From: Sandy Sandfort <sandfort@crl.com>
Date: Mon, 16 Oct 95 07:56:30 PDT
To: Wei Dai <weidai@eskimo.com>
Subject: Re: transaction costs in anonymous markets
In-Reply-To: <Pine.SUN.3.91.951016011355.2241A-100000@eskimo.com>
Message-ID: <Pine.SUN.3.91.951016074023.1719B-100000@crl.crl.com>
MIME-Version: 1.0
Content-Type: text/plain


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                          SANDY SANDFORT
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

C'punks,

On Mon, 16 Oct 1995, Wei Dai wrote:

. . .
> Two main components of transaction cost in an anonymous market will be 
> the cost to maintain anonymity and the cost to evaluate reputations.
. . .
> The second big part of transaction cost in an anonymous market is 
> reputation evaluation.  Of course, normal, everyday transactions require 
> reputations to be evaluated.  However, more effort and cost will be 
> expanded on these evaluations in an anonymous market because the effects 
> of misevaluations will be much more damaging.  Reputations must be 
> constantly reevaluated, as pseodonyms are easily transfered.  Since no 
> good theories of reputation currently exist, these evaluations are 
> difficult to automate.  Perhaps theoretical advances can make these 
> evaluations easier and/or more accurate.  However there does not appear 
> to be any major research effort in this area.

This is only true if you assume everyone will do their own
reputation evaluation.  This is a third-party insurer problem,
really.  Either a "Lloyds of London" or a "Good Housekeeping" 
model would do the trick.


 S a n d y

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~