[ show plain text ]
Re Patrick's [OPE-L:2861]:
> When the revolution comes, you will not be the statistics czar!:)::) I
I want no part of any "revolution" that seeks to bring new Czars to
power! Down with all Czars -- including the "statistics czar"! :-)
> Your exercise A don't provide much meaningful information.
Perhaps, but your prior statistic on median family income in the US didn't
provide much meaningful information either since it could not express
> The median, combined with information on the dispersion and mobility would
> be much more useful and accurate.
How do we calculate dispersion? By percentiles (e.g. 5ths) as the BLS
does? Wouldn't it be much better to calculate dispersion by *class*
and also show dispersions within groups to see group differentiation?
(the later was my intent behind Exercise B).
I think that e.g. statistics for income *and wealth* for the working class
(approximately 80-820f the US population, right?), the top 1%, and the
group in-between (the so-called "middle class") would indeed be meaningful
statistics, don't you? I.e. statistics on income and wealth *by class*
would indeed be meaningful. Moreover, it would be revealing to observe
patterns of differentiation in terms of income and wealth *within*
classes (btw, a subject that was of great interest to Lenin and many
Russian Marxists who studied peasant differentiation). Similarly, as I
suggested previously, it would be meaningful to observe income and wealth
differentials *within* racial and ethnic groups.
I should also note that statistics on the *distribution of wealth* (as
distinct from the distribution of income) are extremely difficult to come
by -- yet such statistics would indeed have great meaning.
While it is true that we have to work with what we've got, we need to
understand the limitations of official government statistics.
In solidarity, Jerry
This archive was generated by hypermail 2b29 : Sun Apr 30 2000 - 19:59:44 EDT