Re: [ietf-nomcom] Changing the candidate selection model

[Eric Rosen <erosen at cisco.com>]

The context for most of the nomcom discussions is the assumption that nomcom
is  some sort  of blue  ribbon commission,  whose task  is to  determine the
objective facts and then to make objective decisions based on those facts.

I  think that  assumption  is just  a  myth.  Choosing  the  IESG/IAB is  an
essentially political process, not a fact-finding exercise.  If we are going
to talk about changing the procedures, we should use our knowledge about how
political processes  actually work  in practice.  I  think it shows  way too
much hubris  to assume that the  IETF is somehow  going to set up  a process
that produces some  objectively superior result, even if  such a thing could
be assumed to exist.  Political processes tend to work in a certain way, and
the way they work is largely independent of the intentions of the people who
drew up the rules.

> Of course, if a Nomcom is lazy or indifferent enough that it
> simply skips the evaluation process and returns all of the
> incumbents, then we are in big trouble.  Should a Nomcom decide
> to behave that way, I believe that we can have exactly the same
> failure case with the current setup so the proposal doesn't make
> things any worse.

I  certainly disagree;  in politics  there's a  maxim that  "you  can't beat
somebody with nobody".   The chances of ousting an  incumbent are much lower
if it is not known who will replace him/her.  

There are  some exceptions to this  maxim; a highly  factionalized group may
agree  to  oust an  incumbent  without  having any  hope  of  agreeing on  a
replacement.   While you  seem to  think that  this is  a valid  outcome, it
usually  doesn't work out  very well  in practice.   (See, e.g.,  the Weimar
Republic.)


> The first thing the Nomcom has to do is to perform a careful and
> thoughtful, area by area and slot by slot performance review
> without the distractions of non-incumbent nominees.

These process details don't matter.   Unless the nomcom happens by chance to
be dominated by  political opponents of the incumbent,  the tendency will be
to keep the incumbent unless there is consensus for a replacement.

> I've got similar reactions to your comment (I'm tempted to say
> "innuendo") about the liaisons.   While I think they are
> necessary, I've always been concerned about the risk of
> unreasonable influences from the bodies to which people are
> going to be appointed

Some people have  strong convictions and are immune  to such influences, but
many  people  are  very  susceptible  to influence  from  the  "leadership",
especially if  they know that the  leadership is watching  their every move.
So I think the liaisons really  add a corrupt influence to the process.  The
nomcom should be able to call on representatives of the leadership when they
require information,  but it's really  outrageous to have liaisons  sit with
them on a regular basis.