[ QUOTE ]
first off, know that we're dealing in averages here. if you are an average american w/ an income tied exactly to average CPI then the CPI applies to you. but the average is a poor descriptor of the individual. for some, it will mean that they are living better off and for others it will mean that they are living worse off as you point out. in some cases the differences will be quite extreme.
what you say makes sense if the bold part above is true.
can you cite where the CPI assumes that a rise in food prices isn't passed entirely through to the index?
if average milk prices increase 10% (and nothing else changes) and households overall report that as such, then the relative importance of milk will rise by 10% and food will increase by 10%*milk's contribution to food if my understanding of the CPI construction is sound.
cna you provide some cite to the contrary?
thanks,
Barron
[/ QUOTE ]
did you google hedonics?
http://www.atimes.com/atimes/Global_.../HF27Dj01.html
[ QUOTE ]
However, in the early 1990s that transparent method of CPI calculation came under assault from Michael Boskin, then chief economist to the administration of US president George H W Bush, and Alan Greenspan, then chairman of the board of governors of the Federal Reserve System.
Their assertion was that the CPI calculation was much too simplistic and resulted in much too high a measure of inflation. They argued that the calculation needed to take into account the real-world phenomenon of substitution, whereby consumers who cannot afford more expensive items switch to buying the less expensive ones, and that the inflation calculation should switch to tracking the costs of the less expensive items whenever substitution was presumed to occur.
That generally would have required moving from the fixed basket of goods to a variable one, but initially, instead of that move the concept of weighting was introduced into the fixed basket in an effort to approximate the phenomenon of substitution.
Straight arithmetic weighting was gradually replaced by geometric weighting by 1999. In the geometric weighting favored by Boskin and Greenspan, the basket items with recently increasing ("volatile") prices receive less weight while those that decrease in price receive more weight. So the result is a lower overall number for inflation.
[/ QUOTE ]
http://en.wikipedia.org/wiki/Hedonic_regression
[ QUOTE ]
Some economists have criticized the US government's use of hedonic regression in computing its CPI, fearing it can be used to mask the "true" inflation rate and thus lower the interest it must pay on Treasury Inflation-Protected Securities (TIPS) and Social Security cost of living adjustments.
[/ QUOTE ]
http://moneycentral.msn.com/content/P92951.asp
[ QUOTE ]
Lastly, Gross points out who is penalized by this cheating: "They disserve, of course, all of those who receive Social Security, as well as other private pensioners dependent on an accurate accounting of prices paid. They disserve buyers and holders of TIPS -- inflation protected securities -- which adjust inadequately to a faulty and near fraudulently calculated CPI that one day could total billions of dollars per year for TIPS holders. And they disserve all owners of U.S. Treasury obligations -- including foreign central banks and institutions. . . ."
[/ QUOTE ]