Re: Follow the Money

From: John Conover <john@email.johncon.com>
Subject: Re: Follow the Money
Date: 21 Aug 2002 18:24:19 -0000



That is an interesting article. See, also:

    http://www.johncon.com/john/correspondence/020217114704.27107.html#example-wealth

Suppose that instead of a progressive tax system, we had a benevolent
king that goes around each and every day, collects all the wealth and
money that everyone in the society made for the day, puts it in a
bucket, and then divides it up and redistributes it equally to
everyone, (i.e., a pure socialism.)  Its a social portfolio scheme
similar to:

    http://www.johncon.com/john/correspondence/020326213456.9658.html#strategy

that would double the rate of the generation of wealth, while cutting
the magnitude of the boom-bust cycles by about a factor of 3, (by
eliminating the leptokurtosis/Pareto distribution of wealth created by
the likes of William Gates' greedy ways.)  The cost would be that
everyone is always equally wealthy, (because it has stopped the
log-normal evolution of the distribution of wealth, which creates the
leptokurtosis.)

Empirically, it seems to work, too-the efficiently run near
socialisms, (Sweden and Norway, Re: WB data, for example,) tend to
have about double the wealth rate generation-and the Soviet Union had
a higher rate of wealth generation, (inferred via GDP growth,) than
the US from the early 20's right up to the end in the mid 80's, (but
there are allegations that the USSR fudged the numbers to make them
look better than they really were.)

So, what's the catch?

The key words in the above diatribe are "benevolent", "king", and
"efficiently". If the cost of the king, (or the government's
redistribution of wealth through central planning,) is more than about
0.04% per day, (about 11% per year, on an annual basis,) then the
whole thing degenerates into negative growth, and negative wealth
generation.

However, if an efficient king is not available, probably the next best
alternative for a government to raise money for
infrastructure/projects is to exploit the log-normal/leptokurtotic
frequency distribution of wealth with a progressive tax scheme;
unconstrained, wealth distribution evolves into a log-normal frequency
distribution, (in a pure capitalism,) where a very few in a society
end up controlling most of the wealth, and will end up supporting the
society-and the inefficiency. The log-normal mechanism is quite
robust, and will tolerate relatively high tax rates, (as a percentage
of GDP,) at the high end of the demographic spectrum, (but at the
expense in growth of median wealth.)

What about a fixed/flat rate tax system? Quite probably, one would end
up with the worst of both worlds-excessive volatility and sensitivity
to inefficiency-both.

Regardless of one's economic epistemology, (perhaps through an
interpretation of the intuitive arguments of Keynes,
neo-Walrasian
general equilibrium, and, Hayek,) as
to which system is best, the point is that the social welfare function
can be an engineered solution-it depends on what one wants to do,
(providing one can decide that without tripping over Kenneth
Arrow's so-called "Impossibility Theorem".) But its interesting
that a progressive tax scheme seems to be the most versatile
compromise, (which probably accounts for why virtually every one of
the couple of hundred countries on the planet use some form of it.)

(Note: the assumptions used are that the growth in GDP is near optimal
in the long run-the square of the deviation of the daily increments of
the GDP being about equal to their average; rms = 0.02, and avg =
0.0004 were used as typical numbers-which are probably manipulatable
through fiscal policy, too-for the GDP, which is close to what the
industrialized nations have run since WWII. Those numbers are
representative for daily data for a lot of things in economics,
including GDP, industrial market characteristics, equity values, etc.)

        John

BTW, if one pays more than 11% per year in trading commissions to
one's used stock salesman, one's equity portfolio growth using the (simple
portfolio management strategy,) would be negative, too. Same
principle.

Jeff Haferman writes:
>
> Nice high-level article by Brian Hayes on Pareto distributions,
> econophysics, etc:
>
> http://americanscientist.org/Issues/Comsci02/02-09Hayes.html
>
--

John Conover, john@email.johncon.com, http://www.johncon.com/


Copyright © 2002 John Conover, john@email.johncon.com. All Rights Reserved.
Last modified: Sat Aug 24 01:41:58 PDT 2002 $Id: 020821112429.22321.html,v 1.0 2002/09/18 21:11:43 conover Exp $
Valid HTML 4.0!