The Delegate Fight: Obama Clinches!

[Gustaf]

So, isn't it time to do some calculations on the remaining states? Some of my guesses on the states where complex district knowledge seems less required..:

Pennsylvania: If Hillary wins with 57-58% of the vote it would mean 31-24 in at-large delegates. I know a couple of people have gone through the districts, so I'll try to add some average of those predictions later on. EDIT: seems like BRTD was the only one who gave a full analysis. He had Clinton +1, but looking at the results out of Wyoming and Mississippi and his idea of a 10% Clinton win I suspect he was a bit biased.  Clinton +8 for now, but I hope someone else can add something to this. 

Indiana: Based on BRTD's reasonings, which seems sensible and assumes a 10-point Clinton victory, she goes +3 in district delegates and another 3 on at-large ones. So, +6 in total. Unless someone else steps in with more info, I'm not gonna change this one.

North Carolina: Seems like a racially polarized, close state. I'm gonna make it a wash right now, till I have more info.

West Virginia: All districts are even-numbered. As is the at-large count. So if Clinton wins somewhere in the 60s she'll win the at-large 6-4 (possibly 7-3) and probably get all of the districts 4-2. I'm guessing Clinton +8.

Kentucky: Closed. Southern. White. Rural. Yeah, not going to be nice for Obama at all. I'd expect Clinton to take the at-large delegates 12-5. Arbitrarily I'll guess the district one break in a similar fashion, Clinton +11. Total gives Clinton +18.

Oregon: Alcon? Anyway, closed primary while Washington was an open caucus. So I'd expect Clinton to do a lot better than she did in Washington. Which she lost bya 2-1 margin. So...I'd expect the at-large ones to split about 10-8 for Obama. Possibly 11-7. I'll just throw a random guess and triple that for the districts until someone enlightens me on the district break-down. So Obama +8 in total.

South Dakota: Obama +2 among the at-large and +3 among the district ones (It's a closed primary so I doubt he breaks two thirds of the vote, but I'm assuming he can break 58%) Obama +5.

Montana: At large will probably split 4-2 for Obama. If we assume each half ends up as an Obama win with between 50% and 70% that gives him 3-2 wins in each. So, overall that's +4 Obama.

Puerto Rico: Closed. Hispanic. Likely to be ugly for Obama but hard to tell how ugly. No fancy stuff, simlpe allocation  by percentage. Given the previous results in the Hispanic vote (and keeping that Dominican result in mind) I'm gonna randomly put Clinton at 60% here. That makes it 33-22 and Clinton +11.

So total from those would be +34 for Clinton.

[they don't love you like i love you]

Indiana: Based on BRTD's reasonings, which seems sensible and assumes a 10-point Clinton victory, she goes +3 in district delegates and another 3 on at-large ones. So, +6 in total. Unless someone else steps in with more info, I'm not gonna change this one.

Actually my initial projections are incorrect because they were based off the raw delegate numbers for Indiana and not the actual numbers that'll be used because of the delegate bonus deal. With the bonus, only one district exists with an odd number of delegates, 6. Hillary will win that one. 3, 4 and 5 are all 4 delegates, the rest are 6. So the issue becomes if Obama can hold Hillary below 58.333% in all of the rural districts and break that number in the Indianapolis one.

North Carolina: Seems like a racially polarized, close state. I'm gonna make it a wash right now, till I have more info.

Unlikely to be a wash with the polling numbers. Also the delegate allocations benefit Obama, simply because he's stronger in most of the districts with odd numbers of delegates.

Oregon: Alcon? Anyway, closed primary while Washington was an open caucus. So I'd expect Clinton to do a lot better than she did in Washington. Which she lost bya 2-1 margin. So...I'd expect the at-large ones to split about 10-8 for Obama. Possibly 11-7. I'll just throw a random guess and triple that for the districts until someone enlightens me on the district break-down. So Obama +8 in total.

I did one earlier debunking J. J. Here's it in a nutshell:

CD 1 - Obama wins 4-3. It contains some blue collar areas but they only make up 20% of the vote. The bulk of the district is part of Portland and affluent Portland suburbia.
CD 2 - Obama wins 3-2. This district is a lot like Idaho and the west, it's a primary unlike those, but the largest city where most of the Democrats live (Bend) favors Obama. Hillary can only win here if the Hispanics actually vote (and they hardly ever do.)
CD 3 - The main Portland district. Obama wins 6-3. 7-2 might even be doable.
CD 4 - Obama wins 4-3. Hillary can't win a district that contains Eugene.
CD 5 - Hillary's best chance for a win probably, but it's also 6 delegates. She can't win it by enough to get more than 3-3 (especially since it contains Corvallis).

South Dakota: Obama +2 among the at-large and +3 among the district ones (It's a closed primary so I doubt he breaks two thirds of the vote, but I'm assuming he can break 58%) Obama +5.

In South Dakota, a closed primary might actually benefit Obama, when you consider the type of people who are actually Democrats and the people who voted for both Bush and Daschle.

Puerto Rico: Closed. Hispanic. Likely to be ugly for Obama but hard to tell how ugly. No fancy stuff, simlpe allocation  by percentage. Given the previous results in the Hispanic vote (and keeping that Dominican result in mind) I'm gonna randomly put Clinton at 60% here. That makes it 33-22 and Clinton +11.

It's not closed. It'd be impossible to be closed actually considering the Democratic party doesn't even officially exist in Puerto Rico.

Remember, 20% of Puerto Ricans are black or mullatto.

For the record I came pretty close in Mississippi. The only mistake I made was calling all the districts for Obama, Hillary narrowly took one.

[jimrtex]

On March 29, Texas is holding its 2nd-tier caucuses (County / SD level).  These (if results are published) should give us a good idea, for the first time, of Obama's exact margin out of the TX caucuses (still only 41% reporting as of now).

By law, all conventions (this included the precinct conventions) are open to the press. 

I received an e-mail from the chairman Texas Democratic Party - apparently because I registered to get access to the precinct caucus results.  It says in part:

You must be logged in to read this quote.

I do not know which "one campaign" made the suggestion.  But even if they go through the formal credentials process, it could make for a pretty ugly scene at a one-day convention with a lot of amateurs involved.

Texas law permits delegates present at 2nd tier convention to cast all votes for their precinct.  So if a precinct were entitled to 18 county delegates and 6 showed up at the county convention, each would cast 3 votes.  The Democrats apportioned one delegate for every 15 votes in the 2006 gubernatorial election so there are huge numbers of potential delegates (88,000) so there is the potential for stacking the conventions, at least for procedural matters.

However, the state convention delegates are selected by election precinct(s) based on the sign ups on election night, with one delegate for every 180 votes in 2006.  There may be some effects due to differential turnout on election night, since precincts will be combined at the county convention.

[Erc]

Pennsylvania: If Hillary wins with 57-58% of the vote it would mean 31-24 in at-large delegates. I know a couple of people have gone through the districts, so I'll try to add some average of those predictions later on. EDIT: seems like BRTD was the only one who gave a full analysis. He had Clinton +1, but looking at the results out of Wyoming and Mississippi and his idea of a 10% Clinton win I suspect he was a bit biased.  Clinton +8 for now, but I hope someone else can add something to this. 

I know nothing about PA politics (why bother, when we have so many experts here), but it seems pretty ludicrous to me that Clinton would only come out +1 in the district delegates if she got a >15% margin, regardless of how well Obama might do in Philly.

Here's the CD breakdown:

CD 9: 3 delegates
CDs 5, 10, 16, 17, 19: 4 delegates each
CDs 3, 4, 11, 12, 15, 18: 5 delegates each
CDs 6: 6 delegates each
CDs 1, 7, 8, 13, 14: 7 delegates each
CD 2: 9 delegates

[Gustaf]

Pennsylvania: If Hillary wins with 57-58% of the vote it would mean 31-24 in at-large delegates. I know a couple of people have gone through the districts, so I'll try to add some average of those predictions later on. EDIT: seems like BRTD was the only one who gave a full analysis. He had Clinton +1, but looking at the results out of Wyoming and Mississippi and his idea of a 10% Clinton win I suspect he was a bit biased.  Clinton +8 for now, but I hope someone else can add something to this. 

I know nothing about PA politics (why bother, when we have so many experts here), but it seems pretty ludicrous to me that Clinton would only come out +1 in the district delegates if she got a >15% margin, regardless of how well Obama might do in Philly.

Here's the CD breakdown:

CD 9: 3 delegates
CDs 5, 10, 16, 17, 19: 4 delegates each
CDs 3, 4, 11, 12, 15, 18: 5 delegates each
CDs 6: 6 delegates each
CDs 1, 7, 8, 13, 14: 7 delegates each
CD 2: 9 delegates

Oh, I agree. I just don't know where to start in calling the districts there.

As for the points BRTD raised:

1. Ok, so Indiana is somewhat in the air again.
2. I understand that it may not be likely. But until I see an analysis I'm not gonna call delegates for anyone. North Carolina is the only state where I'm really unsure about the winner as well.
3. Well, you're analysis of Oregon seems to yield the exact same results I had on pure guesswork.
4. You're entitled to your opinion, but I think Western indpendents and Republicans is a very ugly electorate for Clinton and is part of the reason why she did so badly in those Western states. Before you start looking it up, yeah, Obama has landslided some closed Western states previously. I still think it's a factor meaning SD will not be another Idaho.
5. It says closed on the site. But I've no idea what it means, considering they have their own parties, so I guess you're right. Not that there's likely to be many Republicans in Puerto Rico even if they had American parties. As for the mulatto thing, I didn't know. But are these fostered in the same traditions as African-Americans, with those voting patterns?
6. I don't think your predictions were that bad. But you, as did many others, apparently underestimated white turnout/percentage for Clinton. 

[they don't love you like i love you]

Pennsylvania: If Hillary wins with 57-58% of the vote it would mean 31-24 in at-large delegates. I know a couple of people have gone through the districts, so I'll try to add some average of those predictions later on. EDIT: seems like BRTD was the only one who gave a full analysis. He had Clinton +1, but looking at the results out of Wyoming and Mississippi and his idea of a 10% Clinton win I suspect he was a bit biased.  Clinton +8 for now, but I hope someone else can add something to this. 

I know nothing about PA politics (why bother, when we have so many experts here), but it seems pretty ludicrous to me that Clinton would only come out +1 in the district delegates if she got a >15% margin, regardless of how well Obama might do in Philly.

Here's the CD breakdown:

CD 9: 3 delegates
CDs 5, 10, 16, 17, 19: 4 delegates each
CDs 3, 4, 11, 12, 15, 18: 5 delegates each
CDs 6: 6 delegates each
CDs 1, 7, 8, 13, 14: 7 delegates each
CD 2: 9 delegates

Here's basically what I was thinking:

CD 9: 2-1 Hillary
CDs 5, 10, 16, 17, 19: All split except 10, 3-1 Hillary
CDs 3, 4, 11, 12, 15, 18: All 3-2 Hillary
CD 6: 3-3 split
CDs 1, 7, 8, 13, 14: All 4-3 Obama except 14 (4-3 Hillary)
CD 2: 7-2 Obama (few Hispanics or Reagan Democrats, Obama wins Center City whites easily.)

[Erc]

If Hillary is doing as well as she is statewide given Obama's strengths in those 7-delegate districts, I should think that Hillary would crack 62.5% in a fair number of those four-delegate districts (as has occurred in many other states).

[they don't love you like i love you]

Well looking at them:

5 - Very unlikely, it's got Penn State.
16 - Well maybe but as Rob has pointed out Lancaster County Democrats, while fairly rare, do appear to be more liberal than the state average. Also the part of Chester should favor Obama.
17 - Possible, but this is also probably the blackest rural district outside the south.
19 - I suppose this depends on if there are enough Democrats in the affluent parts of the York area.

[Erc]

Thanks to the discussion here, shifted one delegate to Obama in MS.  Late returns from Hinds County gave him an extra pledged PLEO.

[xzcyhj]

Obama'08 ResultsCenter(Pledged Dels)+AP(Superdelegates)

Obama  1418+215=1633
Hillary  1251+250=1501

[J. J.]

The MI 55 uncommitted delegates.  What are the minimums that each candidate can get?

[Tender Branson]

Which candidate will the remaining 21 uncommitted members of the US Senate back ?

Barbara Boxer - CA (won by Clinton)
Ken Salazar - CO (won by Obama)
Joe Biden - DE (won by Obama)
Tom Carper - DE (won by Obama)
Daniel Akaka - HI (won by Obama)
Tom Harkin - IA (won by Obama)
Mary Landrieu - LA (won by Obama)
Ben Cardin - MD (won by Obama)
Carl Levin - MI ("won" by Clinton)
Amy Klobuchar - MN (won by Obama)
Max Baucus - MT (TBD)
Jon Tester - MT (TBD)
Harry Reid - NV (state won by Clinton)
Frank Lautenberg - NJ (won by Clinton)
Jeff Bingaman - NM (won by Clinton)
Sherrod Brown - OH (won by Clinton)
Ron Wyden - OR (TBD)
Jack Reed - RI (won by Clinton)
Jim Webb - VA (won by Obama)
Robert Byrd - WV (TBD)
Herb Kohl - WI (won by Obama)

.....

So far for Obama (13):

Sen. Bob Casey (D-Pa.)
Sen. Kent Conrad (D-N.D.)
Sen. Chris Dodd (D-Conn.)
Sen. Byron Dorgan (D-N.D.)
Sen. Dick Durbin (D-Ill.)
Sen. Tim Johnson (D-S.D.)
Sen. Edward Kennedy (D-Mass.)
Sen. John Kerry (D-Mass.)
Sen. Patrick Leahy (D-Vt.)
Sen. Claire McCaskill (D-Mo.)
Sen. Ben Nelson (D-Neb.)
Sen. Jay Rockefeller (D-W.Va.)
Sen. Russ Feingold (D-Wisc.)

So far for Clinton (13):

Sen. Evan Bayh (D-Ind.)
Sen. Maria Cantwell (D-Wash.)
Sen. Dianne Feinstein (D-Calif.)
Sen. Daniel Inouye (D-Hawaii)
Sen. Blanche Lincoln (D-Ark.)
Sen. Patty Murray (D-Wash.)
Sen. Robert Menendez (D-N.J.)
Sen. Barbara Mikulski (D-Md.)
Sen. Bill Nelson (D-Fla.)
Sen. Mark Pryor (D-Ark.)
Sen. Charles Schumer (D-N.Y.)
Sen. Debbie Stabenow (D-Mich)
Sen. Sheldon Whitehouse (D-R.I.)

[they don't love you like i love you]

Feingold has already endorsed Obama.

[Tender Branson]

Feingold has already endorsed Obama.

Changed

[Gustaf]

Isn't Landrieu for Clinton?

[© tweed]

I thought Evan Bayh was too.

[Tender Branson]

Isn't Landrieu for Clinton?

Source ?

I thought Evan Bayh was too.

Yes, he endorsed Clinton. He is included in the Clinton column.

[© tweed]

I thought Evan Bayh was too.

Yes, he endorsed Clinton. He is included in the Clinton column.

lol, I'm just mad dumb.

[Erc]

Cantwell (although personally a Clinton backer) has made it clear that she will back the 'delegate leader' at the end of the primaries.

Barring an Obama collapse, that means she's going to be supporting Obama, and this change is now reflected in my counts.


In other news, it looks as though Obama may, after all, go 38-29 in the TX At-Large delegation.  However, with still-incomplete results and no overwhelming evidence to convince me that he will gain that extra delegate, I will (for now) be sticking with my initial estimate of 37-30.

[Erc]

I'm going to be on hiatus until April 16th, so there likely won't be any updates until that time (unless anything really earth-shattering happens, like a MI/FL resolution).

As this is a quiet time in the Presidential race, you won't be missing anything but slightly updated Superdelegate counts.


Final updates:

--The "Pelosi Club" of superdelegates has been added to Obama's column.  Those who are pledging to vote for the leader in pledged delegates are pledging to vote for Obama, even if they don't say so explicitly (I've shown before how Clinton winning in the pledged delegate count is impossible barring him dropping out of the race).

--Burnt Orange Report confirms the 37-30 split in Texas At-Large delegates.  Final confirmation of this number will have to wait for the final 7% of counties, and, more critically, the State Convention on June 7th (where, if I understand the process correctly, there are many ex officio delegates [Texas 'superdelegate equivalents,' I suppose], which could swing the final result somewhat).

[J. J.]

Pennsylvania: If Hillary wins with 57-58% of the vote it would mean 31-24 in at-large delegates. I know a couple of people have gone through the districts, so I'll try to add some average of those predictions later on. EDIT: seems like BRTD was the only one who gave a full analysis. He had Clinton +1, but looking at the results out of Wyoming and Mississippi and his idea of a 10% Clinton win I suspect he was a bit biased.  Clinton +8 for now, but I hope someone else can add something to this. 

I's expect a net gain of 10-12 in district delegates, conservatively.  Obama has four districts, max, where he can get a majority.  Liberally, Obama could have a 25 seat loss in the district seats, but that is unlikely.

[Erc]

The Popular Vote:

Much is beginning to be made of the final 'popular vote count' in the primary and caucus process.

Of course, what this actually is will be a matter of some dispute.

How do you count caucuses? 
First, there's the matter of certain states (IA / NV / WA / ME) not reporting popular vote totals.  They have reasons for doing so (the reallocation process due to viability cuts, making sure that more Democratic parts of the state are represented more fairly in the media reports, etc.), but it means we don't have accurate popular vote totals.  We can get good estimates of support for each candidate where we have good turnout reports, but it is still an estimate.  Thankfully, the margin for error here is not more than a couple thousand votes either way, so it won't affect the results.

Turnouts in each State:  [If anyone has any better numbers, it'd be much appreciated]
Iowa: ~239,000
Nevada: 117,599
Maine: 43,866
Washington: ~250,000

More fundamentally, is it proper to even count caucuses?  Most of them are rather undemocratic institutions...the set times at which they occur, the lack (in certain cases) of a secret ballot, and so forth.  Most of these probably gave an unfair bias towards Obama---there's no way that Obama would have done so well in a state like Minnesota (for example, sorry BRTD) had it been a primary.  Or, compare the two votes in Washington, where Obama did a full 16 points better in the caucus than the primary.  Of course, this advantage is somewhat offset by the extremely depressed turnout in caucuses---Obama can dominate a caucus vote, but if few people show up, his popular vote margin will be pretty minimal.  Thus, on net, the two effects may cancel out and the popular vote margin in the caucus could be rather reflective of the margin in a hypothetical primary.
However, what evidence exists (Washington, again) suggests that Obama may still have come out on top, on net.  In the caucus, Obama had ~91,000 vote lead over Clinton.  In the primary (which, despite the fact that it didn't matter one bit, still had 2.7 times the turnout of the caucus), Obama had only a 38,000 vote lead.
As a result, there is a case to be made (and Clinton will make it) that caucuses are hardly representative and should not be included in the total.

Then there's the entire MI/FL debacle.

Should one include these states in the total, despite the sanctions?  In the case of Florida, the case may be made that, since the primary didn't matter, this depressed turnout, and the election wasn't 'real' in some sense.  However, turnout in FL was incredibly high (despite the fact that one half of the contest didn't count)....larger, percentage-wise, than any of the Super Tuesday contests but CA & MA.  So it hardly seems fair to disenfranchise (in some sense) the voters of Florida.

Michigan is even more ridiculous.  Here, the turnout argument carries more weight.  Turnout was rather low for a primary state (20.2%), with the Democratic turnout much lower (by over 250,000 votes) than the Republican one.  Additionally, Obama wasn't even on the ballot, so how do you even count the state's votes?  You could give all of Uncommitted's support to Obama, but that's hardly a good reflection of reality---Edwards was still in the running at this time, and certainly would have attracted at least 8% support (no poll showed him below 10% while they were still polling him).


Not to mention the question of counting WA's caucus vs. WA's primary results---I'll be doing the former, since that is what matterred, after all.


Personally, I think the best and fairest thing we can do is count everything (& include Uncommitted MI for Obama)---Obama's unfair advantage in caucuses is cancelled out to some extent by depressed turnout, Clinton's unfair advantage in MI/FL is cancelled out to some extent by giving Edwards' MI vote to Obama.  Obama still probably comes out ahead of where he should in caucuses, and Clinton better in MI/FL, but those two should still roughly cancel.

But both campaigns may try to spin the final results, so I'll be keeping track of these permutations.

Primary Vote:
Clinton 12,502,593
Obama 12,974,497 (+471,904)

Caucus Vote: (approximate)
Clinton 388,143
Obama 699,252 (+311,109)

Michigan:
Clinton 328,309
Uncommitted 238,168  (-90,141)

Florida:
Clinton 870,986
Obama 576,214 (-294,772)

And switching from the Washington caucus to the Washington primary results subtracts 52,447 from Obama's lead.

Currently:
Best Case Obama (Primaries + Caucuses, No MI/FL, WA Caucus):
Obama +783,013

Best Case Clinton (Primaries + MI/FL, Uncommitted MI stay Uncommitted, No Caucuses, WA Primary):
Clinton +203,624
[This is, admittedly, a rather silly number, resting on entirely on not counting the Uncommitted votes in MI].

Realistic Count (Primaries + Caucuses + MI/FL, Uncommitted MI for Obama, WA Caucus):
Obama +398,100


I'll be updating this "Realistic Count" on the main page in future.  Note that there are no more caucuses (excepting Guam), so you are free to modify this number to your heart's desired scenario by adding or subtracting the appropriate set amount:
No Michigan:
Obama +90,141
No Florida:
Obama +294,772
No Caucuses:
Clinton +311,199
Uncommitted MI -> Uncommitted:
Clinton +238,168
Washington Primary:
Clinton +52,447

[Angel of Death]

I disagree with the use of the Washington primary to claim that the various biases of counting the primary votes of caucuses cancel themselves out. You have a perfectly fine model with Texas where both the primary and the caucuses counted. Use that.

[Alcon]

I disagree with the use of the Washington primary to claim that the various biases of counting the primary votes of caucuses cancel themselves out. You have a perfectly fine model with Texas where both the primary and the caucuses counted. Use that.

Some people get to double-vote?

[Angel of Death]

I disagree with the use of the Washington primary to claim that the various biases of counting the primary votes of caucuses cancel themselves out. You have a perfectly fine model with Texas where both the primary and the caucuses counted. Use that.

Some people get to double-vote?

What's with the obtuse response? I'm saying that Texas is irreplaceable when it comes to any analysis of the difference in results between primary and caucuses.