[PEN-L:6263] Compounding folly: the Kelvinator fetish

[knowware@xxxxxxxx (Tom Walker)]

1. >The general picture over the past ten
>to fifteen years has been one of failure
>to accomplish the growth-rates of
>the post-WWII period.

2. >BOOMING CONSUMPTION AGAIN LEADS GROWTH

In both cases, the numbers rely on the book keeping fiction that for one
particular "commodity" n = -n. Doug is not going to like my questioning of
bourgeois statistics. But the basic error explains a lot. Let's call it the
Kelvinator fetish, firstly after Lord Kelvin -- whose dictum that science is
measurement graced the title page of Econometrica for the first 20 years of
its existence -- and secondly after Marx's analysis of the fetishism of
commodities.

The commodity in question is 'leisure', or more precisely the free time of
employed labour force participants. The error can best be shown with a
simple example: let's assume that maximum total output can be achieved over
the long run with a standard length of working day of, say, eight hours.
Given a standard day of eight hours, output in any one day can be increased
by working longer hours but such a short term increase will be more than
offset by losses in output on subsequent days.

Competitive pressures on employers and the short-sightedness of some workers
may tend to lengthen the working day beyond its hypothetical optimum for
output. That is to say, that normally hours may be somewhat longer and
output somewhat less than they could otherwise be. However, this
less-than-maximum output could have as easily been achieved with a working
day shorter than the optimal day. For simplicity, let's assume that output
for a day that is one hour "too long" is the same as that for a day that is
half an hour "too short".

Using the example of an optimal eight hour day, then, output during seven
and a half hours would be equal to output during nine hours. Out of the 24
hours in a day, that leaves sixteen and a half hours of free time if the
worker works a too-short day and fifteen hours of free time if the worker
works a too-long day. If workers are paid in proportion to the value of
their total output, the daily wage for the two situations should be the
same. The hourly wage would be correspondingly higher for the shorter day
and lower for the longer day.

In the real world, it is difficult to know exactly what the optimal length
of the working day is, so it is difficult to say whether a day of any given
length is too long or too short, from the perspective of maximizing output.
For the sake of "convenience", the political economy of growth assumes that
the given work day -- whatever it is -- is optimal.

The convenient assumption incorporates the anomaly that, considered as a
commodity, fifteen hours of leisure is "worth the same" as sixteen and a
half hours. In the same vein, fourteen hours might be worth the same as
seventeen and thirteen might be worth the same as eighteen. One could go on
to the end of the day adding to one side and subtracting from the other and
growth economics would take no notice. Workers are assumed to be indifferent
between a lower wage with less free time and a higher wage with more free time.

The grossness and simplicity of this fundamental conceptual error at the
base of post-WWII economic thought is so extreme that people cannot accept
that such a flaw could exist. Oh sure, we'll all admit the world is going to
hell in a handbasket. But to suggest that there is a simple, obvious
programming error driving the thing is unthinkable. Even more unthinkable is
the idea that the introduction of the programming error can be precisely
documented.

Growth? Relative to what?

regards,

Tom Walker
http://www.vcn.bc.ca/timework/covenant.htm