MM, chain-weighting does exactly the OPPOSITE of what Epstein claims it does.
first, from the report "the perfect financial storm" (http://www.financialsense.com/series2/gathering.htm):
<<The next GDP manipulation takes place through a measure called the Hedonic Price Index. This is a statistical maneuver employed by government statisticians to measure computer output and investment. It is meant to capture the increase of computer power in terms of speed and memory. The government takes the actual increase in spending on computer investment and applies a statistical wand which changes the actual number into a higher number reflecting the hypothetical benefits of soaring computer power.
Like corporations, which keep two sets of books, one for financial reporting and another set of books for taxes, the government also keeps two different sets of books. One set is the actual dollars spent on the output of goods and services and the other set is called chained dollars, which is derived after various statistical manipulations have been applied to the actual numbers. As this table shows, actual computer spending in actual dollars went from $86.3 billion during the fourth quarter of 1998 to $114.2 billion in the second quarter of this year. This represented an increase of $28 billion in actual dollars being spent during the last six quarters. (iv)
Investment in Computers & Peripheral Equipment
(Billions of Dollars)
year/qu actual dollars chained dollars
1998
4th quarter 86,3 171,3
1999
1st qu. 88,1 186,1
2nd qu. 92,8 208,5
3rd qu. 97,6 230,9
4th qu. 98,9 243,9
2000
1st qu. 104,3 264,1
2nd qu. 114,2 298,5
Source: Department of Commerce: Survey of Current Business
However, after applying the hedonic deflator, that actual number is changed into $127 billion in chained dollars for the same six quarters. This technique magnifies the actual contribution of computer investment to GDP growth. This manipulated rise in GDP growth doesn’t reflect actual increases to GDP growth. Instead, it reflects the increase in computer power that businesses are getting for their money. As the power of computers increases, so does the impact of the hedonic deflator. Effectually, this creates a statistical mirage, which magnifies modest sums of money spent in actual dollars into giant sums in chain-weighted dollars.
Element #3 Software Shenanigans
Another element of statistical manipulation greatly magnifies the economic growth contribution of the technology sector. This new contrivance happened last year when government statisticians changed accounting procedures for booking computer software. Formerly, spending on software was considered to be a business expense. This acted to reduce corporate profits since expenses are subtracted from revenues. Business expense normally doesn’t enter into GDP accounts. By changing expenditures for software from an expense to an investment, it is now added to GDP. This accomplishes two objectives. It increases GDP growth and it serves to increase corporate profits and government tax revenues since software can now be capitalized instead of expensed, thereby reducing profits.
Investment in Software
(Billions of Chained 1996 Dollars)
year/quarter software dollars
1998
4th qu. 167,3
1999
1st qu. 173,3
2nd qu. 181,1
3rd qu. 192,5
4th qu. 205,3
2000
1st qu. 215,0
2nd qu. 227,5
Source: Department of Commerce: Survey of Current Business
Software spending has been running above $200 billion per year. The combination of inflating the dollars spent on computers, and including software spending as a capital asset, has artificially inflated GDP by a sum of over $500 billion. These statistical manipulations accounted for 32% of the reported GDP growth. (v)>>
------------------snip------------------
how come computer investment inflates so much after being hedoned and chained out of recognition?
from the dept. of commerce web site: (http://www.bea.doc.gov/bea/an/1199gdp/maintext.htm )
<<BEA also prepares measures of real GDP and its components in a dollar-denominated form, designated "chained (1996) dollar estimates." For GDP and most other series, these estimates, which are presented in table 3, are computed by multiplying the 1996 current-dollar value by a corresponding quantity index number and then dividing by 100. For example, if a current-dollar GDP component equaled $100 in 1996 and if real output for this component increased 10 percent in 1997, then the chained (1996) dollar value of this component in 1997 would be $110 ($100 x 1.10). >>
---------snip----------------------
of course this little explanation only touches on the basic principle - there is a lot of devil in the details still. however, as you can see from the above, chainweighting tends to INFLATE the current dollar value of items that are DEFLATING in price, and vice versa. applying the above logic to , say, a new 1gb CPU driven PC, BLS first ASSUMES what it WOULD have cost last year had it existed then, and since all other PC related products are deflating will of course assume a HIGHER price to have been applicable a year ago, and then multiplies this year's unit index number (basically counting unit growth) with LAST year's price.
simply put, something that cost 1,000 last year with 1m. units produced would amount to $1bn. in GDP last year, and costing $800 this year, with 1,2 m. units produced would amount to real(today's) dollar output of $960m., but $1,2bn. in 'chained' dollars.
note again that this is a very simplified version of the actual calculations applied, but in principle this is what's happening. of course it is difficult to properly account for something that is characterized by the big quality improvements and rapidly falling prices that are a feature of computers and other tech hardware. but when the distortions become as large as demonstrated in the above numbers, to the point of skewing GDP growth enormously with no objective yardsticks applicable, something is seriously out of whack.
let's not forget, the error compounds over time. and while one can apply hedonic multipliers and chained dollar estimates to portions of GDP, one can not do the same with the debt that has to be supported by actual cashflows.
your bankster cares little whether your output is three times as high in the govt. statistics than it really is, he wants interest and principal paid, in REAL dollars.
hope this has been cleared up now. |
