Do-Over – How it might still be done.

The successive failures we’ve endured have not been the result of a lack of options, but the result of the MDP’s inability to discriminate between workable and unworkable plans.  Kicking the ball over to the Legislature was only the most recent foolish move.

What can be done?  First, we need to make a hard-headed assessment of the situation, and reject every alternative that requires broad concensus or which depends on obtaining approval from interests which have no reason to grant it.

In short, the MDP must be left out, and neither campaign can be given unilateral veto.  It can’t involve the Legislature, it can’t be subject to scrutiny by the courts, and it can’t depend on raising large amounts of money.

All we can hope for is an objectively fair process, that reflects the attitude of Michigan Democrats.  By “objectively fair”, I mean it would ignore the special pleadings of each campaign and their respective sympathizers, who are prepared to proclaim any principle that seems likely to gain even one delegate for their side.  The standard should be that it falls within the range that reasonable people would agree seems more-or-less fair.  And it should cause Michigan’s delegation to more or less reflect the will of her voters.
So what am I proposing?

In Florida, screening is comparatively simple:  restrict the sample to registered Democrats.  In Michigan, it’s going to be necessary to define a series of questions to winnow out the correct group.  Maybe they’d exclude people (like me) who voted Republican in January.  Maybe they’d initially ask people if they consider themselves Democrats.  The screening wouldn’t be perfect, but perfection isn’t a fair standard at this point.  The fact is that such screening questions are employed all the time in the polling industry, with good results.  Because there would be adequate lead time, I would have the polling firm conduct a pilot study first, so they could test alternatives and recommend the one they think worked best for the real survey.

There are lots of objections, but it’s so late the only alternative is to watch the train jump the track.  Here are the ones I can think of:

• Who would conduct the poll?  I’d suggest Gallup, since they have an unassailable reputation, and also because they don’t work for any candidates.
• How should the questions be phrased?  I’d let Gallup recommend the exact language, including the crucial screening questions.  If a choice must be made, I’d leave it to the leadership of the DNC Rules Committee and/or Howard Dean.  Is that perfectly fair?  No, but it’s the best we can do under the circumstances.
• How would the delegates be awarded?  The method needs to be completely defined in advance, so there’s no room to finagle after the exact results are released.  I assume most of the delegates would be awarded according to the percentages within each CD, with a smaller pot assigned based on the statewide numbers, but it doesn’t really matter.  It could mirror the allocations already established by the MDP.
• Wouldn’t a poll under-represent people who don’t have land-lines?  When I referred to a poll above, I didn’t specify that it would be a conventional, telephone-based poll.  Because adequate funds and time would be available, I would suggest it follow procedures similar to the US Census:  try one method, shift to a second, and then a third.  In this case, I would start with telephone interviewing, if that fails shift to in-person, and finally try First Class and then Express Mail.  I would continue until each selected voter has either been interviewed, has refused, or ten attempted contacts have failed – that’s a lot.  I would “weight” the results by CD, so each district receives as many “votes” as it is entitled to by the MDP’s existing delegate assignment.
• Polls have large margins of errors.  Compared to what?  An election where one of the candidates wasn’t listed on the ballot?  Dividing the delegates exactly equally?  Nobody associated with the process so far, or with either of the campaigns, has any right to complain about the possibility of 5% random error.  I would conduct about 400 interviews per CD, or about 6000 statewide.  Random fluctuations in the sample should amount to a total of three or four delegates – pretty small potatoes compared to the schemes we’ve been hearing.
• But are we going to leave out campaigning?  Should we run a process that has no role for media/ rallies/ mail/ signs?  This is the only point in this proposal where my demented genius really comes into play:  I would announce on April 1 that the poll will be conducted (say) May 30 – June 4, giving the two candidates lots of time to campaign in Michigan, trying to influence all the people who might later be randomly selected.   In effect, we would ALL be targetted by the two campaigns, because they won’t know which one-half percent of us will actually get to vote.
• Where would the money come from?  Total cost would be under $300,000 – chickenfeed by the wasteful standards we’ve seen.  The DNC would be able to provide it, I’m sure.
• How would the specific delegates be chosen?  I suppose the safest method would be to simply inform each campaign they have won XXX delegates and alternates, and tell them how many need to fall into each category, and let them select from their supporters.  If it were up to me, I would NOT seat any of the superdelegates, unless a campaign chose them, as punishment for their parts in bringing about the fiasco we’ve seen to date.  (A zany  alternative would be to offer the seats first to people interviewed in the survey – a genuine cross-section of Michigan Democrats.  It’d be fun, but probably unworkable.)
• Wouldn’t Hillary complain that this is unfair because of X, and Barrack because of Y?  Of course they would but this proposal only needs the votes Dean controls on Rules, plus either of the campaigns.  Whichever campaign is guilty of the most outrageous game-playing gets left out when Gallup comes to ask for guidance, since a working majority won’t require their votes.  Dean really holds the whip hand here.
• What right does the DNC have to impose its own solution?  Essentially, the DNC would be saying they got tired of waiting for the MDP to submit a proper plan under the national rules.  Because Michigan (and Florida) Democrats deserve to be represented in Denver, regardless of the incompetence of their state leaders, the DNC has adopted a stopgap measure.  If the MDP doesn’t like it, the solution is to submit a legal plan.  This is very similar to the DNC’s actions forty years ago in rejecting racially segregated delegations from the South.
• Wouldn’t the public see this as unfair?  Compared to what – the present mess?  Actually, I think a fairly-run process, with Gallup playing a central role, would make a compelling story-line for the media and the voters.  It would be weird, but it would feel even-handed.

Is there any chance for this proposal?  Well, no.  But I decided it was worth the effort to put it forward.