Subject: Re: Alec Nove, computers and socialism 
Date: Thu, 01 Dec 94 09:25:32 +0000
From: wpc-AT-cs.strath.ac.uk


I am in full agreement with Louis on the importance
of computers to socialism. This may be influenced by the 
fact that despite being originally trained as an economist
I have for the last 20 years worked in computing,
particularly on database systems. When Noves book originally
came out I was convinced that his ignorance of computing
made his arguments quite outdated and along with Allin
Cottrell we decided to write a book that was a direct
counter to his. We wanted to show in detail how computer
systems would allow the calculations that he held to
be impossibly complex and which he thought required the
market to be handled simply and rapidly. Our reply to
Nove was eventually published by Spokesman as Towards a
New Socialism, it is also available on the world wide
web in ftp://reports-ftp.cs.strath.ac.uk/papers.

One of the things that is very striking about Nove and
the other advocates of market socialism is their inability
to put precise figures on just how complex a particular
computational problem is. Algorithmic complexity analysis
has long been a standard tool of computer science and when
applied to economic problems it can give one a complexity
measure for how hard a problem is to solve. When one applies
it to the data that Nove himself gives for the complexity
of the Soviet economy one finds that his estimates for the
time to compute a balanced plan are wildly out. He claims
that it would take millions of years to solve the equations
to fully balance the plan for  an economy of 10 million 
products - his estimate for the Soviet economy. Complexity
analysis indicates that a compute time of few minutes 
would be closer to the mark.

My impression is that most economists are stuck in
a 1930s time warp where economic calculations are done
by clerks with pens and India ink.



     ------------------