On Nov 23, 2007 8:33 PM, Jim Devine <[EMAIL PROTECTED]> wrote: > The CPI is supposed to be a measure of the the consumer's "cost of > living." It would thus be wrong to bring in asset-price inflation as > part of the CPI (or the Personal Consumption spending deflator). The > CPI was originally designed (by labor, not by government, BTW) as an > effort to measure the cost of living. The idea was that we should know > when wages aren't keeping up with the prices that workers have to pay. > Workers typically can't afford to buy stocks and bonds, so their > prices shouldn't be counted as part of the CPI. Those paper products > are hardly necessary to human life. >
Fair enough. I have no quarrel with the CPI as a cost-of-living measure (even though I am sure it has some weaknesses there as well). I am mainly disputing the reasoning behind using the CPI to set monetary policy. This reasoning as I understand it, is basically that a high CPI indicates a currency that is losing its value and reducing the purchasing power of people on fixed income, and also people incorporating expectations of a currency losing its value into their decision-making. But does the same logic not apply also for asset prices? If the price of a house increases 25% in a year, does that not mean a dollar buys (roughly) 25% less house compared to last year? Shouldn't the Fed try to preserve the value of a dollar in house-units just as much as in sandwich-units? Also a high CPI does not necessarily decrease purchasing power if wages increase at the same rate. Wage stickiness only holds over short time intervals of months and if the cost of living increases modestly say 5% or so, it is not clear that stickiness over such short times is much of an issue. -raghu.
