(nearly) scale free second hand spam

Date view Thread view Subject view Author view

From: Dave Long (dl@silcom.com)
Date: Mon Oct 23 2000 - 11:33:32 PDT

Maybe there's a social diamond, but in pure wealth
terms, the pyramid is alive and well. I added the
billionaires from the 1995 Forbes 400 and some data
from the 1995 census on "top wealth" (>$600K gross
assets) holders to the 1993 numbers I gave earlier.

Plotted out on a log-log, the distribution looks
more like 1/f between 0 and $100K, and very close
to 1/f**2 between $1M and $10B.

1/f**2 makes a certain amount of sense, if one
views fortunes of that size as being influenced
more by market fluctuation than by lifetime wages.
I think integrating white noise gives brown noise;
whether one can take temporal distributions and
apply them to spatial (personal) ones, I know not.

I'm not sure how to interpret the 1/f part of the
curve. Some possibilities:

- there's an underlying thermal distribution.
  If true, it would be interesting to compare
  current "temperature" with that from other
  historical data sets, but the large 1/f**2
  region makes the overall distribution look
  more scale free than thermal. (Is it much
  of a limitation that the data set is only a
  few hundred million individuals? Would we
  see the curve steepen much more going to
  the right if we added the more infrequent
  events, such as Rockefeller's 2% of GDP?)

- wages are relatively important for those
  without much wealth, which pinkens the
  spectrum in that area.

- (circular reasoning?) simply an artifact of
  avoiding infrared catastrophe: if 14% of the
  population has a net worth under $5K, but
  only 12% has a negative net worth, then the
  distribution has to pass through a maximum,
  which will show up as white noise, and there
  would be a 1/f area on the way to the 1/f**2.
  (but why the maximum, other than common sense
  that we can't have an infinity of debtors?)

Detail that would help: (anyone have sources?)
  - distribution of negative net worths
  - better resolution for the areas around 0
    and half a million


(Scale free is the best distribution of wealth
for tyranny: if economic networks behave like
random ones, one with scale free characteristics
is robust in the sense that one can appropriate
wealth from random individuals without having
much effect on the overall network. Better to
be princeps of an oligarchy than a demagogue,
however, as relative scarcity of elites is what
determines the robustness)

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Mon Oct 23 2000 - 11:21:05 PDT