WTS, sure, on the face of it, hedonic pricing has a certain logic to it. the problem is, that the criteria underlying the methodology are highly subjective. like SB has pointed out, there's a difference between the micro and macro level. essentially what the government is doing, is incorporating anecdotal micro level evidence into macro level economic data. on the macro level, measuring productivity, inflation and GDP growth should be a matter of real, objective output data. it simply makes no sense to transform $8,6 bn. in computer hardware investment growth into an incredible $148 bn. when measured using the government's methodology (the example from '98 GDP). obviously the distortion is so big as to fly in the face of common sense.
imo one of the reasons why government is employing this method is to cut down on cost-of-living-adjusted expenses. in the case of inflation data, the statistical methods employed are even more prone to distortions due to geometric weighting. in essence, perfect substitution is assumed in these figures. meaning that when the price of gasoline goes up while the price of turnips is falling, BLS assumes that you're not driving as much but eat more turnips instead. the effect of this is to understate true CPI inflation, as in reality this assumption is simply wrong.
more later,
regards,
hb |
