Sorry about the confusion. Everyone who works with the CPI, government,
academic, business, ..., is now convinced that the CPI has an upward bias of
perhaps 1%+ per year compared to a Konus (constant utility) index. There
has been a great deal of publicity about this here in DC and I, perhaps
naively, assumed that it was well known elsewhere. There is now a blue
ribbon Senate sponsored commission with such as Grilliches, Jorgensen,
Boskin, ...., to recommend improvements. Since I work here producing the
CPI it is a very big deal for me.
Curiously, during WW2 when shipyard wages were indexed by the CPI, labor was
convinced that inflation was UNDERSTATED by as much as 5-10% per year. It
has always seemed that someone (surely not me) should go back and work my
way through this old discussion. Note, however, that if it were true that
the net bias of the index were on the order of -5% the Fed would use this as
an excuse to throw us into a recession. Indeed, part of their willingness
to allow the modest economic growth we have seen lately is because they now
believe, based on the upward bias argument, that inflation in the Konus
sense is very close to 0. Thus the bias argument cuts both ways.
Oh well, since I have gotten into it this much, perhaps I can look up the
WW2 material. Perhaps in the next week.
Dave Richardson
----------
From: pen-l
Subject: [PEN-L:3895] Re: FW: Daily Report
Date: Monday, April 22, 1996 9:11AM
RE:
> The last item (on NAM) states that if the overestimation of the CPI is
taken
> into account, real wages are up since the 70's. Although I don't
usually
> publicize this, it is correct, and word seems to be getting out.
It would take a lot of work to convince me of this.
According to my calculations, real hourly post-tax wages fell about
15% between 1972 and 1993 (using existing CPI). Real hourly post-tax
compensation (taking account of benefits) is down 10% since 1978
(using existing CPI).
Real hourly compensation (1977 dollars)
1977 1993 % change
My estimates $ 5.57 $ 5.01 -10%
In order to turn this 10% drop into a 15% gain (as claimed by
NAM), the 1993 must be about $6.40. That is, is it really
being argued that a "better" CPI alone is responsible for an
increase in real compensation on an hourly basis equal to $1.39 by
1993? This is a 28% increase in 1993 compensation (relative
to using existing CPI). Perhaps this is possible, but if the BLS
measures of CPI are so bad, why hasn't this been noted until
relatively recently?
The LA Times had an article on this NAM report. They quote
the report as saying (more or less) that the authors knew that they
were picking the most favorable assumptions possible in order
to generate their results. Are we to believe that the NAM economists
have scooped the entire economics profession on this particular
point? Hard to believe.
Eric
..
Eric Nilsson
Department of Economics
California State University
San Bernardino, CA 92407
[EMAIL PROTECTED]
