>At 4:13 PM -0500 3/5/00, Doug Henwood wrote:
>>http://nypostonline.com/business/25545.htm
>>
>>Doug, are things really *this* bad at the BLS? This Crudele guy tends to
>>exaggerate;is it true that without those "ex-anything that goes up"
>>adjustments inflation would have been an annualized 12% last month?
>
>Sorry to take so long to respond to this, but I'm a bit behind.
>
>Crudele strikes me as extremely unreliable. Here's what he said in
>that column:
>
>>Buried deep in a footnote in the CPI report, under a section about
>>seasonal adjustments, is this statement:
>>
>>"Effective with the calculation of the seasonal factors for 1990,
>>the BLS has used an enhanced seasonal adjustment procedure called
>>Intervention Analysis Seasonal Adjustment for the CPI series ...?.
>>For the fuel oil and the motor fuels indexes, this procedure was
>>used (in January) to offset the effects that extreme price
>>volatility would otherwise have had on the estimates of seasonal
>>adjusted data for those series."
>>
>>If this footnote really means what it seems to say, this disclosure
>>is astounding. Washington has been assuring us that that inflation
>>is under control, but it has been reducing -- and maybe even
>>eliminating -- the impact of the rising price of oil in its
>>calculations.
>>
>>That's like leaving the rain out of a weather report. Or the
>>visitor's runs out of a box score.
>>
>>If this is what BLS is doing, it would produce a meaningless number.
>
>Here's the BLS's full note, from the release
><ftp://146.142.4.23/pub/news.release/cpi.txt>:
>
>>Effective with the calculation of the seasonal factors for 1990,
>>the Bureau of Labor Statistics has used an enhanced seasonal
>>adjustment procedure called Intervention Analysis Seasonal
>>Adjustment for some CPI series. Intervention Analysis Seasonal
>>Adjustment allows for better estimates of seasonally adjusted data.
>>Extreme values and/or sharp movements which might distort the
>>seasonal pattern are estimated and removed from the data prior to
>>calculation of seasonal factors. Beginning with the calculation of
>>seasonal factors for 1996, X-12-ARIMA software was used for
>>Intervention Analysis Seasonal Adjustment.
>>
>>For the fuel oil and the motor fuels indexes, this procedure was
>>used to offset the effects that extreme price volatility would
>>otherwise have had on the estimates of seasonally adjusted data for
>>those series. For the breakfast cereal index, the procedure was
>>used to offset the effects of price-cutting among cereal
>>manufacturers. For the educational books and supplies index, the
>>procedure was used to account for greater than normal sale prices
>>on educational reference books. For some alcoholic beverage series,
>>Intervention Analysis Seasonal Adjustment was used to offset the
>>effects of increased brewer's costs along with increased demand for
>>specialty beers. For the nonalcoholic beverages index, the
>>procedure was used to offset the effects of a large increase in
>>coffee prices due to adverse weather. For the fats and oils series,
>>the procedure was used to account for lower domestic butter stocks,
>>lower cold storage supplies, and anticipation of a bumper soybean
>>crop. For the new trucks index, the procedure was applied to
>>account for loyalty rebates offered to customers by American
>>automakers. For the water and sewerage maintenance index, the
>>procedure was used to account for a data collection anomaly.
>
>I have a query into the BLS contact on this, but I read this
>statement to mean that extraordinary movements are stripped out
>before calculating seasonal adjustment factors, since those
>extraordinary movements are nonseasonal. So the technique does
>nothing to disguise sharp price movements - it only prevents them
>from being incorporated into the SA factors and so carried into
>subsequent years.
>
>As it is, SA energy prices were up 1% in January - a 12.7% annual
>rate; that's hardly a small number.
The BLS confirmed my interpretation. And Crudele's claim that the note was buried deep was horseshit; in the electronic version, the footnote appears about 20% of the way into the document, between the end of the introductory text and table 1. That doesn't strike me as "buried deep." It seems like exactly where you'd put such a note.
Doug