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

[John C Klensin <john-ietf at jck.com>]

--On Friday, June 19, 2009 00:49 -0400 Barry Leiba
<barryleiba at computer.org> wrote:

>...
> I suggest that the other cases are either neutral or show the
> new proposal as better than what we have.  So let's consider
> how likely the failure mode is to happen, and the relative
> effect a failure has on the IETF.
> 
> I can't speak to the likelihood, but I can suggest that with
> the current system, it's not just the NomCom that would
> happily reappoint the incumbent.  The community as a whole
> probably knows that she's doing well and will likely be
> reappointed, and so history shows that there are likely to be
> few who volunteer for her spot.  That means that with the
> current system, we probably won't have much of a choice
> anyway, and so the superstar replacement probably won't be
> available anyway.
> 
> In other words, it probably turns into a failure mode for both
> systems, old as well as new.
> 
> As to the effect the failure has: By definition, the IETF has
> not gotten the optimal leader in that position.  But the IETF
> *does* have an effective leader whom the NomCom is happy with.

> It strikes me that that's not so bad -- not a very serious
> failure.

Right.  And, fwiw, if the "ask your management for permission to
stand as a nominee" scenario I outlined in an earlier note
applies and the superstar replacement is persuaded to be a
Nominee in year N but the incumbent is chosen anyway, the odds
of the superstar being available in year N+1 (when there might
be a weaker incumbent or someone ready to step down) go down at
least slightly.  That is one of the scenarios that leads me to
believe that "incumbents first" would sometimes improve the
quality of the nominee pool when it was important to have a
good-quality pool, i.e., when either there is a vacancy
(incumbent not willing to serve again) or a problem (incumbent
fairly clearly in need of replacement).


--On Friday, June 19, 2009 09:22 -0400 "Joel M. Halpern"
<jmh at joelhalpern.com> wrote:

> Returning to John's proposal, as I said, my concern is not
> with the case of a strong AD and a hypothetical superstar new
> nominee.  My concern is on the other end.  And I suspect
> (although it is always hard to tell, even in face-to-face
> discussions much less email), that John's and my basic
> disagreement comes much more when there is no superstar, or
> even outstanding candidate.

Actually, I think our most basic disagreement lies somewhere
else (see below).
 
> THe basic problem is when it is not clear cut whether the
> nomcom should re-appoint a given sitting individual.  Assume
> that the person has served two terms.  (The single term cases
> confuse the issue.)  We do not, and should not in my opinion,
> have a rule that says "out after two."  So the nomcom has to
> evaluate the person.  They have to evaluate how well that
> person is doing their job.  How well that person will fit with
> the partially known set of other people both in terms of style
> and in terms of skills.  (I have been told of cases where the
> nomcom litterally had to chose between people because two
> people could not serve together.  Also, sometimes nomcoms will
> use other appointments to compensate for weaknesses in one
> appointment.)
> And I consider it unacceptable to just leave the seat unfilled
> if we have someone who can probably do a decent job.  But I
> would hate to have to do an initial round appointing someone
> who is only expected to probably do a decent job.

Also ignoring the single term case, I think that the empirical
experience -- in the community, not necessarily the Nomcoms--
has been that, when an AD or other incumbent has served for two
or more terms and can only be expected to do a "decent" job, it
is time to get him or her out because, statistically and
observationally, the behavior and quality of job performance
always get worse.   That may not be true for incumbent
superstars, but they aren't at issue in this example.

This case is where we have a fundamental difference of opinion.
I think the IETF is put at risk when a Nomcom has few enough
qualified nominees that is has to start making decisions on the
basis of "this particular nominee is only the best of a bad lot,
but is he or she better than nothing".    Obviously, sometimes
those appointments work out, but that isn't the issue.

I believe that, if the Nomcom doesn't have (and can't find) an
adequate number of Nominees who are both willing to serve and
highly qualified, it is actually very important to take that as
a symptom of "not enough depth in this particular area" and
force the relevant body and the community to take that as a fact
and deal with it.  I am willing to leave the definitions of
"adequate" and "highly qualified" to the Nomcom, but I don't
want to see any appointments made for which the Nomcom perceives
that they had to scrape the bottom of the proverbial barrel.

> Sometimes, if the nomcom beats the bushes a bit, good nominees
> surface.   Sometimes, when the nomcom beats the bushes, they
> just get scratches on their hands.  Failing to appoint an AD
> because the procedure says "you can not go back to the decent
> but not great candidate you passed over because he is an
> incumbent" seems like a bad box to put yourself in.


I don't want to put words into your mouth here, but it sounds to
me that you are willing to go quite a bit further in the
direction of a probably-unsatisfactory candidate in order to
make sure that a slot gets filled, while I would prefer to
empower the Nomcom to say "we have examined all of the
possibilities, beaten the bushes, and can't find a Nominee we
are really happy with -- so the community needs to make another
plan.

Put differently, I can see lots of reasons to either retain an
incumbent or to drop one.  But I don't see "after performing an
evaluation and weighing the issues, we'd really prefer to retire
this person, but are willing to live with another term if we
can't find anyone else or if all the people we can find are
worse" as a acceptable reason for retention.  That is especially
so because I believe I've seen situations in which a Nomcom has
returned someone who had burned out because they couldn't find
anyone better and that the result of that decision has usually
been unfortunate.  While I can accept that there possible
situations very near that boundary that you would consider
failure cases, I consider any "no choice" situation to be a
sufficient failure that there is no win either way.

   john