[ExI] Middle Class Doomed?

[Anders Sandberg]

On 09/10/2013 21:37, Kelly Anderson wrote:
> On Mon, Sep 30, 2013 at 2:17 PM, Anders Sandberg <anders at aleph.se 
> <mailto:anders at aleph.se>> wrote:
>
>     On 2013-09-30 15:02, Eugen Leitl wrote:
>
>         On Mon, Sep 30, 2013 at 07:29:34AM -0600, Kelly Anderson wrote:
>
>             I believe that incomes almost always follow a power curve,
>             with a few
>
>         Depends on the country. It's not a natural law.
>
>
>     Whether it is a natural law is a good question, actually. 
>
>
> The folks at the Santa Fe Institute certainly seem to think that it 
> is. They are smart people, pursuing original lines of thinking, but it 
> makes a lot of sense to me.

That is just a disguised argument from authority. Santa Fe is great, but 
remember that they also are motivated to hope for universal laws of 
complexity. We need better arguments for it than that it is popular at 
cool places.

>     It is not just that power-law tails are found in all
>     industrialised economies, but they seem to follow robustly from a
>     lot of models too (e.g. http://arxiv.org/abs/condmat/0002374 ). 
>
>
> Do you happen to know what they mean by "distribution of wealth tends 
> to be very broadly distributed when exchanges are limited" Anders?

When the interaction between agents is more limited than "everybody 
trades with everybody" the distribution gets more lumpy and unequal.

> The question sort of comes down to this, "Is it always a good thing 
> when the exponent is adjusted so that the middle class is larger?"
>
> It seems obvious, but perhaps it isn't so obvious. I'm not entirely 
> sure, but my gut says a healthy middle class is a good thing.

OK, here is a utilitarian argument: wellbeing as a function of wealth is 
a very convex function. So the sum total wellbeing is maximized if the 
distribution of wealth is equal.

Of course, one might counter by pointing out that (1) maybe we cannot 
sum or compare individual wellbeing, (2) maybe it is not the sum that 
should be maximized, and (3) reallocation schemes might be impermissible 
for deontological reasons.

A classical leftist argument is that wealth is power, so a more equal 
distribution distributes power in society widely. The problem is that it 
is not clear how power actually scales with wealth. It could be that it 
is convex like sqrt(W) or concave like W^2. If it is convex even power 
law tails are not too bad, while concave might make even very equal 
societies look falsely egalitarian while small coalitions rule. And a 
realistic view that things are a messy combination of skill, ambition 
and wealth might imply that in different domains different forms hold.

>  What I'm seeing is that globalization, the Internet and 
> dematerialization are things that push the winner-take-all paradigm to 
> levels that it hasn't previously attained.

Yes. This is true. It also reaches the limit: it is not possible to be 
more global than totally global. Once we have good translation everybody 
will be in the same big domain.


-- 
Anders Sandberg,
Future of Humanity Institute
Oxford Martin School
Faculty of Philosophy
Oxford University

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20131010/fc3d4ca8/attachment.html>