With the Iowa Caucus county convention results now in, and Obama leading 25-14-6, Barack O'Bama leads Hillary Clinton 1418-1251 in pledged delegates.
That's a lead of 171.
Update: 169, based on California's adjustment being +3 for Obama rather than +4. All numbers you see below are changed by this change.
John Edwards' 26 pledged delegates dropped to 18. 6 are in Iowa, 4 are in New Hampshire, and 8 are in South Carolina.
There are 10 certain contests left (8 states, tiny Guam and Puerto Rico), and 2 uncertain ones. I'll briefly run through a couple of scenarios on the Michigan and Florida situations at the end, but the diary will mainly focus on the contests we know will occur.
[Monday afternoon update]: Obama gains 1 superdelegate today, and with some small fluctuation in Colorado and other post-certification numbers, Obama has updated its election results center site, so I thought I'd go through and re-update the entire diary with these slight adjustments, just for accuracy. The number is now 167, which is reflected in all totals and percentages.
Pledged Delegates
O'Bama's spreadsheet is a decent place to start. It has been pretty accurate. Had the dandyjack Irishman from Hawaii gained another .335% of the vote in Mississippi (did Limbaugh generate 2207 or more voters out of Clinton's 155,686, or 1.42% of her votes?) then Obama would have hit his spreadsheet's 20-13 prediction there.
For the rest of the way, O'Bama's sheet says a dead even pledged delegate count: 283-283. When they have erred, they have generally erred on the side of caution. For instance, they predict only +7 in North Carolina, which has seen its neighbors Virginia and South Carolina go by 29% and 28% margins respectively. O'Bama starts with a clear polling lead in North Carolina, he has won easy double digit popular vote wins in every state in the Virginia-to-Louisiana arc, so I would nominate North Carolina to be a state where the sheet under-predicts his actual win.
But let's use a net gain of zero for either candidate as the baseline. O'Bama's team predicts wins in Indiana, North Carolina, Oregon, Montana, and South Dakota. They predict Clinton wins in Pennsylvania, West Virginia, Kentucky, and Puerto Rico, with a 2-2 delegate tie in Guam.
Here's the prediction list:
PA: O 75, C 83 - Closed
Guam: O 2, C 2
IN: O 39, C 33 - Open
NC: O 61, C 54 - Semi-open
WV: O 13, C 15 - Semi-open
KY: O 23, C 28 - Closed
OR: O 28, C 24 - Closed
PR: O 25, C 30 - Open
MT: O 9, C 7 - Open
SD: O 8, C 7 - Closed
Total: O 283, C 283
Added info: Closed = Dem only; Semi-open = Dems and Independents; Open = anyone.
Caucus Net Delegate Gains/Losses
First, let's be clear. Obama expanded his lead by 10 in Iowa, as NBC confirms. Obama's delegate totals are almost always right, while it takes the news organizations a longer time to finally catch up with them.
Now, there have been several diaries discussing Clinton open efforts to try and sway delegates moving up the chain from the election nights to the county conventions and state convention. Iowa should be a very calming tonic for anyone worrying about the lanky Irish lad's ability to handle a close-in fight. He might not win every skirmish, but we at least have evidence of one thing: the Clinton folks are publicly talking about it, the O'Bama folks are shutting up and just getting the job done. Silence in news articles does not mean inaction.
As we will see momentarily, Iowa was really the only major state with serious delegate fluidity because the Edwards freebies were up for grabs. Colorado is the only other state, but a massive statewide success at this at best nets Clinton 2 more delegates. This will be a negligible impact if it happens.
Add-on Superdelegates.
There will be 76 or 81 add-ons (MI = 2, FL = 3). There is a stupendously good explanation here of how the caucus states do their add-on selections. It even specifically responds to the notable Colorado anxiousness diary here and explains why it's not a big deal.
If you want the abbreviated version, the 5 states where caucus numbers can move in any real, fluid way are Iowa, Nevada, Colorado, Alaska and Maine. Call them the Zero Locked Delegate states. However, only Iowa (because only Iowa had Edwards delegates) will see significant movement as FlyontheWall explains:
The Colorado experience is instructive in this regard - one county showed a five-point swing toward Hillary, another a four-point gain, and a third ran true to the precinct results. A two- or three-point swing statewide for Hillary would enable her to take a delegate from Obama; a five- or six-point swing would allow her to take two. And that in the largest of the undetermined states. We're just not talking about huge numbers of delegates, and it's unclear that the dynamic will favor Hillary across the board.
Kansas, Nebraska, Washington and Wyoming are in another category of caucus state - call this the Mostly Locked Delegate states. Basically by state rules most of the pledged delegates are already locked in and about a third are fluid (95 are locked, 51 are potentially vulnerable to bad attendance and switching, etc.). This differs with the previous category because none of those delegates are locked.
A third category is caucus states - the Fully Locked Delegate states - whose tallies directly translate to the ultimate totals. When these states hold state conventions, they are not determining allegiances, just identity. Allegiance was locked in on the initial vote - meaning the Minnesota 72 caucus delegates were as set in stone as Missouri's 72 from the primary. These states are American Samoa, Idaho, Minnesota, North Dakota, Hawaii and the Virgin Islands.
In all caucus cases, whoever wins more than 50% of the delegates at the state convention wins the add-on. Only states and DC get add-ons (so none for American Samoa, Virgin Islands, etc.).
Thus, we can basically now say O'Bama gets the following add-ons for sure: O'Bama gets 14 from all the above list (1 each except Minnesota and Washington state each get 2). The only one I haven't allocated is Nevada. You would think Obama would get the add-on for having more state convention delegates, but since it is in the fluid group of states it's unclear, especially with the chaos at the initial attempt. FlyontheWall gives that add-on to Clinton. So, 15-0 or 14-1 in caucus state add-ons, in favor of O'Bama.
Oh, and there's also Texas, with its 3 add-ons. The caucus process controls the add-ons, meaning O'Bama gets them. The bottom line is O'Bama is looking at 17-1 or 18-0 in the add-ons won from caucus states. For what it's worth, FlyontheWall assigns the Texas 3 to O'Bama but the Nevada 1 to Clinton, making it 17-1 for a net +16 in caucus-state add-ons.
Primaries are a different story. Some of this vote has to do with who controls the process. Using Alabama as an example, Fly writes:
Given that some of these will be awarded by a vote of a body whose composition is already a matter of public record - a state convention, the members of the DNConvention delegation - we can actually be as certain of these UADs votes as of those of their pledged peers, even if we don't yet know their names.
Under this criteria, Fly projects Arkansas (1) and California (5) for Clinton. For O'Bama, he projects Alabama (1), Delaware (1), Illinois (3), Utah (1), DC (1), Virginia (2) and Vermont (1). Alabama is known, and the others are apparently processes where the choosers are already on public record, and we know which way they endorse. Since all states use different criteria, these are the only ones whose process is known enough and the endorsers clear enough to project the winner of the add-on. At any rate, this is an additional 10-6 split, on top of the 17-1 caucus split. So 27-7 O'Bama edge so far, with 42 or 47 undetermined. (Note: as of March 6, Fly had not yet allocated Iowa's 1, which we now know belongs to O'Bama since he is over 50% of delegates in Iowa, or Hawaii's or North Dakota's 1 add-on. Based on Fly's March 11 update referencing the processes in Hawaii and North Dakota, I feel comfortable assigning these to O'Bama also).
Again, let me stress that I am using 0 as the baseline number for the following analysis, even though we can kind of speculate that O'Bama will have an edge when all is said and done. It's a nice counterweight to the "what if he loses PA by more than the 8 delegates his spreadsheet predicts?" questioners.
Undeclared Superdelegates
The campaigns have provided lists of superdelegates to different news outlets. Chuck Todd explicitly uses lists the campaigns have provided him plus public endorsements. Other news orgs have slightly different numbers, as does Democratic Convention Watch, which links each name to an announcement of endorsement or of neutrality (on the undeclared page). Todd's current number as of this morning is 253-217, but it is also clear from noting that Michigan has 25 and Florida 29 that he lumps add-ons into the superdelegate total, which makes it hard to give a clean breakdown the way that Dem Convention Watch does.
The bottom line is 795 minus 76 add-ons = 719 supers. Add Florida and Michigan's 49 supers and 5 add-ons and you get 849 minus 81 add-ons = 768 supers. (And if roughly 36 million vote in the primaries and caucuses, then those approximately 720 supers get votes 50,000 times more powerful than yours or mine.)
Of these 719 or 768 supers, most have declared. As I wrote last weekend, there is zero evidence of O'Bama-to-Clinton defections, and a small handful of Clinton-to-O'Bama defections such as David Scott, John Lewis and Roz Samuels. Because we can't be certain of the MSNBC, AP, CNN, CBS pollution factor of including add-ons in the total, we'll use Dem Convention Watch's number of 244-207 (they show 209 minus 2 add-ons). This does not include Florida and Michigan, which they have 15-5 for Clinton.
This means, according to Dem Convention Watch, there are 451 declared supers and 268 undeclared ones, or 471 declared supers and 297 undeclared ones, varying with the inclusion of FL/MI.
Without FL/MI, plus the pledged 1418-1251 number: O'Bama 1624, Clinton 1495. Lead of 129 with 18 unallocated Edwards, 566 unallocated pledged, 76 unallocated add-ons, 268 undeclared supers.
With FL/MI, plus 1418-1251: O'Bama 1629, Clinton 1510. Lead of 119 with 31 unallocated Edwards, 879 unallocated pledged, 81 unallocated add-ons, 297 undeclared supers.
Let's take the without FL/MI scenario first.
A lead of 129, with a baseline assumption of a tie in unallocated pledged, and a baseline assumption of a tie in add-ons (which is our built in cushion), and a default assumption of a 9-9 tie in Edwards votes in Iowa, New Hampshire and South Carolina, means Clinton has to win the remaining 268 undeclared supers by 129 just to tie. Which would be 199-69.
Under this scenario, she has to win 74.3% of the remainder, and the add-on edge still cushions O'Bama to get a first-ballot win at the convention.
Now, Florida and Michigan have some different scenarios to look at.
If you start with the baseline of a full re-do, plus the default add-on tie and Edwards tie and the net tie in pledged delegates in the other 10 contests, you begin with a lead of 119, 313 unallocated pledged in Michigan and Florida, and 297 undeclared supers.
The range of reasonable results is probably +5 for Clinton in Michigan and +30 at the Florida extreme, all the way to +7 for O'Bama in Michigan and +15 for Clinton in Florida. In other words, a net of +8 to +35 for Clinton in unallocated pledged. That would bring O'Bama's lead to somewhere between 84 and 111 delegates, with 297 undeclared supers.
That means the 297 would have to break for her 191-106 (best case) up to 204-93. Which is still 64.3% to 68.7% of this group she would have to win to get a tie. (Remember, we've still got an add-on cushion for O'Bama in there.)
There's also the Michigan 50-50 pledged delegate tie scenario and the half-vote Florida scenario. That would give Clinton +19 in pledged delegates, meaning his lead would be 100 (119-19) and all the default tie assumptions and 297 undeclared supers. 199-98 would have to be the win to get a tie, and that's 67.0%.
Undeclared Superdelegate Summary
What this means is we have a range, from Clinton's most optimistic scenario of needing 64.3% of 297 undeclared supers to a (more realistic) scenario of 68.7% with Florida and Michigan. And that is with some unrealistic assumptions of add-on ties thrown in. Basically 2-to-1.
Without Florida and Michigan, she needs 74.3% of 268 undeclared supers, assuming unrealistic add-on ties and such. And see below, maybe as much as 77.2% of 241. Basically 3-to-1.
Here is a key argument point - if that significant an edge between 2-to-1 and 3-to-1 undeclared superdelegates is breaking hard for Clinton, that means something really strange is happening in the campaign. Some other factor is going on in the world to make the supers break that definitively.
There is always a lot of bluff and bluster from the Clintons about trying to poach delegates and whatnot, but the truth is if they were so forceful and powerful they'd simply get the undeclared supers they need and not have to worry about stealing elected pledged delegates. I know a bluff when I see one. The fact is, the supers are not kowtowing, because if they were, they'd be declaring. Does anyone think the Clintons have superdelegates they're asking to keep silent? They'd be out there trumpeting these people in front of the world to show how the supers are "rejecting" O'Bama after Clinton won a state that matters.
Additional note: I used Dem Convention Watch's numbers, which are lower than Chuck Todd's. If Todd's 253-217 number is right, reflecting lists that the campaigns have provided plus public declarations, then there are 18 or 20 (depending on how many add-ons are in there) more supers to reduce the 269 number (which have split 9-9 or 11-9 for O'Bama). For example, First Read notes that the superdelegate from Illinois today was already on the campaign-provided list, if not public. Taking a chunk of 18-20 away from the 269 gets us to 239-241 left, and the percentages spike upward to overcome a 131 deficit, reflecting on the range of percentage of undeclared supers. It would be 186-55 Clinton would need, or 77.2%.
A Discussion of Popular Vote
There has been a lot of talk (the execrable Pat Buchanan is an example of a TV talker pushing this argument) that if Clinton wins the popular vote she will have a legitimate argument to make to uncommitted superdelegates. Buchanan, for one, thinks this clinches the nomination for her. You hear adjectives like "possible" and "realistic" thrown at this scenario, and it's all bunk.
I will now show why.
Kos did a popular vote tally here. The only issue I have with it is for some reason he gave Clinton 30% of the popular vote in Iowa when that was Edwards. Clinton got 29%.
I emailed Chuck Todd this past week over how he is planning to count popular vote totals when caucus states like Iowa, Nevada, Maine and Washington did not report totals. Todd replied that they had a projection based on Iowa and Nevada, that they were working on Washington state. He didn't say about Maine (thank you ClaudeB), but I am guessing that was by omission in the email not omission in the calculating. I also asked him about Texas, which he said because everyone who voted in the caucus had definitionally already voted, he didn't think counting them twice was fair. I accept that argument as a fair one.
I also learned that his middle initial (it's on his email) is, in fact, D.
At any rate, we'll take Kos' 813K number and do a little projecting, because I know it'll come up in the comments. One way to make predictions about turnout is by congressional district. Each congressional district has roughly 650K or so people in it. Ohio has 18 CDs. PA has 19. Virginia has 11 CDs. North Carolina has 15.
PA - 19 CDs (1 more than Ohio)
Guam - negligible
IN - 9 CDs (same as Missouri)
NC - 15 CDs (4 more than VA)
WV - 3 CDs
KY - 6 CDs (combined with WV has 9 CDs like Tennessee)
OR - 5 CDs
MT - 1 CD (biggest CD in America)
SD - 1 CD (with OR and MT approximately the # of votes as Wisconsin's 8 CDs, since MT and SD are bigger than normal CDs)
In fact, I think PA is analogous to OH, Indiana is analogous to Missouri, NC is analogous to VA, WV + KY is analogous to Tennessee, and OR/MT/SD will be won by an overall Wisconsin-like margin.
Clinton won by 228,000 in Ohio (PA would be 250,000)
O'Bama won by 10,000 in Missouri
O'Bama won by 275,000 in Virginia
Clinton won by 82,000 in Tennessee
O'Bama won by 193,000 in Wisconsin
I know this is rough and dirty. Overall population is not a perfect match state to state in terms of voting. Ohio was open while PA is closed, for example, though PA has a full extra CD. O'Bama will win NC but maybe not by as wide a percentage as VA (though NC has a full 4 extra CDs).
But it kind of feels about right, doesn't it? If we took those numbers and added them, O'Bama actually gains 146,000 votes. Then there's Puerto Rico to factor in. I don't know how many people will vote in Puerto Rico or how wide the margin will be. Let's use up this whole 146,000 on Puerto Rico just to be safe.
The overall picture is clear - O'Bama is going to be at least 813,000 ahead over those 10 contests.
Then there's Florida and Michigan. Michigan with O'Bama's name on the ballot would be so close as to be negligible (it has 15 CDs, by the way, two fewer than Wisconsin + Indiana). To me, Michigan is Ohio minus the Ohio River Valley where he does poorly, plus the Upper Peninsula which is more like Wisconsin where he did well.
Florida with Edwards on the ballot siphoning off 14% of the vote had a Clinton ceiling of 288,000 gain.
Is this popular vote thing clear now? I think you can see that the popular vote has about a nil chance of being won by Clinton, despite countless bloggers and pundits suggesting she has a shot at winning it.
A final word.
In the first paragraph of the NYT story today about the uncertainty that looms among undeclared supers, I almost went on a full out rant diary over the bolded part of the opening paragraph:
WASHINGTON — Lacking a clear route to the selection of a Democratic presidential nominee, the party’s uncommitted superdelegates say they are growing increasingly concerned about the risks of a prolonged fight between Senators Hillary Rodham Clinton and Barack Obama, and perplexed about how to resolve the conflict.
Perplexed? What. The. Hell. You're an undeclared superdelegate and you just can't figure out how to resolve the conflict? What possible options are there, after all? Is this The Onion? The rest of the article talks about how they'd really like it if Al Gore came riding to the rescue or Howard Dean or Nancy Pelosi.
I am not going to launch into that rant (I'm a feisty Irishman myself), but simply and soberly say that I pray for the day when Democrats have a spine again, when it is back in their DNA. All this risk-aversion is despicable. When you know exactly who you are, when you know your values and can articulate them, you are free to go out and achieve great things. I wrote a diary earlier this week that openly pleaded with two undeclared superdelegates who I have a personal connection with and who this community holds dear to seize a leadership moment in a vacuum of non-leadership. I know this - someone will have to do it.
And that's where we stand, the day before St. Patrick's Day.
(Minor text edits and a couple of links added.)