Inflation Ex-Inflation Ex-Inflation
I stumbled across a data point yesterday that was quite fascinating: The percentage of food and health care in the Consumer Price Index.
It turns out that Inflation may be even worse that I previously thought.
According to Mark Faber (via Bill Fleckenstein), food and health care are significantly under-weighted in the CPI. That's based on the actual consumption of food and health care by real (as opposed to theoretically modeled) people.
Faber notes that:
"In the U.S. counts food as only 8% of the CPI index. Whereas, it counts for about 10% in the United Kingdom, about 15% in the rest of Europe and more than 18% in Japan."
I have yet to validate it that percentage, but if it turns out to be true, we have Inflation (as reported) on top of the misleading hedonic/substitution adjustments (ex-inflation) on a disproportionate spending pie (ex-inflation X 2). UPDATE: Food at home is 7.66% of the CPI according to the BLS (Thanks, Mike)
Hence, doubly understated Inflation. Unless you go Core, which is the original Inflation Ex-Inflation.
Now here's where things get really interesting. Fleck points out that if you consider the proportion of U.S. household spending on food by income quintile, all but the top 20% of earners spend at least 20% of their paychecks on food.
Hence, what the weighting versus the reality are very different. This means that for 80% of the country's populace, the CPI weightings are dramatically understates what the average American is experiencing in terms of their inflation versus the official CPI measure.
This goes a long way to explaining the difference between the official inflation data and the generally poor consumer sentiment data.
CPI data is Wednesday, with a consensus of 0.3%.
>
NYT Interactive Inflation Chart
>
Previously:
Is Inflation Really Understated? (No!) May 08, 2008 http://bigpicture.typepad.com/comments/2008/05/inflation-infla.html
Inflation Abounds April 29, 2008
http://bigpicture.typepad.com/comments/2008/04/inflation-aboun.html
Is the Fed Causing a Global Food Crisis? April 25, 2008
http://bigpicture.typepad.com/comments/2008/04/is-the-fed-caus.html
>
Source:
Why all roads lead to inflation
Bill Fleckenstein
Contrarian Chronicles
MSN Money, 5/12/2008 12:01 AM ET
http://articles.moneycentral.msn.com/Investing/ContrarianChronicles/WhyAllRoadsLeadToInflation.aspx
Tuesday, May 13, 2008 | 06:55 AM | Permalink
| Comments (68)
| TrackBack (1)
add to de.li.cious | 
digg this! | 
add to technorati | 
email this post
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c52a953ef00e55237f64b8834
Listed below are links to weblogs that reference Inflation Ex-Inflation Ex-Inflation:
» Como la inflación roba a los pobres y da a los ricos from meneame.net
Este artículo explica muy bien y con las cifras en la mano (y gráficos interactivos) como las medidas oficiales de inflación subestiman el alza de los precios para el 80% más pobre de la población (que gastan mayor parte de sus ingresos en alimentos).... [Read More]
Tracked on May 14, 2008 1:08:06 AM
Comments
Sure as hell looks like food is 15% on the NYT chart?
~~~
BR: I am not sure if that's their take on all consumption (home and dining).
UPDATE: That 15% weighting includes restaurants also.
Posted by: Owner Earnings | May 13, 2008 7:39:29 AM
"Fleck points out that if you consider the proportion of U.S. household spending on food by income quintile, all but the top 20% of earners spend at least 20% of their paychecks on food."
What is the source for that fact? I don't find it credible.
~~~
BR: People, don't just tell me "I doubt this guy." Do a little research and disprove him!
This isn't a popularity contest, its a discussion about analysis and rational reasoning.
Sweat a little!
Posted by: jult52 | May 13, 2008 7:42:57 AM
The 20% spent on food for the 80th percentile would be reasonable if it includes eating outside the home, i.e. McDonalds, etc. This has been a growing trend in recent years (eating out) and one that is ripe for a downturn in a slow economy - which is already happening to an extent.
Posted by: austincompany | May 13, 2008 8:06:38 AM
OE,
Here is my personal example.
Take home pay, $3600/month
Monthly food budget $650/month (18% of take-home)
Seems like allot but for my family of four it works out to be....
$1.78 per person per meal
Posted by: Formerly Known As... | May 13, 2008 8:09:06 AM
The way this should work is:
The BLS compiles all the raw data and puts it into .xls, .csv. maybe a database format or two.
Then, private companies create their own CPI, unemployment rate, etc. They could brand them: 'The JP Morgan CPI,' 'The Big Picture Real Rate' 'Kudlow & Co Real American Real Rate" etc... Whichever one became most reliable, predictive (or helped with whichever partisan political agenda) could be used. So then over time, the best rates would win out.
The government could still use their rate for now for COLA etc. But after a decade or so, people might get up in arms enough to demand change to a private index (e.g. two guys at a bar: "Our company uses the 'Kudlow & Co Real American Real Rate' for our pension plan discounted cash flow analysis." "Oh you lucky guy, we've got the "Gold Bugs Hyper Rate'" .)
Posted by: VennData | May 13, 2008 8:29:06 AM
20% sounds about right.
You have to avoid considering personal experience in these things; I am willing to bet that Barry's audience is not indicative of the wealth distribution of the U.S.
The portion that amazes me is that our policymakers are constantly in disbelief about consumer sentiment and behavior as a result of some of these things. I can picture them staring at the CPI data and repeatedly saying "There is no inflation, why are they so mad about it....?"
Sad.
Posted by: Blutskralle | May 13, 2008 8:33:39 AM
Barry,
Count me as well as skeptical about that 20% level. I recall in Michael Pollan's latest book, he cited a statistic that the US spent the lowest % of income per capita than anywhere else in the industrialized world. I will go find that passage, but his argument was that industrialization of food, resulting in "meals" made up of "food products" was driving that percentage down not up. I think that 20% might be more accurate as a global number.
Posted by: r squared | May 13, 2008 8:35:26 AM
20% isn't insane. It's the bottom 80%--for a large chunk of those people they're already spending their max on food and as food costs rise and incomes stagnate, it's not hard to hit 20% involuntarily.
He might also be including food stamps in some odd way, who knows.
Also, remember it's not hard to jump from $400 a month for two to $650 (20%). That's only an additional 2 nice dinners out on the town. All it takes is a few drinks and bam... 20%. Kind of like withdrawing equity to buy a flatscreen.
Posted by: Patrick | May 13, 2008 8:35:35 AM
Of course, regardless of the percentage level, it is hard to deny there is not food inflation. I my building there is a "salad bar" run by four Korean women, fairly steady traffic, high quality. 4 weeks ago, they raised the price per pound by 10%. Last time I checked that would qualify as inflation.
Posted by: r squared | May 13, 2008 8:41:00 AM
Amanda,
Why are you quoting % of CPI? Isn't the entire point that the CPI is off and actual spending is what matters?
~~~
BR: Patrick, Amanda has been outed -- by a reader! -- as our little troll friend. Sure enough, his/her IP address shows multiple names.
Posted by: Patrick | May 13, 2008 8:48:26 AM
Price index minutiae 101: The CPI focuses on OUT-OF-POCKET expenditures, and therefore by design does not include medical expenses paid by third parties, ie insurance companies and medicaid. The PCE price index, however, focuses on consumption expenditure regardless of who pays. It therefore has a much larger weight for medical expenses.
Posted by: rhw | May 13, 2008 8:49:18 AM
The most obese nation buys less food than any other nation!
Slight-weighted geniuses work for the US government.
Posted by: Aurora Borealis | May 13, 2008 8:50:31 AM
Say it out loud to yourself 'THE BOTTOM 80%'
Since when did 80% of the population become the bottom!!!
Riddle me this Batman...20% of $87,000 gross income feeds my family of four what?
.
.
.
.
3 square meals a day at McDonalds
Posted by: Formerly Known As... | May 13, 2008 8:53:49 AM
2006 median household income = $48,201
20% of MHI = $9,640.20
That comes out to $26/day per household. That's hardly an egregious number.
Some of you folks live in la-la land....
Posted by: Mr. Obvious | May 13, 2008 9:11:11 AM
It's not clear how they are treating food prepared at home vs. restaurant and take out. The latter is far from trivial. Also, a significant piece of the grocery bill is non-food.
Posted by: steve | May 13, 2008 9:11:44 AM
Request for source of data:
Money quote in Fleckenstein's article: "Interestingly, if you look at the proportion of U.S. household spending on food, by income quintile, all but the top 20% of earners spend at least 20% of their paychecks on food. Thus the CPI weightings understate what is already an understated rate."
The fundamaental data point in the article, totally unsourced. Anybody know where it comes from?
Posted by: ndd | May 13, 2008 9:23:01 AM
The food-buying experience of someone and his/her significant other, or a family with one or two toddler-aged children, is completely different from a family of five with two rapidly growing teenagers. Anyone with a 16-yr-old teenage boy or girl busy in sports or other activities can attest to how rapidly large quantities of food can disappear from the fridge, irrespective of how frugal and sensible the purchases are.
Posted by: bluestatedon | May 13, 2008 9:24:12 AM
Amanda, right on. Inflation at the consumer level is much more about perception than academic interpretations of what is or is not valid data. Consumers drive the ability to pass on cost. As long as they continue to spend, PPI is a much more significant data point because that basically tells you what the economy is capable of producing (growing). Inflation is a widely misapplied notion - over time, clearly it is the central driver of time valuing assets, but that's not main street. I do believe that we are in the very beginning of a new supercycle backlash against home ownership, but one that will take a few years to really take hold at the retail level.
Posted by: Fred | May 13, 2008 9:35:24 AM
Food at home is 7.66% of the CPI
Food away from home is 6.173%
Combine both and you get 14.914% of spending is on food. This is the 2005-2006 average weight and the most recent data we have. As food inflation has been higher than general inflation this % will rise in the next couple of years.
As it so happens I just wrote a post on the weights in the CPI. Energy is up by 3.5%, so other items have been cut out of the family budget. Food is one of them.
Links to the CPI weights can be found here
http://pbp.typepad.com/economy/2008/05/how-do-folks-pa.html
Posted by: Michael Donnelly | May 13, 2008 9:38:03 AM
As Mr. Donnelly above pointed out, the statement that food is only 8% of the US CPI is false, as easily seen in the NYT link that Barry provided.
As for Fleck's unsourced statement about the bottom 80% spending 20% or more on food, this may well be true, but so what? The CPI -U is, after all, titled "for all urban consumers," not just the bottom 80%.
Seriously, this conspiracy-mongering about CPI distortions is getting tiresome.
Posted by: gustav | May 13, 2008 9:53:36 AM
Remember, the Fed targets the PCE deflator not the CPI, and that deflator gives a weight of 17% to health care. So where's the problem?
Posted by: PA | May 13, 2008 10:27:12 AM
Personally, anecdotally, Quicken and Excel tells me our upper 20% family (75K < income < 100k/annual) of 2 adults and a toddler spends 16% of take-home on food and dining out. We eat most meals at home, buy first from a warehouse club, 2nd from a local grocer, and our big weekly meal out is Saturday breakfast. I have no trouble believing 20% of income for smaller incomes.
Posted by: pmorrisonfl | May 13, 2008 10:33:13 AM
Just to add more anecdotal evidence to the fire... Math on Groceries and Dining [out] stats from Quicken for myself, my parents and 3 of my friends:
Me [3 kids, wife]: 14% food
Parents [0 kids at home]: 12% food
Friend 1 [single]: 20% food
Friend 2 [1 wife]: 16% food
Friend 3 [1 kid, 1 wife]: 22% food
Order of highest income is:
Friend 1
Me
Parents
Friend 2
Friend 3
I know #1 goes out for nice dinners and throws parties for which he typically pays for all the food. I'm a cheapass, but we still eat out several times a week.
All of the people on the list live in the middle of the US and make more than the median household income.
Looking at myself, I know the wife and I cut back on the fancier outtings, to help pay for the new house... but with kids getting bigger and McD's raising prices on their sandwiches about 10% in the past year... our efforts that got us from 16% to 12% now have us back at 14% because of rising costs.
Fortunately for us, that's the only inflation that's hurt us at all.. as we only fill up our vehicles every 2-3 months.
Posted by: Chad K | May 13, 2008 10:48:02 AM
Total family income=$95410 after taxes
Weekly food allowance(always used up)=$300x52 weeks=$15600
percentage of income used on food= 16.35%
this is for a family of four, 2 dogs and a cat.
My mortgage is $128o/month x 12 months= $15360
All I can say is wow...my food bill is higher then my mortgage!!
Posted by: David | May 13, 2008 10:49:26 AM
The main reason the weight on food is lower in the US than the UK and Europe is because the US CPI includes owner-occupied housing costs, which are a big chunk of the index, whereas the harmonised CPI's now used in the euro-zone and the UK completely omit housing.
Posted by: PA | May 13, 2008 11:02:04 AM
