[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[wg-d] More Robert's Rules: Streamlining Voting
Those of you who took the time to read my article on motions may be
wondering about voting. It was mentioned, but never discussed in the
previous article. I'll attempt to address it here.
For our purposes, there are two opposing forces pulling at either
end of the voting issue; Fairness and speed. I think we can all
agree that we'd like everything to proceed fairly for everyone
concerned. However, I think we'd also agree that we need a streamlined
process that doesn't grind to a halt every time there's a decision to
For the past two weeks, we've discussed consensus in and outside of
the IETF process, and we've batted Robert's Rules about a bit. I
think the reason these two have come up (besides personal prejudices
and experience) are that each addresses one of these two concerns:
Consensus procedures sacrifice fairness and accuracy for speed, whereas
Formal voting procedures such as those offered by Robert's Rules of
Order give accuracy in return for time.
Perhaps there's a happy medium.
Let's start by considering the need for voting: As I proposed in
my discussion of Motions, there are several places where a vote would
need to be called (I haven't gone into too much detail because it's
tied up in the formality of precedence, and I'd like to rake the
percentage votes over the coals here as well). If we relied solely
on Robert's Rules, this could become a long process, even after efforts
are made to speed things up. What we need is a bit of consensus process.
So, something occurs that necessitates a vote according to Robert's
Rules. Under Robert's Rules, the chair would then proceed to call
some form of vote (verbal, show of hands, secret ballot, etc.). What
I propose instead is this:
1) After a reasonable amount of time has passed since the event
requiring a vote has occurred, and no one has objected, the Chair
may claim consensus and move on.
 To be determined by us.
2) If someone does object before or within a reasonable amount of
time after the chair claims consensus, the Chair must look for
support of this position within a reasonable amount of time.
3) If support for the objection is found in the form of X number
of people agreeing with the objection (this could be considered
"seconding", but it would be by more than one person), then the
Chair calls the issue to a vote.
The easiest, simplest, most open form of voting I can come up with
* When the chair calls a vote, the chair announces exactly what's being
voted on, and posts a list of all current members of the group. (by
"current list", I mean a list obtained at the time of calling the vote,
not some list saved for this purpose).
* Everyone on the list is allowed, but not required, to vote yea or nay,
or any other means of indicating assent or dissent within reason and
common decency. :) The vote ends at a predetermined time announced
by the Chair when the vote is called.
Ok, there are some problems with this. For example: This takes too
long. Perhaps, perhaps not. Here's something that will speed things up.
The chair posts a list of all valid voters when the vote is called.
Assuming the vote is on something that requires a simple majority and
not some odd supermajority, *as soon as that majority is reached*, the
Chair may move on. The requirements of Robert's Rules have been
satisfied (simple majority vote or whatever majority was required), and
the requirements of consensus have been satisfied (more that 51%, less
than 100%). There's really no need for business to wait for a full
Furthermore, this behooves those who wish for a speedy process to vote
Now, there's another issue: Someone can just object to every call for
consensus, thereby driving the group to a vote on each and every issue.
I've seen it done in physical meetings, and it's very annoying to
sit through when it's not a matter of utter substance.
Therefore, I propose that we limit everyone to X number of objections
to a call for consensus (X being small). If everyone is only allowed
to object to only one or two calls for consensus, hopefully they'll
refrain from wasting them on efforts to block progress, while still
having a tool to legitimately question the existence of consensus.
If something really wrong is going on, there are enough calls allocated
to each person to ensure proper votes are taken on important issues.
The earliest objection to the call for consensus (as timestamped by
the mailserver dnso.dnso.org and inserted into the header of every
message sent to the list) will be the one "counted" as the official
objection, to prevent several people from losing an objection over
the same call.
This is just something I roughed up over lunch today, but I think it
has some promise, at least as a possible alternative to "all consensus
all the time" or "let's all vote every single time".
Mark C. Langston Let your voice be heard:
Systems Admin http://www.icann.org
San Jose, CA http://www.dnso.org