing-money-like-mad has charts from the St. Louis Fed of monetary base which shows the monetary base has been falling and in a downtrend since the end of 2004. If excessive money is inflationary, these charts appear to show that money creation is not the reason for rising prices. Then what is the reason?

Plainly, the only substitute for money supply is debt expansion. Tame consumer price index with a debt bubble - that sounds vaguely familiar....

Posted by: Winston Munn | Oct 4, 2007 9:26:30 AM

1. Inflation really is a monetary phenomenon. But this is only is a half of the full thruth. Monetary supply is driven by monetary demand, not vice versa. (othervise the Fed would use this control mechanism all the time to make infaltion curve. ). So, something behind monetary demand actually drives inflation as a process. The Fed are just reactive not proactive.

2. The difference between core CPI and CPI (indices themselves not rates of change, i.e. inflation rates)is a piecewise linear function time.

This is obvious from figures 5,6, and 7 in my old post
http://inflationusa.blogspot.com/2007/07/what-is-difference-between-cpi-and-core.html

Therefore, quantitatively, there is no difference between core CPI and CPI, the former is teh sum of the latter and linear time fucntion A(t-t0), where t0 is 2003,
and A=-1.5

3. Inflation is an almost precisely measured macroeconomic variable in sense of quantitative modeling. It is not the same than to measure TRUE inflation, but the BLS measures consistently the SAME portion of this true inflation. For numerical modeling, one needs just to calibrate using linear coefficient, as we know.

On the other hand, one can have an unlimited discussion on social, demographic, economic effects of inflation and what is the distance between measured and true inflation.

4. The precision of the inflation measurements is proved by its stict dependence on the change rate of labor force level (i.e. employment + unemployment).

http://inflationusa.blogspot.com/2007/08/can-we-predict-inflation-and.html

(Links in this post go to working papers with related analysis and data sources)

Thus, inflation is a process driven by only one force - the change in labor force level.

Posted by: KIO | Oct 4, 2007 9:32:31 AM

All rather silly. Inflation isn't a real measurable thing of course. There is not "basket" of measures that net out to any actual reality.

Decide what you want to happen and then find a policy measure to fit it. It's just a lie.

Housing prices were all over and inflation didn't move. You telling me that's accurate?

What is needed is a "price matrix" that measures prices for a few hundred common items in 10 or 20 most populated places. This would then be released...not any one aggregate that makes it look like an oversimplified idiotic function.

When oh when will economists quit acting like they know something from oversimplified and idiotic "functions?" What a joke.

~~~

BR: We actually made repeated comments about how OER understated Housing Inflation as prices were rising.

See these posts for more details

Posted by: Ryan Lanham | Oct 4, 2007 9:35:00 AM

Therefore, quantitatively, there is no difference between core CPI and CPI, the former is teh sum of the latter and linear time fucntion A(t-t0), where t0 is 2003,
and A=-1.5

You think so? A constant unidirectional error over time (how many years now?) has accumulative effects.

And what's to say that in the future, when the direction becomes inconvenient, the measurement wont be reconstituted ala Boksin?

Posted by: cranky | Oct 4, 2007 9:54:35 AM

The idea of excluding food and energy because they are more volatile has always troubled me. Suppose you're trying to evaluate a company with 10 divisions, and 2 of those divisions have extremely volatile quarterly results. Would you simply exclude those divisions from the analysis? Few if any analysts would take such an approach; most would simply apply some form of smoothing technique to the divisions in question.

Posted by: david foster | Oct 4, 2007 9:58:05 AM

kharris doesn't make good points; and the heavy rhetoric redounds against, rather than in his or her favor.

The essence of the argument, as I have read it over the past year or so, is that the market utilizes measures provided by the government as givens for established mathematical formulae which accurately illustrate certain financial realities. The free market reflects the aggregated responses of actors to these realities. The government has, by virtue of its assumption of the power "To coin money, regulate the value thereof, and of foreign coin, and fix the standard of weights and measures;" (U.S. Const.,Art 1, sec 8.) has created a situation where an efficient market demands that "the value thereof" be accessible, and the government has 'fully occupied the field'.

When the information that the govenrment provides is false or incomplete, the market acts without correlating to reality; much like a person jumping of a 100 foot cliff while thinking it is a 1 foot drop.

The offense is the mis-defining of inflation. dblwyo attempts to mitigate this by stating that CPI and I are separate terms, separately defined. This is correct as far as it goes, but the government employs CPI as a given for policy decisions for which the formula requires I, such as setting entitlement rates, and taking monetary and political policy decisions.

Where the rubber hits the road, in these cases, is the political and financial expectations of the market actors. Social Security was accepted by the voting public because it was a program designed to alleviate the scarcity experienced by pensioners. When the government undermines the intent of that negotiation with the public by not alleviating said scarcity to the extent promised or implied because they only adjust the 'return' of the deal to the 'other side' of the negotiation according to CPI in the face of more rapidly declining purchasing power per dollar, then the scarcity experienced by the public and pensioners is not alleviated.

In short, it is a lie that frustrates the intent of and injures one of the parties, i.e., fraud.

And debasement is a precise term meaning "to reduce the exchange value of (a monetary unit)" (Merriam-Webster). When 1 dollar will not pay the same fraction of my rent, or purchase the same amount of a commodity because the government has "regulated its value" lower by excessive money creation, then it has been, indeed, debased. And, since the government is the sole entity with the power to do so, it must also be the exclusive party in blame.

So, no, the claims are not 'merely rhetorical', and the complaint is not that "inflation is more than inflation"; rather it is that "the actual debasement of our currency and economy is more than what the lie that the sole arbiter of the mutually agreed to term 'inflation' is telling us it is." The implications of that have been the substance of several months of posting.

Posted by: XON | Oct 4, 2007 10:02:42 AM

Not related to this thread, but the bad-news-is-good-news thing happened again today. From the Yahoo finance page:

"Stock prices are higher in early trading today as some investors read a larger-than-expected increase in unemployment claims as support for another interest rate cut."

Posted by: Rusty | Oct 4, 2007 10:04:48 AM

All that I really need to know, as just a simple Middle-Class worker, making $60K, and trying to hold onto what savings I have, is that when I go to the grocery store, the bill is higher week over week for the same list of stuff. When I pay my energy bills, they are higher month over month, even though I do my best to use less and less.

The price of a Frididaire means nothing to me, clothing is put out in a wide range of prices and quality, and I don't buy new cars. Plasma Screen TVs? Oh, please... I don't even have cable. Very little of what is referred to as "the core" means anything to me.

What DOES effect me, and the very poor, and you rich investor-types up on the Street, is the cost of Food and Energy-- period. Those two things effect ALL Americans, and they must be included in the mix. The problem is that we consumers don't even get to hear about the REAL inflation rate, anymore, as the news agencies have stopped even mentioning it anymore. It's as if we're all supposed to believe that we are just imagining that our CORE BILLS aren't constantly going up.

Many people in America have been priced out of buying many of those items considered a part of the CORE, and THAT is the problem. The CORE applies to a slowly diminishing class of consumers, any more. We NEED both sets of data-- the REAL and the CORE to accurately surmise the purchase power of our dollars-- otherwise, our ability to make proper financial judgements goes out the door.

So, put the esoteric books and charts away, as trying to justify the switch to CORE Inflation as the only measure that matters only breeds cynicism from those of us who actually have to live in the economy that the Street Poindexters have created.

Obviously, no one on the Street actually buys a gallon of milk any more. If the Poindexters actually did, they'd notice that the price of oil to fuel the farm machines, and the demand for corn ethanol, has driven up the price of cattle feed, and blown the price of milk through the roof. That is the inflation that matters when you don't get a Million dollar bonus every year. Believe me-- your nanny feels that pain when she goes home to her own house and family at night.

--mf

Posted by: Monkeyfister | Oct 4, 2007 10:04:56 AM

Begging the question, I still want to know if these people are using the core inflation rate from the BLS as the inflation assumption in their retirement portfolio?

If so, expecting fuel, food, and medical to go up at the government spun rate will leave you pretty short.

Posted by: me | Oct 4, 2007 10:19:44 AM

seems the numbers and reports are for the purpose of purporting SALES

in the land of capitalism and jungle laws - can't you fudge for your own cause?

bank conglomerates / hugh retailers know - they acquired that knowledge - not you all - Uncle Sam asks, but does he buy?

Posted by: Greg0658 | Oct 4, 2007 10:22:59 AM

The CPI and the CPI-XFE decoupled around 1986. Since that time the trend has been for the CPI-XFE to overshoot the overall CPI. Indeed, the core CPI level is still above the overall CPI level.

That is, there has been more total inflation in core-CPI than in overall CPI.

This extends naturally from the fact that as the technology advances commodities prices tend to fall in real terms.

The big question is whether you believe the current run up in commodities is the result of a weak dollar or the result of a rapidly industrializing China. I stand with the latter in which case it makes little sense to restrict money supply growth.

Moreover, I don't understand how people can simultaneously be convinced that housing is overheated and due for a long price correction and have a problem with OER.

If you believe the housing market will correct to fundamentals won't we see falling or flat housing prices until they come back in line with rents?

In that case isn't the appropriate long run view of the cost of housing rents?

Now you might argue that the FED should target asset prices and not allow housing bubbles to be created but that is different from assuming that the run-up in housing has dramatically changed the long run price trend.

Posted by: karl smith | Oct 4, 2007 10:24:08 AM

Well put, mf.

As the masses begin to question the published government statistics, given their obvious disconnect from the stark reality at the local store, cries will be heard up and down the entitlement roster for higher cost of living adjustments. The game will be up.

Posted by: gn | Oct 4, 2007 10:28:04 AM

IMHO, the Fed is fighting inflation by deliberately understating CPI so that all the COLAs (cost of living adjustments) in a whole lot of people's pensions etc will be lower than otherwise.

Posted by: Bill | Oct 4, 2007 10:29:53 AM

http://finance.yahoo.com/q/bc?s=%5ETNX&t=1y

http://finance.yahoo.com/q/bc?s=%5ETNX&t=my&l=on&z=m&q=l&c=

While I'll admit that taking the debating position that eschews inflation is ultimately untenable, for now the two charts above are the judges of whether it's leading to inflationary expectations or not, and the answer is not at the present.

True, the dollar has already deflated the purchasing power of U.S. citizens much more than core would indicate, but the dollar's been low against world currencies before, only to rebound. Much of that potential rebound (should it occur) would offset the dollar's losses against other currencies.

China and India alone are still v-a-s-t depositories of untapped d-e-f-l-a-t-i-o-n, worldwide. It's like they are bottomless pits of deflation, denominated in labor wages.

In the late 80s and early 90s, commodity prices took a nosedive. Farmland fell 50% in some areas. Feeder cattle dropped under 50 cents.
--
Woodshedder,

What you may not know is that Barringo is being a bit sarcastic about the change in Goldman Sachs' commodity index that reclassified the standard gasoline commodity contract deliverable in New York Harbor standard terms. Many suggested at the time that the abrupt and unexpected change was politically motivated just for the purpose of demonstrating low inflation (that whopping drop!) prior to the mid-term elections. I have no opinion about that, but just recant that debate for you.

Consequently, Panzer wouldn't be m-i-s-r-e-p-r-e-s-e-n-t-i-n-g the chart to establish his opinion that core inflation is not worrisome... he'd just be using actual data, h-o-n-o-r-a-b-l-y, to attempt to establish intellectually his own interpretation that it's not worrisome now, and his interpretation that the Fed a-l-s-o interprets it as not worrisome. That'd help clear the way for another rate decrease, and the market almost always interprets that as bullish. That's what you want isn't it? I think you must know that without further cuts, the bullish case isn't so strong.

You ought to send Panzner some flowers.

Maybe these:

http://www.theflowerexpert.com/content/growingflowers/flowersandseasons/pansy

BTW, unless I'm in error, Panzner's not worried inbout inflation now. "It's the least of their [the Fed's] worries, Larry," he said recently on Kudlow.

Posted by: Eclectic | Oct 4, 2007 10:32:45 AM

Thank you, gn.

All that before my first cuppa.

Inre:Bill-- For discussion's sake, Federal Government Servants still working under the GS/FERs system will receive a 3.5% COLA (tied to CPI). But, talking about coming up short: Government Employees working under the nefarious NSPS system will only get a 1.5% COLA, with the balance between the two systems being put into the NSPS "Pay Pool," where those employees will have to "perform to earn" the rest. The Pay Pool determines whatever BONUS an employee will receive vis the performace of his/her co-workers (shares of the Pool). So, under the NSPS System, 1/2 COLA = an "earned bonus." MAGIC! Dark Magic.

THOSE people are going to really feel the bite of REAL inflation, as they lag behind year-to-year. A one-lump-sum, half of which is eaten up by simple COL, coming just in time for the Holiday and Heating Bills, won't stimulate the economy much, and will barely keep them floating. Another whole set of educated cynics with a jaundiced eye for the CPI.

This is just one, off-the-path, demonstration of how the CORE effects many in a very negative manner.

--mf

Posted by: Monkeyfister | Oct 4, 2007 11:01:09 AM