On Tuesday, in article <2008021918272211272-pbrazee@[EMAIL PROTECTED]
>
pbrazee@[EMAIL PROTECTED]
"Howard Brazee" wrote:
> On 2008-02-19 13:12:19 -0700, David Friedman
> <ddfr@[EMAIL PROTECTED]
> said:
>
> > I don't know where your "conventional wisdom" is coming from, but the
> > U.S. Statistical Abstract shows per capita money income in constant
> > (i.e. inflation adjusted) dollars roughly doubling from 1967 to 2005.
>
> Part of that is because we have lots more rich people today. But
> having the average go up because other people got rich leads towards
> dissatisfaction and often revolution.
>
> A better check on "conventional wisdom" is to see how much things got
> better (or worse) for the average Joe.
There are different sorts of average, with different meanings, and most
people think of the arithmetic mean, which is the case here, and which
has that weakness. Consider the per capita income of a slave plantation
as an extreme example.
Two other sorts of average are the mode and the median, and I sometimes
get confused by which is which.
One simply ranks samples in order, low to high, and takes the one in the
middle. What that means depends a little on how things are measured and
recorded. You might, for instance, split incomes by $1000 dollar steps:
think of all those marketing surveys. So it might turn up the paid staff
on the plantation, rather than the slaves or the owner. And, if you work
on individuals, how do you rank more than one person with the same
income.
The other, which I think requires that sort of lumping together of the
measurements, is simply the biggest group. Which in the plantation
example would be the slaves. In the modern USA, it might be workers on
minimum wage, or it might be a little above that (and people who are
working through temp agencies, with lots of little gaps in their
employment, are going to be lower down the range than their hourly rate
suggests).
I think I've got all that right enough. Now, what's im****tant is that
these different sorts of average measure different things. And which you
use, when you want to say how things have changed, can say a lot about
what you're trying to do. For example, with the plantation, ending
slavery might lead to the mean income dropping, or it might increase.
But the other two measures will give different results.
Most obviously, the income of the largest group, the former slaves,
increases. What happens to the middle group? I suspect that, since the
slaves are no longer all at zero, and some will earn more than others,
the middle group drops down a few steps, but it depends on how the
groups are defined. And the new middle group won't be the same people.
And those changes will happen without any necessity for a change in the
mean income.
So, pick the right average, and you have "correct" evidence for slavery
being good or bad for average incomes.
--
David G. Bell -- SF Fan, Filker, and Punslinger.
On the horizon, a carrier task force of the Salvation Navy was
turning into the wind, preparing to launch Zeppelins.
|
