Tim Duy: Misunderstanding the CPI

Tim Duy takes issue with David Leonhardt's discussion of inflation:

Misunderstanding the CPI, by Tim Duy: David Leonhardt at the New York Times purports to explain the CPI – but by the end of the article, I began to wonder if he really understands it himself. True, much of Leonhardt’s analysis is reasonably accurate, and is summed up by a quote from Stephen Cecchetti:

It’s about as accurate as anybody is going to get it.

Unfortunately, Leonhardt’s ardent defense of the CPI is spoiled by his attempt to explain that incoming numbers actually exaggerate inflation:

That said, there is one way in which the official numbers were clearly understating inflation. To track housing costs, the Consumer Price Index analyzes rents, not home prices. (Why? Long story.) And rents didn’t go up anywhere near as much as house prices during the real estate boom. So the index missed the huge run-up in home values that made life harder on anyone trying to buy a first home.

The CPI is a measure of the average price level. Unless you just happen to be the average consumer, your personal CPI will be different, and possibly very different, from the headline numbers. But Leonhardt uses a specific consumer – a first time homebuyer – to justify his claim that using rents, not prices, to measure housing costs understated inflation during the housing bubble. Let’s take a different consumer. Me, for example. My mortgage payment has not changed in 5 years. Assuming I stay in my current house, my mortgage payment will not change for the next 25 years. All else equal, CPI was overstating inflation during the bubble according to my personal experience, and, unless rents fall or I move, will continue to overstate inflation.

Leonhardt extends a faulty analysis further to claim that the CPI will now be overstating inflation since it does not capture falling home prices:

Since 2006, of course, home prices have been falling. But rents have kept rising slowly, which means that, as far as the Consumer Price Index is concerned, housing has somehow gotten more expensive during the real estate crash.

So when the new inflation numbers come out next week, they will indeed be misleading. They will be artificially high.

Inflation may appear overstated to the first time homeowner, but what about the renter? If rents go up even while housing prices go down, how is inflation overstated from their point of view? Again, one cannot use a specific consumer to claim the CPI will be overstating inflation. Either Leonhardt doesn’t really understand inflation, or he is deliberately mischaracterizing inflation to dismiss incoming data as flawed.

Moreover, by dismissing as a “(l)ong story” the issue of why the BLS uses rents, not house prices, to measure housing costs, Leonhardt is missing an opportunity to explain the justification for the use of owner’s equivalent rent (OER). Alternatively, he just doesn’t want to explain OER because the explanation undermines his criticism of the CPI.

The CPI is a measure of the market basket of goods and services consumed by the average consumer. The service the consumer receives from housing is the shelter it provides. If a household can rent their home for $1,000 a month, then that is the price of the service component of their housing. Whatever mortgage amount the homeowner is paying above that $1,000 represents something other than a housing service. From the BLS – the “(l)ong story”:

Until the early 1980s, the CPI used what is called the asset price method to measure the change in the costs of owner-occupied housing. The asset price method treats the purchase of an asset, such as a house, as it does the purchase of any consumer good. Because the asset price method can lead to inappropriate results for goods that are purchased largely for investment reasons, the CPI implemented the rental equivalence approach to measuring price change for owner-occupied housing.

Increasingly, society views the purchase of a home as primarily an investment, not for the service it provides (don’t even get me started on this topic). The investment motive is justification for the wide discrepancy in many locals between house prices and rents. If my house rents for $1,000 a month, why would I pay $2,000 a month unless I saw it as an investment?

The BLS is properly trying to separate the shelter and investment components of housing; only the former is properly included in a measure of consumer prices (there is, however, a separate issue of measurement error). The investment component should be excluded, the same way we exclude the price of gold or a share of GE stock from the consumer price index. By criticizing the use of owner’s equivalent rent to claim that inflation is over- or understated, Leonhardt reveals that he really doesn’t understand the purpose of the CPI. It is a measure of consumer prices, not asset prices. Perhaps he can be forgiven; it is, after all, a “(l)ong story.”

Moreover, the debate over the use of OER in the CPI is something of a false debate. In my opinion, it misses the point entirely. The debate is not whether housing costs are miscalculated in the CPI – the BLS’s basic methodology is appropriate to achieve their objective. The debate is whether or not the Fed should include assets prices, such as home prices, in their policy objective of price stability. Just because there is a valid argument that the Fed should be using a measure other than (or in addition to) consumer prices does not imply that the CPI is flawed. It implies that the construction of monetary policy is flawed. In effect, the BLS is unfairly criticized for the Fed’s policy error.

Comments

[ndd]

I'm with Leonhardt on this one. People who DID purchase houses during the boom really DID pay much higher prices. That is a real cost to real consumers (and 2/3 of Americans own houses), and should be reflected in the CPI.
I understand Duy's argument, but if it is correct that the average American moves every 7 years, then in the last 7 years 50% of Americans have actually experienced housing inflation.
Under Duy's analysis (and owner's equivalent rent), none have. That's a profound distortion of the inflation rate ordinary Americans who purchased houses have faced, imho.
Seems to me the way to accomodate the real world is to use some weighted average, e.g., something like assigning a 10% weighting to changes in median home prices for the last year, 2 years, 3 years,... 10 years for 2/3 of the housing inflation rate, and rent changes for the remaining 1/3.

[reason]

ndd...
I'm not sure about that. Only part of what you paid for a house is owners equivalent rent. Part of it is speculative investment. (You may end up being winning if rents go up in the future.) I don't think the DJIX belongs in the CPI, so I doubt that housing prices do either.

[Enigma]

I'm not sure how much into Austrian school economics, but Ludwig von Mises argued that the entire idea of price indices is logically invalid. Basically, it is a meaningless number that is pseudo-scientific.

There is a long article here attacking the idea of price indices from a deductive reasoning perspective. What do you think?

[ndd]

Tim Duy's link about house price inflation leads to this article by Lloyd Norris in the NY Times:

in practice, in an economy where most people own and renting is not easy in many areas, it greatly understated the rate of inflation during the early part of this decade. And now it is overstating it.

From the end of 2001 through last June, the rent figure rose at an annual rate of 2.7 percent a year, below the C.P.I. average of 3.0 percent. But the average cost of an existing home rose at a rate of 7.6 percent a year, and another measure, the S.&P./Case-Shiller index, climbed at an even faster rate, of 11.5 percent a year. Given that the house figure accounts for more than a fifth of overall C.P.I., using real house price data would have made inflation look worse, and might have encouraged the Fed to back away from its policy of super-low interest rates earlier than it did — and perhaps take out some of the air in the bubble in home prices.

Now, however, the reverse is going on. Since June, the owners’ equivalent rent has risen at an annual rate of 3.7 percent, double that of the Consumer Price Index, which has risen at a 1.8 percent rate. But home prices are falling in many places. Through February, the latest data available, the average price of existing home sales has fallen at an annual rate of 9.4 percent.

http://norris.blogs.nytimes.com/2007/04/17/exaggerated-inflation/

Unfortunately, the comment format doesn't allow us to post graphs, but Tim Iacono of The Mess that Greenspan Made has an excellent graphic representation of what Norris and Leonhardt are talking about:

http://www.iaconoresearch.com/BlogImages/07-08-24_cpi_with_cshpi_or_oer.png

Seems to me, the same house prices are faced both by investors and those simply looking for a place to live. Personally, I prefer that the Consumer Price Index measure, you know, actual prices.

[reason]

ndd

Personally, I prefer that the Consumer Price Index measure, you know, actual prices.

It does - the disagreement is just about which particular prices belong in this particular index.

[reason]

enigma...
all well and good - in fact in general I agree, but it is just as difficult to define what we mean by the "money supply". All he says (if he was consistant) is all economic statistics are measures of not very good proxies. Using them, we may not be flying blind, but the focus is not very sharp and we are as often as not looking backwards.

[ndd]

Tim Duy:

Leonhardt reveals that he really doesn’t understand the purpose of the CPI. It is a measure of consumer prices,

reason:

the disagreement is just about which particular prices belong in this particular index.

Let me get this straight:

"The bast way to measure consumer prices is to ignore the actual prices that actual consumers actually pay for the biggest item in the index."

I see.

[snark on] Where are Hillary Clinton and Bob Menendez when you need them? [/snark off]

[reason]

ndd...
No the disagreement is what in the fact biggest item actually represents (i.e. removing from the combination imputed rent, speculation, inflation hedge the bits that aren't imputed rent).

[ndd]

reason, thanks as always. But at this point we simply disagree, and that disagreement has had profound effects. As Tim Iacono said at his blog, if the Fed had taken into account the surge in inflation due to housing prices, it would have treated interest rates far differently, and we would probably not have the worst-since-the-depression credit/liquidity/solvency crisis we are having now.

[reason]

ndd...
I'm not sure how we could actually do this particular part of the CPI any other way. Very many home owners in fact make no mortgage payments because they have paid upfront. What you are paying now, also includes payments for the future. Looking at the rental price seems to me the fairest way to do this.

[reason]

ndd...
In fact (responding to the bit where we crossed) this argument could be applied to any asset price bubble. This argument is hardly new. The fact that in this case a change in the definition of the CPI may have helped is incidental but not in principle correct.

[kharris]

Once you recognize that housing serves more than one purpose - shelter, investment and whatever else it may do - then you are implicitly recognizing that putting a price on it for the purposes of assessing consumer prices is complicated. That leaves anybody who brings a "should" to the discussion with some obligation to addressing the advantages of the BLS process relative to any other process for accounting for housing prices.

For instance, do we want the Fed to respond to a speculative rise in land values by tightening monetary policy? If so, are we willing for the result to be that a wide variety of other prices have to fall persistently in order to bring overall CPI down to target when land prices are rising? Are we willing for the Fed to ease to the point that a wide variety of other prices rise persistently when land values are falling? (Which, by the way, seems to be the case right now, though largely as a result of events outside the Fed's control.) Do we want federal benefits to be subject to swings in land prices, when some beneficiaries may enjoy a rapid rise in wealth due to rising land prices, and can realize gains on that wealth by selling their home and renting?

If we don't want the Fed to jerk the prices of consumer goods and services around in response to land price fluctuations, and we don't want benefits payments to be increased and decreased largely in response to land price fluctuations, then what is the point of linking CPI more strongly to land prices?

[paine]

good post
as good as you can expect on this subject
given the briar patch at its core

teasing us with the service /investment dichotomy

subtracting the estimated
capitalized expected ground rent stream ...eh reason ?

some high style metastatics

[paine]

"I understand Duy's argument, but if it is correct that the average American moves every 7 years, then in the last 7 years 50% of Americans have actually experienced housing inflation"

some prolly wrong assumptions here

the actual annual residence move
might be a useful number however

but look at it this way
if the rental market for a comp unit is n bucks a month
that's the shelter cost

now if the comp unit is not for rent
cause rentals and self owned are always non comp shelters...

if self ownership has a value in itself ...

[paine]

"do we want the Fed to respond to a speculative rise in land values by tightening monetary policy? "

great question

if its a lot value bubble then ...yes

another double haeded beaut:

" are we willing for the result to be that a wide variety of other prices have to fall persistently in order to bring overall CPI down to target when land prices are rising? Are we willing for the Fed to ease to the point that a wide variety of other prices rise persistently when land values are falling?"

brings out the question
whast is the dynamic relationship
between asset prices and product prices

the science needs to do
more work on this
much more work

wealth effects have a deus ex machina
aspect to them at least as far as
inside
the policy mac-models
in common use today

[Gerard MacDonell]

Breath of fresh air there. The only thing I would add is that the general price level is the ratio of money spending to the utility derived from that. Utility cannot be aggregated across individuals and so is not measurable. Accordingly, the general price level is not measurable and inflation is thus not measurable. To quibble over various attempts to do the impossible is a waste of time. Maybe we can say inflation is now sort of low, but beyond that it is really impossible to say.

[bakho]

This is a silly argument. The Fed needs to use all the information available to understand how monetary policy in the present will affect the economy in the future. The CPI is only one measure of the economy, but it is important for setting entitlements and setting income increases for many people. CPI also has many components. Rent is only one. Small changes in rent inflation will be even smaller when averaged into the CPI.

The housing bubble primarily applies to a few coastal areas of the US. In the Midwest, housing prices have been relatively flat during the bubble years.

[d_rumsfeld]

Isn't the CPI composed of some 82,000 items measured monthly? It seems reasonable to measure the largest component with a bit more resolution.

As mentioned above, two-thirds of americans own homes. One-third of those have paid off their mortgage. So one-sixth of americans have only maintenance costs. That leaves half of Americans who have mortgage payments. Those mortgage payments can be estimated using a distribution of home prices and housing turnover. The final third of american's rent. These folks should be accounted for using rental indices.

That doesn't seem much more complicated, but it does seem more accurate than OER.

[reason]

One interesting argument I saw elsewhere, pointed out that home owner occupiers may pay no rent, but they are also (to the same degree) forgoing rental income that they would otherwise receive. Does forgone income belong in a price index?

[vimothy]

Anyone familiar with John Williams? The US government has been manipulating CPI stats for a long time. Here's an interesting and recent interview with Williams from this month's Harper's Mag:

http://www.tampabay.com/news/article473596.ece

Pretty germane...

[Eric Dewey, Portland OR]

Probably an unpopular reminder, but anyway - isn't price just a substitute marker for perceived value?

Does our ability to measure prices suffer when we are in fundamental disagreement over the underlying questions of value - for example, the value of a house as shelter as opposed to investment?

If "we" are now viewing houses more as investment than as shelter, does that shift in fundamental values drive some of what we've been experiencing in the housing economy?

[oas smoas]

"ndd says...
...if it is correct that the average American moves every 7 years, then in the last 7 years 50% of Americans have actually experienced housing inflation. "

er, no. either they move to relocate so transaction is neutral in price, or they move to up or down size, the price difference of which is predominantly an investment component.

[kharris]

Eric,

Assuming that price is a substitute for perceived value ignores consumer surplus. Price is equal to perceived value only for the marginal buyer. All other buyers have enjoyed a price lower than they would be willing to pay, which I take to be what you have in mind with "perceived value". That is the point exploited by monopolists, oligopolists and the like.

[Charles]

Tim Duy might be right from a purely methodological standpoint: the home buyer is paying the premium over what he could rent the house for because s/he sees it as an investment and investment is not measured by the CPI.

But statistics need to be anchored in reality. The collapse of the savings rate, for example, could mean that we've all become irresponsible spendthrifts. Or it could mean that real wages are somehow being miscalculated. The rise in bankruptcies could be because there's not enough moral stigma against it... or it could be because people don't have enough money. Americans might work longer hours (versus the 1960s) because we hate our spouses... or it could be that we're desperately trying to stay afloat.

When one looks at the broad mass of statistics describing Americans, we look like a people getting poorer and poorer, at least in the lower four quartiles.

And a factor that would provide a unifying explanation is a miscalculation of the inflation index.

(Crossposted from Calculated Risk)

[cm]

reason: Is opportunity cost a "consumer price"? If yes it belongs in the "consumer price index". (And BTW why stop at rental opportunity costs? I can think of many other categories.)

Remember, the CPI supposedly was initially created to approximate cost of living trends during a period of (direct) wage/price controls.

[plschwartz]

I want to sort of nail down something that Kharris said:
"Once you recognize that housing serves more than one purpose - shelter, investment and whatever else it may do - then you are implicitly recognizing that putting a price on it for the purposes of assessing consumer prices is complicated."
I fell that it is an insight that goes beyond CPI to a truer understanding of housing and something that might possibly be considered when discussing a housding "bailout"
plschwartz

[bakho]

The Fed does not base monetary policy on CPI alone. They are not stupid robots running a simple calculator. They have ways of considering asset prices in setting monetary policy. Making large changes in calculating economic parameters creates difficulties in interpreting historic data. During the 1980s, the way unemployment is calculated was changed. Unemployment rates in 2008 cannot be directly compared to 1975 without recalculation. There is no problem with the CPI and there is no monetary policy failure. There is only a problem for models built around CPI, that fail to include enough other information.

Models will always fail in some situations. This is why monetary policy is set by a board of humans and not a computer. Computer programs can produce good monetary policy most of the time, but only if the conditions fall within the parameters that the model anticipated. Crisis such as the current one can be used to refine the models to account for new situations.

[a]

"The debate is not whether housing costs are miscalculated in the CPI – the BLS’s basic methodology is appropriate to achieve their objective. The debate is whether or not the Fed should include assets prices, such as home prices, in their policy objective of price stability. "

Isn't the CPI-W used to adjust Social Security benefits? So to say that the debate is only about what the Fed does or does not do with the CPI seems to miss the point. It's not just the Fed which uses it.

[ScottB]

I think that BLS has it right this time. I agree with those above who note that many homeowners had no increase in housing costs (mortgage paid off, or mortgage unchanged), and that homebuyers who were already homeowners didn't see that much of a net price increase due to appreciation of their previous homes. First-time homebuyers in higher-priced areas were the ones who were forked, but judging by the foreclosure rate, they won't be for much longer... if we were able to adjust for them, we'd also have to adjust for all the folks who will buy at the over-corrected bottom coming up in a couple of years...

[kthomas]

Reminds me of something one of my college English profs once told me - that those who make the definitions, rule the world.

[me]

You can see the difference at John Williams Shadowstats. He shows the new hedonic crap versus the original way it was calculated. You have to be brain dead to accept owner equivalent and this other BS.

Every body worships Paul Volker, does he think inflation in under control?

http://www.shadowstats.com/

[Syaloch]

I had this very argument a couple of years back with a former co-worker at BLS (he works on the ECI, I used to work on the PPI).

Duy clearly has a better understanding of the purpose of the CPI than Lloyd Norris, who confuses the paper asset price of a house with what the typical consumer actually pays for housing services on a monthly basis. Indeed, wasn't the housing bubble was largely the result of low interest rates and creative financing that allowed people to get more house for the same monthly payment?

However, I would quibble with Duy that the difference between the rental price and mortgage payment is entirely due to the investment value. Anyone who's ever made the transition from renting to owning knows that you have a much greater ability to customize a property you own, and that this has significant value.

[robertdfeinman]

Factoid:
I own my house and the mortgage is paid off, but the costs have gone up just the same: electricity, gas, repairs, homeowner's insurance and real estate taxes have all risen.

If I was still paying off my mortgage then the monthly payments for this (it was a fixed rate loan) would be the only component that was constant.

I don't see why the actual cost of home ownership can't be included in the CPI in addition to those who rent. They just need to weight them in proportion to the population.

If the nominal value of my home goes up, but I don't sell then how does this affect my expenses? Even if you include the fraction of those who sell each year and then include their capital gains in the calculations this shouldn't make much difference, since the majority of sellers just buy somewhere else, presumably at a price equal or higher than the prior home.

Maybe there needs to be a cash flow number which is tracked in addition to the CPI.

[Jesse Weiher]

All of you need to stop reading opinion pieces on the CPI and start reading actual academic research. My first suggestion is the Boskin Commission report: http://www.stanford.edu/~boskin/Publications/CpiCommission.pdf

The CPI is meant to measure movements in "consumer" prices because it is meant to track changes in consumption costs. Ideally, it helps assess changes in the Cost of Living. The purchase of a durable good like a home is not a consumption purchase. Ergo, home prices are not included in the CPI. A change in the value of the home that is not accompanied by a change in rental rates does not change someone's Cost of Living, because they can simply switch from living in owned housing to living in rented housing. What it does change is the return they can expect to get from their capital investment in housing stock.

The point made in previous posts that buying a home is one of the biggest purchases an individual can make is moot. It is not a consumer purchase. The fact that it is a big purchase does not make it a consumer purchase. Just like yelling vociferously doesn't actually make someone right.

People buy stocks, art, gold, wheat futures, mutual funds, certificates of deposit, bonds, and homes as part of investment. None of these are included in the CPI, because none of them are consumer purchases (unless you'd like to burn or eat the paper your stock certificate is printed on).

The NIPA account tables define the purchase of investment goods as saving. Including investment purchases into the CPI would not make the CPI more accurately reflect of the cost of living. It will simply conflate the implication consumption and savings decisions. Savings decisions affect your future income levels and not the current Cost of Living.

There are many reasons (highlighted in the Bosking Commission report and subsequent research) why the CPI may not be an accurate measure of the Cost of Living. The fact that housing is not included is not one of them.

[Patricia Shannon]

I'm surprised nobody's talking about the exclusion of food and gas from so-called "core" inflation, a definition I consider purely political. If it is not, than it is purely idiotic. I have to have food to live. In our society, we (including people on social security) have to buy gas, directly and/or indirectly. We can do w/o things like cell phones.

[Eric Dewey, Portland OR]

kharris, thanks for clarifying for a non-economist. But I'm not sure that the clarification doesn't amplify my point.

Is it possible that we've experienced oligopolistic behaviors in the consumer housing market that have skewed prices, including the OER?

The standard rebuttal to arguing that there is a national real estate market has always been that real estate pricing is regional. But buying a house for most people is *really* buying the money to buy a house - in other words, a mortgage.

And the mortgage market has been becoming increasingly national over the past 20 years - powered by the MBS market.

If there has been extensive exploitation of "perceived value" over an extended period of time by an oligopolistic mortgage market, why wouldn't that skew the OER, and thus the CPI measure?

(The terms "buyer" and "seller" may actually confuse things here: in this argument a "buyer" would purchase mortgage-backed securities, which would have the effect of simultaneously "selling" funds to make additional mortgages in a constant feedback loop - so the traditional roles become blurred).

[Jesse Weiher]

Patricia Shannon, you are mixing the word "core" with the word "important." Don't do that. Just because someone used the word "core" doesn't mean they have a political agenda. It is a term invented by egg-head scientists, not politicians.

Gasoline and food prices fluctuate dramatically over the year. Holding the price of everything else constant, the CPI can grow and shrink dramatically because of these seasonal fluctuations.

The term "core" inflation simply refers to that part of inflation that does not change in a dramatically seasonal manner. Often times people will say "inflation was at 2.5% this quarter, but most of that was due to seasonal changes in food costs. Core inflation only rose by 1.5%"

It is not meant to say that the dramatic price increase in food or gas in unimportant. It is simply meant to say that these prices are more highly variable.

[paine]

"Every body worships Paul Volker"

please ...say it ain't so

[paine]

buying a house ???

the durable productpar excellence

beyond the car tv pc bath tub

but you don't buy any of them every cpi reporting period either

so ....
should all durables be figured at current lease rate ??

notice
absurdity looms and rather too quickly
if we move in either direction here

[paine]

"All of you need to stop reading
opinion pieces on the CPI
and start reading actual academic research"

indeed look what fruit it bares...

" The purchase of a durable good like a home is not a consumption purchase"

[cs]

Tim Duy explains this nicely. I live in SoCal and one of my neighbors moved to CO and is renting his house. The rent pays both mortgages.

If the newsmedia understood the BLS methodology, maybe they would have spent some time explaining to the public years ago that the difference between their mortgage payments and the relevant rent constitutes a punt on real estate. The problem is in the public perception, not the reality -- which the BLS captures reasonably well.

[paine]

pat s

i know you don't read my stuff cause of its format
but .....

the core rate is a proxy for unit wage costs

oft repeated proposition:

the ultimo purpose
of the fed's inflation watch
is to control
the rate of change in wages
and what prices can wage increases change ??
non commodity product prices
right ???

once the fed;s aim is absorbed
excluding commodity prices
asset prices
any other prices in fact
follows immediately

the cpi
is really an official index
primarily used
to adjust stuff
stuff like adjustable t rates
and social security payments

politics ???

hey what would you have it be ???
yes we could maybe change the politics
to ...say ...
your jobbled class politics
not the rentiers politics
we see reflected in fed index watching today

[JKH]

Jesse Weiher:

Thank you.

Your explanation @ 10:43 is the clearest and most correct of all the comments here, including the post itself.

Everybody who visits this post should read it.

(BTW, one of the flip sides to this misunderstanding about the purpose of CPI is the equally incorrect suggestion that the definition of NIPA saving should include marked to market securities gains and capital gains on other types of assets, including housing. That is not the purpose underlying the definition of saving. That idea is captured under wealth rather than saving.)

[paine]

"It is simply meant to say that these prices are more highly variable."

that's the superficial glib daily paper take

seasonal flux ???

food prices and energy prices aren't in seasonal flux
right now

the underlying rate of inflation or trend rate of inflation
is clearly higher then core rate calcs indicate

the core index is all about unit labor costs
in other words
price change in products
after crudely abstracting
"the magnificent dynamics" of various economic rents

[paine]

"Your explanation @ 10:43 is the clearest and most correct of all the comments here, including the post itself."

here's the nugget of truth in that claim:

" A change in the value of the home that is not accompanied by a change in rental rates does not change someone's Cost of Living "
hold on to that tightly
and you oughta stay sane
thru most roller rides

[paine]

NIPA talk depresses me
what are we here
but a gaggle
of suspicious crotchety pensioneers

[d_rumsfeld]

Ignore my post from above. The CPI for 2008 does include both a rental portion and the Owner's Equivalent Rent portion, at 5.8% and 23.9% of the CPI total, as per the NYTimes infographic currently linked to on BigPicture and Calculated Risk.

So two questions for anyone knowledgeable. First, if (round numbers) 33% of the population rents, why is rent less than 25% of the OER weight in the CPI index? Is this weighting related to the distribution size of rental housing vs. housing?

Second, why would rent change at a different rate than owner's equivalent rent?

[Jesse Weiher]

paine

Actually, food and gas prices ARE in seasonal flux right now. They also happen to be experiencing some pretty strong price increases that are NOT due to seasonalities.

The fact that gas and food prices are at record highs does not negate the fact that prices for these goods tend to move seasonally.

Energy and food crises do not suddenly change harvest seasons and peak driving seasons.

And OF COURSE core inflation is not tracking actual inflation very well. Since we are currently experiencing a food and energy crisis. What is it with these Johnny-Come-Lately/What-Have-You-Done-For-Me-Lately types that think the only thing worth paying attention to is the current state of the world. It's not like the term "core inflation" was invented with this particular situation in mind. It's also not like core-inflation will never again be a worthwhile statistic to look at. It simply happens to be that now is not the time for focusing on core inflation.

I swear, some of you wouldn't last a minute if you actually had to set economic policy

[kthomas]

James, you seem passionate, but that fourth paragraph....what the heck are you talking about?

[Bruce Wilder]

JKH: "the definition of NIPA saving should include marked to market securities gains and capital gains on other types of assets, including housing"

Kind of has to, doesn't it?

It is a consequence of double-entry bookkeeping. The income statement of a business enterprise is reconciliation of a beginning and ending balance sheets for the same reason, and a capital gain or loss, therefore, appears on the income statement.

The logic of NIPA makes NIPA savings a plug to keep the accounts balanced.

And, it is not completely wrong as a model of economic behavior. A large number of people did take advantage of the derangement of the mortgage market, to extract equity from their homes, in lieu of savings or to finance consumption.

What's your point?

[paine]

jesse

you might not need to justify your take on
the cpi rate vs core rate
if you focus on my point

wage control is fed policy sine qua non

all else is eye wash
but if you are a retired pensioner
then for u
cpi indexed vs non indexed income streams
become sine qua non
and if the composition of the relevent index
shows the biggest adjustment due
that might make your mouth water
but if another index composition
would increase your due adjustment
why then a brutal injustice has been served unto u

" some argue as if
justice itself
outha work like a magic spell "

[paine]

"I swear, some of you wouldn't last a minute if you actually had to set economic policy"

you're right
i'd get hanged by a wall street lynch mob

[ndd]

Well, this is what I get for leaving for work....

P

" The purchase of a durable good like a home is not a consumption purchase"

Of course, then we don't "consume" flat screen tvs, furniture, cars, books, etc., either.
______

cs:

"If the newsmedia understood the BLS methodology, maybe they would have spent some time explaining to the public years ago that the difference between their mortgage payments and the relevant rent constitutes a punt on real estate. The problem is in the public perception...."

Maybe the public doesn't perceive that they are making a "punt on real estate" --given the value they place in owning a residence and all that ownership includes (including changes to the physical property) -- because, they're not?

[paine]

ndd

your wit is so dry here
it blows right past me
are you underlining my sarcasm
or after a mis reading
independently rediscovering it

[me]

Jesse Weiher

You don't think Boskin and Greenspan were trying to eliminate/limit increases in social security payments do you? You opinion piece is no more credible that the ones we read.

Originally there was a fixed basket of goods that was measured. Now it is basically hocus pocus. Equivalent rent, cheaper computers, substitution of hamburger for lamb, whatever is cheapest. Not even close to the real world.

[ndd]

Paine

You are delphic

I'm covering my bets.

[SanFranciscoJim]

I just had an "a-ha" moment reading these comments. The reason that American savings rate has seemed so low the last decade is because Americans have been investing in their homes instead of more traditional investment vehicles. Savings haven't really been that low, they have just been placed in what was perceived to be a less risky and better returning vehicle than stocks and bonds and CDs.

It took the comment Once you recognize that housing serves more than one purpose - shelter, investment and whatever else it may do - then you are implicitly recognizing that putting a price on it for the purposes of assessing consumer prices is complicated.. for me to really put that together.

Thanks, kharris for your insight.

[JKH]

Bruce Wilder:

For the first 7 years of this decade, household wealth as calculated in the Federal Reserve Flow of Funds reports increased by about $ 20 trillion (roughly 50 per cent), whereas the household savings rate was near 0. Wealth is contracting now due to declining prices, but the savings rate isn’t going negative.

Household NIPA savings are calculated from various types of income, not capital gains.

(As far as households and equities are concerned, only dividends from non tax sheltered accounts are included in household income.)

[richard]

I'd assumed some time ago that real estate purchases could be partitioned into (shelter, consumer item, investment), with the proportions difficult to identify.

We know it's a shelter hedge, because buying a small house and paying down the mortgage is one sure-fire way the poor and elderly have lived within their means.

We know it's considered a consumer item, because people will buy a much nicer place to live in than they will to rent.

And we know it's a speculative investment item, based upon recent behavior.

Making estimates to allocate that triple, even in the aggregate, is not easy.

[don]

Tim - nice post. And you use the power of understatement.
"but by the end of the article, I began to wonder if he really understands it himself."

As you so clearly show, either the author doesn't understand the CPI at all, or he is deliberately confusing people.

Asset prices are the present discounted value of anticipated future returns. Thus, 'Consumer price targeting' and 'asset price targeting' can yield different monetary policy responses, although in both cases, there is an element of expectations being targeted. But the fed has only one instrument (the money supply) to target prices, employment and anticipated future returns to assets. I'm not sure they would gain anything by adding more targets to their list. Given their track record, I think they would do better to just target consumer prices.

[paine]

don
the fed has regs to control credit flow rates into various sectors

its micro managing but its necessary
if sectoral asset bubbles are to be pre empted

general rules and regs
created by congress
specifically enforced
by the fed

needless to add
this will require
a large
"independent "***
fed staff
with brains and ...guts

btw ****
i don't mean independent of congress
i mean independent
of those organizations they are regulating

[don]

Paine -

Pat S isn't the only one that has been ignoring you on account of your format, but I've now learned some respect.

Neat point about why the fed uses 'core' inflation.

[cm]

Jesse Weiher:

From Merriam-Webster:

Main Entry: fluc·tu·ate

intransitive verb
1 : to shift back and forth uncertainly
2 : to ebb and flow in waves

Since I came to the US, I have sure seen the "forth" and "flow" parts, the "back" and "ebb" parts not so much, when it comes to food, services, and transportation. (With the notable and pretty much sole exception of seasonal fruits and vegetables.)

Over the whole period as well as annually.

[Charles]

San Francisco Jim says, The reason that American savings rate has seemed so low the last decade is because Americans have been investing in their homes instead of more traditional investment vehicles.

That's one interpretation, Jim. But one really has to ask why so many Americans are investing themselves into bankruptcy or foregoing medical insurance or their children's educations for the sake of investment.

Another interpretation is that the CPI adjustment is seriously off. Certainly if you talk to people, they feel poorer, more anxious, less optimistic about the future. These are not common sentiments among investors.

[BJ Feng]

I don't think people or the news media understands that CPI doesn't measure what it used to measure. CPI no longer measures price changes in a basket of goods, it tracks price changes for a given standard of living or utility level. So if a computer gets better, then you derive more utility from it, your standard of living increases, so the CPI has to be adjusted downward to hold your standard of living and utility constant. Unfortunately, this adjustment is no longer transparent, we're getting less and less information on what they are adjusting probably because some of the former things they adjusted were questionable.

Just because you have 200 cable channels, is your cable service 10x better than when you had only 20 channels? Interestingly enough, the cost of tuition has also been adjusted downwards to reflect the greater quality of college education year after year. I can't remember the explanation, perhaps something to do with more money spent by the universities on students, only that it wasn't very good or satisfying.

So what you ask? Surely a desktop computer today is much much better than one from 10 years ago so to compare apples with apples, we need to adjust downward.

The problem is that a computer that cost $1500 ten years ago can't run jack today! That computer is absolutely useless and worthless. A computer of comparative power relative to today's software needs would cost around $900, but the CPI doesn't take into consideration comparative power, only that absolute computing power doubles roughly every 1.5 years, so huge downward adjustments are made to computers on top of the decreases in prices.

A young rich person who puts all his wealth in an account that adjusts for CPI will be poor by the time he is old. The CPI tries to hold standard of living constant, but we're used to increasing standards of living as time passes. Think of the standard of living for a wealthy individual in 1800. He might have an indoor "water closet", and correct me if I'm wrong, but he probably doesn't have indoor plumbing! He has no incandescent light, only candlelight that is multiples fainter. His horses can't compare to today's cars.

A person today who lives like the wealthy did in 1800 would be considered dirt poor! No plumbing? No light bulb? Yet going by CPI, all those inventions would have to be adjusted for to keep constant the standard of living. Because standard of living rises over time with new technology, holding it constant guarantees poverty.

Those of us who will retire 40 years from now should expect new inventions that better their quality of life. If 40 years from now, you have the same standard of living as you do today, you will be unbelievably poor by the standards of the future, even if your standard of living today is quite good. Thus, anyone who only increases his savings by CPI will find that he is growing poorer over time. His savings won't buy him the new stuff that he will need in the future to survive.

Beware, the CPI no longer is what it used to be! The numbers of the past can't be compared with the numbers of today! As usual, journalists are too ignorant to understand and still report CPI as if it was still the CPI of old.

[reason]

cm
I don't think you understood my point fully.
It is possible to decompose owner occupied rent into:

a: - Rent paid to self
b: + Income from Rent paid to self.

So the rent is a real cost, but we have changed the income part of the equation. The income doesn't belong in the CPI however, so treating the cost to home owners of their pre-paid property as zero is effectively including income (that might otherwise have been an investment income for instance) in the CPI. Imagine if owning your own house was banned and everybody paid rent. Then people would have alternative investments for their savings, (so there would be a measured income change) but rental costs would be the same. Essentially, I'm saying why should peoples portfolio decisions effect the CPI.

[BJ Feng]

http://www.post-gazette.com/pg/05129/501565.stm

"An inflation debate brews over intangibles at the mall
Monday, May 09, 2005
By Timothy Aeppel, The Wall Street Journal

WASHINGTON -- To most people, when the price of a 27-inch television set remains $329.99 from one month to the next, the price hasn't changed.

But not to Tim LaFleur. He's a commodity specialist for televisions at the Bureau of Labor Statistics, the government agency that assembles the Consumer Price Index. In this case, which landed on his desk last December, he decided the newer set had important improvements, including a better screen. After running the changes through a complex government computer model, he determined that the improvement in the screen was valued at more than $135. Factoring that in, he concluded the price of the TV had actually fallen 29 percent.

Mr. LaFleur was applying the principles of hedonics, an arcane statistical technique that's become a flashpoint in a debate over how the U.S. government measures inflation. Hedonics is essentially a way of accounting for the changing quality of products when calculating price movements. That's vital in the dynamic U.S. economy, marked by rapid technological advances. Without hedonics, the effect of consumers getting more for their money wouldn't get fully reflected in inflation numbers."

What is disturbing is that a relatively small number of people are making largely subjective decisions that greatly influence the CPI and the COLA payments to Social Security retirees. I had the same thoughts as the blogger below, this is from 2006.

http://bigpicture.typepad.com/comments/2006/11/
ppi_hedonic_adj.html

"Consider this quandry: How is it possible that prices throughout the entire pipeline, from raw materials to finished goods, managed to stay tame no matter how high energy prices went during the past 3 years? The riddle is answered by observing that government models are gamed to show as little inflation as possible; Otherwise, the COLA obligations would be going through the roof.

Bill King (of The King Report) makes a similar observation: "Isn’t it interesting that PPI didn’t surge when oil did, but when oil declines sharply PPI plunges?"

Consider his observations on the BLS PPI data:

“Prices for light motor trucks fell 9.7 percent following a 3.5-percent gain in the preceding month. From October 2005 to October 2006, the index for light motor trucks dropped 12.4 percent…In accordance with usual practice, most new-model-year passenger cars and light motor trucks were introduced into the PPI in October. (See Report on Quality Changes for 2007 Model Vehicles, USDL 06-1973.)” Quality changes produce hedonic adjustments to prices. Ergo the large drop in vehicle prices is fiction. It’s the work of BLS bureaucrats, the Winston Smiths from “1984”.

The ‘quality’ or hedonic adjustment to light vehicles is $392.10/vehicle. The BLS reduced the actual costs of these vehicles by $392.10 ERV. For autos the BLS adjusted the real price $139.96 lower. So as we have maintained for years, PPI and especially CPI are constructed so that they can’t show actual inflationary changes or pressure." (emphasis added)

So the anti-inflation question at hand is simply this: Did real prices fall, or was this function of statistical sleight of hand? You may recall the headlines for October Retail Sales noted the role of autos and light trucks: Autos Save Retail Sales From Sharp Decline."

I remember reading so time ago that the BLS isn't publishing their hedonic adjustments anymore, I certainly couldn't find it in the detailed tables of published CPI data. That's why it's next to impossible to isolate the effect of the adjustments on today's CPI. I could be wrong, but there are a lot of dedicated bloggers who are aware of the adjustments, but I haven't seen actual numbers like the one above, for quite some time. PIMCO's Bill Gross also thinks the hedonic adjustment methodology is flawed.

http://moneycentral.msn.com/content/P92951.asp

But as you'll see, the core rate is not Gross' major beef. It's the nonsensical extremes to which hedonic adjustments have been taken, something that my good friend Jim Grant has written about on many occasions as well. Though the hedonic adjustments began as a way to turn computer horsepower into price reductions, they have since been applied to lots of other goods. As Gross notes:

"In 1998, the methodology was adopted for computers -- surely the biggest step backward in realistic inflation calculations. Since then, the BLS has expanded the concept to include audio equipment, video equipment, washers/dryers, DVDs, refrigerators and, of all things, college textbooks! Today, no less than 46% (my emphasis) of the weight of the U.S. CPI comes from products subject to hedonic adjustments," Gross wrote.

[Patricia Shannon]

Paine,
I don't entirely skip your comments. I usually at least look at them. I just don't spend a lot of time trying to figure out what you mean when I don't get it right away. I know you often have good ideas.

[cm]

reason: I see your point, but then every purchase/expenditure or consumption pattern ("choice") has an opportunity cost effect (and not necessarily only a financial one).

I generally agree that imputed prices can always be questioned for the very reason that the imputation imposes a "subjective" value judgement. But so does what is included in the index, weighting coefficients, etc.

[cm]

Patricia: No offense please, but sometimes insights are not fed by the spoon.

Having said that, lack of time is an entirely legitimate reason for skipping the effort.

I also skip many contributions by specific authors, but for the reason that I expect drivel, not tough fare.

[reason]

cm.
I take your last comment as flattery.-) It seems at least I sometimes don't write drivel.

[Borrowers]

"My mortgage payment has not changed in 5 years."

You are a heavy borrower. A policy that results in you not having to pay back everything you borrowed works to your benefit. (You do not have to pay an inflation adjust on the principal remaining on your mortgage, so inflation slowly erodes the purchasing power of remaining debt.) Thus you think inflation is just great.

However, inflation working to the benefit of borrowers does not mean that it is a productive policy for the nation as a whole.

[Debt]

The argument seems to be, "Inflation is over stated because I don't pay any annual increase on my debt."

Debt is not a consumer item. It is being used as a zero sum transfer of purchasing power. Fixed price debtors gain real purchasing power from an inflation of consumer prices, but someone else necessarily loses an equal and opposite quantity of purchasing power. Necessarily. That is, the personal inflation rate of heavy debtors is lower than average, but the personal inflation rates of counter parties in the transaction are higher.

[Quality]

"A young rich person who puts all his wealth in an account that adjusts for CPI will be poor by the time he is old."

This statement would seem to be borne out by the experience of real life people. The CPI adjusted wage is higher than it was 40 years ago, but one wage is no longer enough to support the average family with. In the post WWII era, one wage could easily support a family unit. If the CPI accurately measured consumer prices, the same CPI adjusted wage should easily support the same family unit today.

The argument seems to be that computers and penicillin did not exist at the turn of the century. The fact that they exist now is justification for creating enough new money to drive the price of food/homes/essentials to 40 times what they used to be at the turn of the century. This is nonsense. Computers and food are completely different products. The price of food should be held constant, not increased when engineers figure out how to put more circuits on a chip.

Maintaining a steady price of essential items is the only way to maintain the standard of living per hour worked. This is so because newly created money is not distributed proportionately to the entire population. Only a small percentage of the population directly receives the new money. This increases the purchasing power of the recipients, but reduces the purchasing power (standard of living per hour worked) of those who don't directly receive the new money.

The only way for everyone to receive an inflation adjustment to their income would be for every single citizen to receive a proportional share of newly created money. If the newly created money is directly distributed to only a few, only the recipients gain. Everyone else loses, since this is a zero sum transfer.

[BJ Feng]

Inflation is understated because the hedonic adjustments that the BLS makes is subjective. The changes have been gradually made since the 1990's, perhaps people haven't had enough time for it to sink in and realize what is going on. There are two CPIs, one before the 1990's and one afterwards, the two simply cannot be compared because the way they are calculated are completely different.

I don't understand why this is a non-issue for liberals who finally have a legitimate complaint, perhaps because it was implemented during the Clinton years. Perhaps because they trust that government workers can accurately tell how much a given product has improved by. After all, the same kind of people would be setting medical prices and oil prices if they had their way.

[cm]

reason: I cannot imagine I was unclear, but I used 'reason' literally, not in reference to your moniker. And I didn't want to single out particular names, so I left it at 'specific authors'. I guess you get my drift.

[reason]

cm...
You are forgetting that I can (surprisingly) reason. You had earlier responded to me, so logically that meant I sometimes don't speak drivel (as you read my comments), but clearly some other people always speak drivel.