20131118 (ON)
Journal: November 18, 2013
     Index     
Return to:   Site   or   Journal   Description

Government (CPI)

"The inflation rate used on the chart is the government CPI number until 1993. Beginning 1994, I have added 2.7% per year to the government CPI number. This should better reflect the true inflation rate since the government number has not been accurate since around 1993. The added 2.7% corrects the geometric weighting formula used by the government to calculate CPI. The government does other adjustments that constitute another 4% reduction in the inflation figure, but I will assume those adjustments are valid. For more information on inflation, see the following articles: http://www.shadowstats.com/article/56 , http://www.safehaven.com/article-8848.htm" from http://home.earthlink.net/~intelligentbear/com-dj-infl.htm

I compiled this information from the Bureau of Labor Statistics, the organization that calculates the Consumer Price Index or CPI. There is not a single CPI, it has changed so a historical record is being lost. Changes are in the works again. The attached word document below highlights two of the more important changes:
    • Arithmetic to geometric weighting
    • Hedonistic Adjustments

The simple "basket of goods" is now a complex, value-laden basket. The change from arithmetic to geometric weighting can be expressed by the equations used to calculate the CPI.

Official Equations (PDF): The left equation is so simple. Take the summation sign away (it cancels) and it just says "the price now divided by the price then (any base period) is the inflation fraction" (divide this by the time and get a rate, multiply by 100 and get the inflation rate in percent).

The "new" equation (right) took simplicity and replaced it with nearly impossible to understand mathematical gobeldygook. Comparison of the two methods in the table below shows that, relative to the arithmetic method, the geometric method exaggerates differences both positive and negative, and several sources estimate the net difference is that the geometric method underestimates the rate of inflation of the "typical basket by 3 to 4 percentage points. This means that more often than not, the situation is probably more similar to examples 3 and 4, where most items rise somewhat the same but a few rise hardly at all or even depreciate.

The PDF document shows calculations to arrive at the table above, as well as other stuff from the Bureau of Labor Statistics web site. Supporting the idea that examples 3 and 4 are common, is the concept called "hedonistic adjustment", which was formally added to the CPI gradually from 2007 through 2009. This "adjustment" often reduces the apparent price increases for selected items because of "hedonistic value" due to product substitutions. The "distorted" or some say "improved" index by geometric weighting is thus further modified by applying this equations to selected items in the "basket";


The b's (not the B's, which are the "current", adjusted price for an item while A's are the old, base-line price) are an arbitrary weighting factor based on someone's judgment (committee vetted no doubt) of "equivalent" value "perceived" by the "typical" consumer in terms of "living standard" when buying a functionally similar but different item (i.e. item B, above, rather than item A, which is no longer available or bought). The quotes indicate BLS's words. These weighting factors, arbitrary as they are, are applied, arbitrarily, to when differences in items in the "basket" are sufficient to apply the weighing factors in the first place. As a result, only certain selected items are even considered for weighting, and these are all restricted to certain defined groups of rapidly changing consumer items such as apparel and electronic equipment.

So now the index is clearly seen to be weighted by nearly invisible factors, or people's judgments, about such things as the increased value of digital vrs analog televisions for the same screens size (BLS weighs that increased value quite high, so high in fact that it more than offsets a price increase of nearly $300; so, based on someone's judgment, a price increase of $300 is actually a decrease of $100. Orwell called this one, and named it newspeak, not doubleplusungood, which he could have.

Then there is of course the fact that only 84% of Americans are considered in the CPI; only those that live in an SMSA (standard metropolitan statistical area) or town with more than 2500 people. Farmers, your costs don't count, so your increased drive expensed to town are not included. In fact the CPI is really called the CPI-U, the "U" stands for "Urban". Then there is the CPI-W for wage and clerical workers, and the new C-CPI-U for chained CPI-U and talk of a CC-CPI_UW. I am reminded of Dante lamenting the sinful ways of politics:

How often, in the time you can remember,
    have you changed laws and coinage,
    offices and customs,
    and revised your citizens!

And if your memory has some clarity,
    then you will see yourself
    like that sick woman
    who find no rest upon her feather bed,
    but, turning, tossing,
    tries to east her pain."
            The Commedia, Purgatorio, VI:145-151

The constant "improvements" to the CPI have had one certain effect: they make historical comparisons much more difficult. Perhaps that is part of the intent, conscious or unconscious.