On Ginis and 1947-1998 US Ginis

Dear WSN/PSN people,
Recently I read a letter to the WS or PSN list with the article "The Super 
Rich Are Out of Sight" by Michael Parenti speaking about the way in which 
U.S. Census Beaureau is hiding the presence of highest incomes in 
statistics 
used to determine US Gini's Indexes. A gini equal zero means perfect 
equality, a gini of one means that one person/family received all income 
(perfect inequality), a gini equal 0.5 is a very unequal society, and a 
gini 
equal to 0.2 is a very decent society, with minor class differences.
This topic is important to sociologists with interests on distribution 
inequalities and their measurement. One person that pionered the field was 
Pareto, who showed 100 year ago that income distribution, arranged from 
richer to poorer, has an exponentially decreasing curve in societies of 
different cultures and ages. Later, Lorentz introduced his famous curves 
normally made from poor to rich people. Today those statistics are 
collected 
and treated in the form of quintils (simple or accumulated) from which Gini 
value is calculated.
For example: from Census Beaureau in 1994, the poorest 20% of population 
received 4.2% of national income, the second 20% received 9.9%, the third 
quintil got 15.7 % of total income,  the forth 20% group got 23.3%, and the 
richest 20% people of US got 46.9% of national income.

Gini Index may be calculated  in different ways, and there are notorious 
differences in the reports. For example,  the World Bank reported following 
US-1994 data:
Survey year   Gini  Low 10%    1st 20%   2d. 20%    3d. 20%      4th. 20%   
 
   Highest 20%    Highest 10%
U.S. 1994 c,d  40.1          1.5        4.8          10.5         16.0      
 
      23.5                45.2                         28.5

However, at
http://william-king.www.drexel.edu/top/prin/txt/Factors/dist5.html it 
appears a gini value of
45.6% for the same year 1994.

Also, I measured the same data applying a Pareto´s correlation with a 
Lorentz curve of the shape
Y = t^ ((1-g)/(1+g)) where Y=accumulated fraction of income, t=people's 
fraction from rich to poor, and    g=gini's index 0 <= g < = 1 and 0 <= t 
<= 
  1. I derived the formula, which is long to explain in this letter.

For the year 1994 I obtained a US-Gini = 0.535 or 53.5 %
The consequences are quite serious, because g=0.535 corresponds, according 
to my particular estimates, to an average surplus rate of  2.30, and 
average 
markup gross sells/gross costs= 1.70 for the US-economy of 1994. In simpler 
words, it means that 30% of national income is made of wages and state 
social aid (real one, not militar expenses), and 70% is made of incomes 
from 
capital gains (surplus).

Using this technique I prepared a table with 1947-1998 historical data of 
US-ginis. It shows the same historical tendence of other reports observed, 
but the absolute values of ginis are around 10-20 % higher than them and 
around 30% over World Bank reports.

You may derive your own conclussions from the graph attached. There are 
also 
other studies that consider women, Black and Hispanoamerican people there, 
and they tend to crowd the poorer quintils. I think it may be useful for 
unions, workers, sociologists, women, discriminated ethnic-american groups 
and those, including me, that are against IMF/WB/WTO policies and 
misleading 
reports..

I also ask a favor to those that may give me addresses of web places that 
deal with the official methodology to derive official ginis from quintiles.

Excuse me if this looked a very technical letter. My regards to each one 
and 
all of you. Emilio

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

000-Cartagini.xls