One of my diary drafts is entitled "The Florida/Michigan Revote Card, and the State of the Race". The diary was going to argue that holding fresh primary votes in both Florida and Michigan was unambiguously in the interests of the Clintons. Moreover, the diary was going to predict that we were going to begin to see some Clinton surrogates start to send out trial balloons about potential Florida and Michigan votes at about the time of the Ohio and Texas primaries.
I never finished the diary, and I can't really take credit for a prediction that I never got onto the public record. But, indeed, this is exactly what has happened. I want to at least complete the second part of the diary, and explain why the Clintons are doing this. In fact, I would argue that Florida and Michigan revotes are an essential part of any potential Clinton path to the nomination.
Let's step back and look at this from Clinton's sort of mathematical point of view. Her basic calculation is as follows:
It is in Clinton's best interest to hold revotes in Florida and Michigan when the left side of the equation exceeds the right side of the equation. Namely, her expected gain in pledged delegates plus superdelegates in the event of a revote exceeds her expected gain in the event that no new vote is conducted.
This should be fairly self-evident. What's trickier is estimating the individual parameters. It is not as simple, for example, as simply assigning Clinton the number of "phantom" delegates that she received from the unsanctioned Florida and Michigan primaries in the event that there is no revote (P_n), because we have to assess Clinton's likelihood of success in front of the credentials committee.
In fact, let's tackle this parameter -- (P_n) -- first. Marc Ambinder does a good job of explicating the mechanics of the credentials committee:
So early this summer, the 25 members of the DNC's credential committee will join 161 other members to be selected by the states.
The members do not have to be Convention delegates.
The presidential candidates get to choose them, based on the delegates allocated to them in that particular state primary or caucus.
It's proportional (of course).
So if a state has 4 members, and one candidate wins 50 percent and another candidate wins 50 percent, then each candidate gets 2 members.
What does this mean?
Let's say that ALL 25 of Howard Dean's appointees vote AGAINST seating Florida and Michigan. Let's say that 80 additional members are appointed by Obama and 81 by Clinton. 25 + 80 is more than 81. You can fiddle with the numbers and get to a scenario that might seat the Florida delegates. But it's safe to safe to say that the credential committee option for Sen. Clinton to get Michigan and Florida seated would require her to have won more delegates than she will win.
Most of the members of the credentials committee are allocated to the individual campaigns, in proportion to the number of pledged delegates that each campaign receives. Moreover, the candidates can select whomever they please to seat at the credentials committee, so defections of any kind are unlikely. There is also a minority of delegates who are selected in essence by Howard Dean. Since it was Dean's DNC who applied the Florida and Michigan sanctions in the first place, one would think it unlikely that these appointees would vote to seat the Florida and Michigan delegations in any sort of manner that Mrs. Clinton would be pleased with.
Now, politics are unpredictable, and there are also some end-around scenarios by which Clinton could attempt to force a floor vote on Florida and Michigan by filing a minority report, so we can't entirely rule out Clinton gaining some delegates through this path. However, it is not all that likely, and in some of the scenarios where Clinton "wins", perhaps only some fraction of the Florida/Michigan delegates would be seated. We might hypothesize probabilities like the following:
I would imagine that the majority of the time, either no delegates at all are seated, or the delegtes are seated, but split equally between the two candidates (these two scenarios are functionally equivalent). On the other end of the spectrum, Clinton's best-case scenario -- both Florida and Michigan are seated in full, and Obama gets no credit at all for the Michigan uncommitted delegates -- seems rather unlikely, as it wouldn't pass most people's "sniff test" in terms of fairness to Obama. In between these two cases are a whole host of compromise possibilities.
Of course, these numbers are just guesses. But, if they are somewhere in the vicinity of being correct, I would guess that Clinton's expected value from attempting to seat the Florida and Michigan delegaets without a revote is somewhere in the neighborhood of 20 pledged delegates.
Next, let's look at (P_r): Clinton's expected gain in pledged delegates in the event of a revote. This one is more straightforward. Rasmussen polls out today suggest that Clinton would win by 16 points in Florida, whereas Michigan would be a tie. Obviously, Clinton would gain no ground in Michigan if Rasmussen's prediction were true. But a 16% margin in Florida is worth approximately (16% x 185) = 29.6 pledged delegates. This is actually more than our estimate of the number of delegates that Clinton would pick up by pressing her case with the credentials committee!
Of course, Rasmussen's numbers could be wrong, since we might be as many as three months away from a prospective revote in Florida and Michigan. Obama has gained ground in literally every state in which he's had the chance to campaign, including states like New Hampshire and Ohio which he eventually lost.
Nevertheless, I think we can acknowledge that Florida is inherently a pretty tough state for Obama, with its combination of older voters, Hispanic voters, Jewish voters, and Southern Baptists whites, all groups that have tended to favor her in the primary process. Florida would poll, I would guess, somewhere around 10 points behind the national trend; if Obama outright wins a vote in Florida, that means that the overall national trend has become at least a minor landslide in his favor.
Michigan, on the other hand, could well be a pretty decent state for Obama. Michigan is my home state, and I would argue that it falls somewhere between Wisconsin and Ohio on the political spectrum, but closer to the Wisconsin end of the spectrum than the Ohio end. It does not have the Appalachian and Southern pockets that Ohio has. It does have two of the twenty largest Universities in the country. It has of course tended more blue than Ohio in recent Presidential elections. It has one of the largest African-American populations outside of the South.
We have gotten off on a bit of a tangent. Michigan is a state Obama could win -- but it is also a state Obama could lose, and that's what matters from Clinton's perspective. Neither candidate has any reason to be afraid of a revote there.
The other part of the equation is the superdelegates. I am not talking about the handful of superdelegates from Florida and Michigan themselves -- but rather, the moral claim that Clinton could make to the nomination under each of these scenarios, which could in turn sway superdelegates in all 50 states. I would argue that Clinton's moral claim is potentially much stronger in the event of a revote than trying to stand by the initial results.
If there is no revote in Florida and Michigan, Clinton's argument is that those voters have essentially been disenfranchised, and that she deserves some sort of extra credit for that. This is S_n in our nomenclature. However, this claim is not likely to be especially persuasive. For one thing, if there are no revotes in Florida and Michigan, Clinton will likely be perceived as having undermined that effort, and Obama will be able to make the case that she is afraid of a fair fight. This is particularly the case in Michigan, where his name was not on the ballot. For another thing, an overwhelming majority of Democrats in both Michigan and Florida want a revote; a fresh vote is favored by a 62-24 margin in Michigan and a 63-28 margin in Florida, according to Rasmussen.
Of course, before appealing to the superdelegates, Clinton will first try and seat the delegates through the credentials committee. If she succeeds in her effort, she will have no claim to "extra credit" at all. And whether she succeeds or she fails, if the process becomes sufficiently nasty, some superdelegates could actually gravitate toward Obama as a result. Thus, P_n and S_n are not entirely independent of one another; Clinton might be able to seat some delegates from Florida and Michigan or she might have an argument to press to superdelegates, but she can't really have both.
On the other hand, Clinton will be able to make a different sort of moral claim (S_r) in the event of a revote -- provided that she wins both Michigan and Florida. This will be a version of the argument we've heard already: Clinton is capable of winning the "big", "important" states, and/or she has the momentum and has become the consensus choice of Democrats, even if she hadn't been throughout the primary season. Clinton would be able to argue that she had defeated Obama in each of Ohio, Texas, Pennsylvania, Florida, and Michigan. This is potentially a rather persuasive argument, and it's something that Obama supporters should be prepared for. In fact, I suspect this argument is at least as likely as not to succeed.
On top of that, Obama's pledged delegate lead will have been eroded, though probably not overturned, by these victories, so Clinton's threshold for persuasion is much lower. Let's plot out an optimistic, though not inconceivable, delegate scenario for Clinton:
State Obama Clinton Popular Vote
WY 7 5 Obama +17
MS 18 15 Obama +9
PA 67 91 Clinton +15
Guam 2 2 TIE
IN 33 37 Clinton +6
NC 58 57 Obama +1
WV 11 17 Clinton +21
KY 20 31 Clinton +22
OR 27 25 Obama +4
PR 17 38 Clinton +38
MT 9 7 Obama +13
SD 8 7 Obama +7
FL 75 110 Clinton +19
MI 59 69 Clinton +8
411 511
Under this scenario, Clinton wins Pennsylvania pretty big. This is enough to hinder Obama's momentum in North Carolina and Indiana; they essentially tie the former state whereas Clinton comes out ahead in Indiana. Clinton cleans up in Puerto Rico, and then wins Florida and Michigan by margins somewhat larger than in the current Rasmussen poll.
Under this scenario, Clinton would gain a net of exactly 100 pledged delegates. Obama would still have a net advantage of 50-60 pledged delegates, which is slightly larger than Clinton's advantage in superdelegates now. However, we might reasonably expect that the balance of undecided superdelegates would gravitate toward Clinton, if such a scenario were to transpire. This is basically her path to the nomination.
On the other hand, if there are no Florida and Michigan revotes, I don't think Clinton can build up enough momentum to have a strong enough moral claim on the nomination; the clock would essentially run out on her. She would probably have to win North Carolina and Indiana outright, while finding some way to significantly close the delegate gap elsewhere along the way. If Clinton is to make a persuasive argument that she has run the table, she needs Florida and Michigan back on that table.
So what can Obama and his supporters do?
#1, Obama should recognize he's negotiating from a position of strength. This does not mean that Obama should try and block a Florida/Michigan revote; such efforts would make him look inconsistent, and would only weaken his moral claim to the nomination. But he should recognize that Clinton needs these revotes far more than he does, and use that as leverage both in the way he negotiates things like dates and rules, as well as how he manages the PR aspects of the campaign.
#2, Obama needs to start working on a counter-narrative, in the event that something like the "worst case" scenario I outlined above transpired. In particular, he needs to talk up the importance of North Carolina, and to a lesser extent, Indiana. More delegates will be awarded on May 6 (185) than in Pennsylvania (158). And if there's any sort of arbitrary dividing line between "big states" and "small states", North Carolina belongs on the "big state" side of that dividing line; it has nearly the same number of delegates (115) as Michigan (128).
#3, Obama could consider offering to seat the Florida and Michigan delegations, subject to the requirement that Michigan's uncommitted delegates are pledged to Obama. This would cost him 56 delegates -- more than Clinton would probably expect to receive through the credentials committee. However, it would completely neuter any moral claims that Clinton could make to Michigan and Florida. In fact, Obama would look completely magnanimous, and would probably have an easier time framing the case that superdelegates should line up with the pledged delegate winner.
And #4, Obama needs as many donations as possible, because these revotes are a'comin, and they're going to be expensive.