the cost of living

[Jim Devine <jdevine@xxxxxxxxxxxxxxx>]

Thanks, Eric Nilsson, for telling me about your excellent article in THE
REVIEW OF RADICAL POLITICAL ECONOMICS, vol. 31 no. 4 (Fall 1999). It is
quite well done, in a very clear and systematic way. It's truly Gordonian,
while if Brad ever gets to rewriting his book on the standard of living of
people in the 20th century, he should read it and learn.

However, there's a confusion between Eric's and my meaning of the phrase
"cost of living."

The usual orthodox economist simply assumes that the consumer price index
measures the "cost of living" without spending much time thinking what that
phrase means. Eric brings in Robert Pollak's distinction between the
expenditure cost of living (E Coli) and the income cost of living (I Coli).
If anything, the standard consumer price index measures the E Coli. But as
Eric so clearly shows, one can suffer a fall in real income even if both
one's nominal income and the CPI are constant, if the needed expenditures
that aren't counted as part of the CPI become more expensive. For example,
if home insurance prices rise drastically (raising the I Coli), that hurts
the consumer's ability to buy the "basket" that is used to define the CPI
and thus hurts his or her real standard of living. So Eric develops a
measure of the I Coli, which unfortunately did not show up in the appendix,
because the latter didn't get printed.

Eric also brings in the key assumption of the orthodox view that the CPI
can be used to measure "real" standards of living, i.e., the constancy of
tastes. But that's nonsense, as Dean Baker has also argued. New needs arise
regularly, changing people's tastes. As a first approximation, following
the work of Alan Krueger and Aaron Siskind, Eric argues that the CPI-U-X1
(without the Boskin "improvements") represents a pretty good estimate of
the income cost of living once changing tastes are taken into account. The
changing tastes effect and the I Coli effect cancel out the reasonable part
of the Boskin effect.

Anyway, both Eric and I have different definitions of what we mean by the
cost of living. He's emphasizing the income cost of living, where I'm
interested in something that might be called the "utility cost of living"
(U Coli). This includes not only marketed products as part of the story but
also non-marketed factors. My ideas is noted by the US Bureau of Labor
Statistics when they say: ""a more complete cost-of-living index would go
beyond [the CPI] to take into account the changes in other governmental or
environmental factors that affect consumers' well-being" (Gibson,
http://stats.bls.gov/cpi1998g.htm ). My measure brings in such things as
environmental costs: when the smog gets worse, one has to wash the clothes
more often, etc. This lowers someone's standard of living just like as with
the story explaining the I Coli.  I try to get a handle on this by using
the consumption-oriented component of the Genuine Progress Indicator as an
indicator of what people are actually getting from the economy as a whole
(both market and non-market segments). Again, my article is in the new URPE
book just published by M.E. Sharpe and will be in CHALLENGE in September.
I'd appreciate comments.

As I mentioned, my U Coli measure is not the kind of "cost of living" that
measures the performance of markets, since it includes the effects of
external costs & benefits, non-marketed goods and bads, "necessary costs"
of living, etc. Even though it says something about standards of living --
i.e., that real wages have fallen much more drastically than indicated by
the ratio of the nominal wage to the CPI  -- it's not something that Alan
Greenspan should care about. He should be concerned with something like the
CPI-U-X1, which reflects the behavior of markets.

BTW, the US system of CPIs seems really messed up. The CPI-U (for urban
consumers) isn't comparable between before 1983 and after, since it uses
different methods to measure the cost of owner-occupied housing. The
CPI-U-X1 is consistent between these two periods -- but introduces
inconsistencies as soon as the Boskin "reforms" are introduced. Then,
there's the new CPI-U-RS, which brings in the dubious Boskin changes all
the way back -- to the 1970s. Before that, you're on your own.
Jim Devine jdevine@xxxxxxx & http://bellarmine.lmu.edu/~JDevine