The Delegate Fight: 2016

[Bacon King]

Speaking of the GOP party leader delegates, will you be keeping track of their public endorsements (in states that don't bind them to the public vote) like you are for DNC superdelegates?

Per Wikipedia's endorsement page, Christie and Rubio have one each; Huckabee had one before dropping out

[Erc]

More detailed Iowa results:

Going through and analyzing the results on a county-by-county level, you get the following (approximate) results:

Clinton 703 - Sanders 696 - O'Malley 7

These numbers may strike some as odd.  O'Malley got essentially wiped out due to the 15% threshold; how is there any way he could get any delegates in the next round?

The answer is tactical voting.  In small counties that send only 3 or 5 delegates to the state convention, it may be worth it for one candidate to send supporters over to the O'Malley camp, raising him to viability and giving him a delegate at the expense of the other candidate.  (Note that I'm assuming that O'Malley needs to have at least one delegate at the county level for this to be relevant).  Do we really expect this to happen now that O'Malley is out of the race? Possibly.  Not all of Edwards' support flaked away after he dropped out in Iowa, though all of Biden's/Richardson's did.  O'Malley does seem more similar to the latter case...

If we don't expect this sort of tactical voting to happen, what then?  Let's say all the O'Malley delegates just stay home.  Sanders gets 2 of O'Malley's delegates to Clinton's 5, and Clinton wins the state:

Clinton 708 - Sanders 698

Let's think about another scenario.  In the best case scenario for Sanders, all of O'Malley's delegates vote for Sanders.  This not only makes a difference for a few counties where tactical voting may have been important, but also in some larger counties (like Polk) where O'Malley, who never had any hope of making viability, still has enough supporters to throw a delegate or two to Sanders.

The best case scenario here for Sanders is

Sanders 706 - Clinton 700

Note, however, that 6 of those delegates are the result of coin flips, which I'm assuming here for the sake of argument that Sanders sweeps.  However, if Clinton continues her luck and wins a majority of them, she wins the state.  These numbers also assume that all the O'Malley supporters actually show up and that they actually caucus entirely for Sanders, not really a reasonable assumption.  Clinton could do slightly worse than this number suggests if there's also some tactical voting, but this is pretty much Sanders' ceiling.

A more realistic, but still Sanders-favorable scenario, has him winning 85% of O'Malley supporters and splitting the coin tosses evenly.  In this case, the count is

Clinton 704 - O'Malley 702

There are only two counties requiring coin flips in this scenario, I've split them evenly.  Sanders would need to win both coin flips and another one at the convention itself in order to win the 23rd delegate.  12.5% chance of this, and if he loses any of the 4 delegates he's counting on to win with O'Malley support (in Crawford, Linn, Polk, and Jasper) Clinton still wins.  The highest bar to clear here is in Polk County (Des Moines), where he'd need to get around 82% of O'Malley's support to win.

Note that there are also still some clearly incomplete results: in Kossuth, Hancock, and Fremont counties at the very least.  Only Fremont could actually have a bearing on the results, as this is one of the counties where tactical voting is possible given the current results.  I'm also assuming a 3-2 split for Clinton among the tele-/satellite-caucus delegates, which seems likely but is not guaranteed.

TL;DR:  Clinton seems quite likely to have a slim plurality at the State Convention.  However, tactical voting, O'Malley supporters, differential turnout (i.e. people forgetting to show up to County Conventions), and even coin flips, could make a difference.

Sanders seems highly unlikely to win unless he gets the overwhelming majority (5/6ths or more) of O'Malley supporters and wins at least three coin flips.  As a result, I'm calling the 44th delegate for Clinton, though we may return to this after March 12.

[Erc]

Speaking of the GOP party leader delegates, will you be keeping track of their public endorsements (in states that don't bind them to the public vote) like you are for DNC superdelegates?

Per Wikipedia's endorsement page, Christie and Rubio have one each; Huckabee had one before dropping out

This is a trickier beast.  Many states don't explicitly bind the RNC members, while others do.  There's also the open question of whether national GOP rules require them to be bound in states that have any sort of primary/straw poll.  Frontloading HQ suggests, from his conversations with the national GOP, that they should be, but how this is going to be enforced is another question.

As a result, I'm going to take a purely state-by-state approach, and not incorporate endorsements of these RNC members until after their states vote (if at all).

The only RNC members guaranteed to have freedom of conscience at Cleveland are those from states that are not holding any sort of public vote (CO, WY, ND, Virgin Islands, Guam, and American Samoa).

[Erc]

More Iowa Analysis

This afternoon, the Iowa Democratic Party released the complete (well, almost) precinct-level data, letting us do a proper county-by-county analysis.

The State Convention will have 1406 delegates present, so 704 is necessary for majority and control of the 44th delegate.

At present, we can project with near-certitude that Clinton and Sanders will get at least the following vote totals:

Clinton 699 - Sanders 690

There are a couple of caveats here; this assumes that attendance at the County Conventions is perfect (even if it isn't, there is a system of alternates in place), that none of the Sanders or Clinton delegates will switch sides before March 12, and that there's no "surprise" tactical voting (i.e. half of the Sanders delegation switching to "Uncommitted" and stealing a delegate from Clinton, when Clinton could prevent this by switching a greater number to "Uncommitted").

This leaves 17 delegates remaining, of which Clinton would only need 5 to secure a majority.  Let's break down these delegates by county.

Black Hawk County

County: Sanders 222 - Clinton 196 - O'Malley 1 - Uncommitted 1
State: Sanders 36 - Clinton 32 - Unallocated 1

Clinton could pick up that last delegate if both the O'Malley and Uncommitted delegates agree to support her.

Bremer County

County: Sanders 34 - Clinton 40 - O'Malley 1
State: Sanders 5 - Clinton 6 - Unallocated 1

Clinton could pick up that last delegate if the O'Malley delegate supports her.

Butler County

County: Sanders 35 - Clinton 25
State: Sanders 3 - Clinton 2 - Unallocated 1

This one's a coin flip.  Whoever wins gets the last delegate.

Crawford County

County: Sanders 25 - Clinton 27 - O'Malley 8
State: Clinton 2 - Sanders 2 - Unallocated 1

O'Malley's best county in the state, but he still falls one short of viability.  He could pick up the last delegate if any one delegate (presumably from the Sanders camp) switches.  Alternatively, the O'Malley supporters could recaucus and give the last delegate to either Clinton or Sanders.

Decatur County

County: Sanders 33 - Clinton 35 - O'Malley 1 - Uncommitted 1
State: Sanders 1 - Clinton 1 - Unallocated 1

If the O'Malley and Uncommitted delegates switch to Sanders, he can force a coin flip for the last delegate, which would otherwise go to Clinton.

Delaware County

County: 43 Sanders - 55 Clinton - 2 O'Malley
State: 3 Sanders - 4 Clinton - 1 Unallocated

Clinton picks up the last delegate if both O'Malley delegates support her.

Fremont County

County: Sanders 9 - Clinton 8 - O'Malley 1 - Unknown 2
State: Sanders 1 - Clinton 1 - Unallocated 1

No one showed up to the caucus in Riverton precinct, so its two delegates are treated as if they don't exist.  This means that the O'Malley supporter could force a coin flip by siding with Clinton; otherwise the delegate goes to Sanders.

Henry County

County: Sanders 55 - Clinton 55
State: Sanders 3 - Clinton 3 - Unallocated 1

This one's a coin flip.

Jackson County

County: Sanders 23 - Clinton 25 - O'Malley 2
State: Sanders 4 - Clinton 4 - Unallocated 1

If both O'Malley supporters side with Sanders, they could force a coin flip; otherwise Clinton wins the delegate.

Jasper County

County: Sanders 90 - Clinton 103 - O'Malley 7
State: Sanders 8 - Clinton 9 - Unallocated

If the O'Malley camp breaks strongly for Sanders (5-2 or more) he can pick up that last delegate; otherwise Clinton does.

Keokuk County

County: Sanders 14 - Clinton 25 - O'Malley 1
State: Sanders 1 - Clinton 2 - Unallocated 1

If the O'Malley delegate sides with Sanders, it would force a coin toss for the last delegate; otherwise Clinton wins it.

Linn County

County: Sanders 180 - Clinton 163 - O'Malley 1
State: Sanders 63 - Clinton 57 - Unallocated 1

If the O'Malley delegate sides with Sanders, they pick up the last delegate; otherwise Clinton does.

Marion County

County: Sanders 48 - Clinton 50 - O'Malley 2
State: Sanders 6 - Clinton 6 - Unallocated 1

If both O'Malley delegates side with Sanders, they could force a coin flip for the last delegate; otherwise Clinton wins it.

Monroe County

County: Sanders 29 - Clinton 28 - O'Malley 8
State: Sanders 1 - Clinton 1 - Unallocated 1

Any of the three candidates could pick up the last delegate, depending on tactical voting or lack thereof.

Polk County

County: Sanders 554 - Clinton 638 - O'Malley 8
State: Sanders 105 - Clinton 121 - Unallocated 2

The O'Malley supporters will decide where the last two delegates go.

Scott County

County: Sanders 139 - Clinton 135 - O'Malley 1
State: Sanders 41 - Clinton 40 - Unallocated 1

If the O'Malley delegate sides with Clinton, she wins the last delegate; otherwise Sanders does.


Precinct CR 1 in Kossuth County, Precinct 5 in Hancock County, and Competine Township in Wapello County also did not report, but cannot make a difference in the final outcome.

Analysis to follow.

[cinyc]

Precinct CR 1 in Kossuth County, Precinct 5 in Hancock County, and Competine Township in Wapello County also did not report, but cannot make a difference in the final outcome.

It's possible that there were no Democratic caucus goers in those precincts.  They are rural, and there were barely any Republican votes tallied in all but Precinct 5 in Hancock County.  Only 5 Republican votes were tallied in CR 1, for example. 

In the Republican caucus, Indian Settlement in Tama County recorded no Republican votes according to the Microsoft map, which is also possible given how Native Americans tend to vote Democratic if they vote at all.

[Erc]

Iowa Analysis: What are Sanders' Chances?

As examined to death with exhaustive detail in the previous post, there are 17 state delegates still potentially up for grabs, if the few O'Malley and Uncommitted supporters at the County Conventions pick a side.  (If they don't pick a side, Clinton wins 708 - 698 ± 1).

Sanders needs to win 14 or more out of the 17 in order to win outright, or 13 to have it go to a coin flip.  So clearly, he needs to win a clear majority of the O'Malley / Uncommitted support.  How much, though?

I ran a few simulations to see what Sanders' probability of winning the 44th delegate are given varying probabilities of a given O'Malley/Uncommitted delegate defecting to Sanders--a value which we'll call p.  

I made a few assumptions here.  First, all O'Malley/Uncommitted delegates pick a side, rather than just refusing to re-caucus.  Second, the probability of each picking a side is independent of the others.  These are almost certainly not true, given that Sanders may try harder in some places than others, but it should give us a subjective feel for the problem.

p	Chance of Sanders Victory
50%	<0.1%
60%	0.4%
70%	3.2%
75%	8.0%
80%	16.8%
85%	30.0%
90%	46.3%
95%	64.1%
100%	77.3%

These are the most diehard O'Malley people we are talking about here.  I just don't think they break so overwhelmingly (90%+) to Sanders to give him the win.

As a result, I'm sticking with my initial call here, unless events on March 12 prove otherwise.  Keep an eye out for those 16 counties on March 12!

[Erc]

Alaska (R): March 1

Overview
28 Delegates (1.13% of total)
Closed Caucus (Binding)
28 At-Large (Proportional, 13% threshold)

Delegate Allocation

Delegates are bound according to the results of the Presidential Preference Poll held on March 1.  Like for the Democrats in Iowa and Nevada, the delegate results are based on State Delegate Equivalents, even though, in this case, the media reports the raw vote totals.  Essentially, the vote in each district is weighted by the Republican vote in 2014.  Delegates are allocated among all candidates receiving at least 13% of the state delegate equivalents in the poll, proportional to their share of the vote among candidates meeting the threshold.  Rounding is done to the nearest whole number of delegates.  Any rounding errors are given to the state winner, or taken away from the qualified candidate with the least number of votes, as appropriate.

Delegate Selection
Delegates are selected by the State Convention on April 28 - 30, the end of a caucus/convention process beginning on March 1.

Results (3/2)

Candidate	Percent	Delegates
Cruz	35.3%	12
Trump	33.3%	11
Rubio	15.7%	5

Rubio is considered to have "drop[ped] out" of the race, the above calculation is done with just Cruz and Trump, yielding

Cruz 14 - Trump 14

Note that the use of State Delegate Equivalents is crucial here; if it had been based off the raw vote, Cruz would have an additional delegate.

Withdrawal of Candidates / Brokered Convention

If a candidate drops out before the National Convention, their delegates are reallocated as if they had received no votes in the Presidential Preference Poll.  All delegates are bound on the first two ballots, as long as their candidate “maintains an active campaign.” After the second ballot, delegates bound to the candidate receiving the fewest votes on the previous ballot become unbound.

Useful Links
The Green Papers: AK
Frontloading HQ: AK
AK GOP Rules

[Erc]

Alabama (R): March 1

Overview
50 Delegates (2.02% of total)
Primary
29 At-Large (Proportional, 20% threshold; WTA if majority)
21 District (“Winner-Take-Most”)

Delegate Allocation
If any candidate receives a majority of the statewide vote in the primary, or is the only candidate to receive at least 20% of the vote, they receive all 29 At-Large delegates.  Otherwise, delegates are allocated among all candidates receiving at least 20% of the vote in the poll, proportional to their share of the vote among candidates meeting the threshold.  In the unlikely event no candidate receives 20% of the vote, delegates are allocated proportionally among all candidates.

If any candidate receives a majority of the vote in a Congressional District, or is the only candidate to receive more than 20% of the vote, they receive all 3 delegates from that CD.  Otherwise, the winner receives two delegates and the runner-up receives one.  If no candidate receives 20% of the vote, the three delegates are allocated proportionally among all candidates.

Rounding is always to the nearest whole number.  Any rounding errors are given to (or taken away from) the candidate with the most (or fewest) votes.  This “Alabama Method” of allocation and rounding is commonly used in many other states, as well.

Voters may also vote for “Uncommitted,” who is treated as a candidate for all allocation purposes.

Delegate Selection
The primary ballot also allows voters to vote on the actual delegates who will be sent to the National Convention.  Voters are only allowed to vote for delegates who correspond to the candidate (including “Uncommitted”) who they voted for in the Presidential Preference Poll.  Some candidates have incomplete slates of delegates; in the event a candidate is entitled to more delegates than he has on his slate, that slot will presumably be filled by an alternate delegate.  Such alternates are chosen by the Alabama Republican Executive Committee and must similarly pledge themselves to the candidate they are bound to support; they must also pay a $150 filing fee.

Withdrawal of Candidates / Brokered Convention
Delegates remain bound until the candidate releases them, unless two-thirds of a candidate’s Alabama delegates vote to release themselves.

Results (as of 3/2)

Rubio failed to meet the statewide threshold, so the At-Large delegates are apportioned 20 for Trump and 9 for Cruz.  Cruz fell about 325 votes short of an additional At-Large delegate.

Trump won all the Congressional Districts, as was the only candidate above threshold in CD 1 and CD 4, picking up 14 delegates.  Rubio was above threshold (and Cruz) in CD 5, picking up 1 delegate.  Cruz took second place and was above threshold in the remaining CDs, winning 4 delegates.

Trump	36
Cruz	13
Rubio	1

This count is in agreement with The Green Papers and FHQ.

In the case of Rubio, his 1 delegate is effectively released, as he can vote to release himself at any time.

Delegates

Trump (36)
[1] Jim Bonner
[2] Jim Carns
[3] Judy Carns
[4] Ed Henry
[5] Shaun McCutcheon (donated to Rubio in Jan. 2015)
[6] James Matthews
[7] Bob Baccus
[8] Perry O. Hooper
[9] Will Matthews
[10] Barry Moore
[11] Catherine Crosby Long
[12] James Henderson
[13] Chess Bedsole
[14] Dennis H. Beavers
[15] Chad Tucker
[16] Laura E. Payne
[17] W. Brent Woodall
[18] Frank H. Long, Jr.
[19] J. Holland (a presumably different Justin Holland donated to Cruz in Dec. 2015)
[20] Carmen Moore-Ziegler
[1-1] Frank Burt, Jr.
[1-2] Unfilled
[1-3] Trey Oliver
[2-1] Jeana Boggs
[2-2] Jeremy Adams
[3-1] Charles Whatley
[3-2] Kathleen Moore
[4-1] Bradley L. "Brad" Williams
[4-2] William R. Ingram
[4-3] Jesse David Ochocki
[5-1] Jim Peters
[5-2] Shannon Matthew Moore
[6-1] Joe Freeman
[6-2] Jef Freeman
[7-1] Teresa I. Beeker
[7-2] Chris Beeker III

Cruz (13)
[1] Tom Parker
[2] Chip Brown
[3] David Wilson
[4] Mo Brooks
[5] Chad Mathis
[6] Arnold Mooney
[7] Carla S. King
[8] David Pinkleton
[9] Sue Alexander
[2-1] Mike Holmes
[3-1] Tim Sprayberry
[6-1] Collin Luke
[7-1] Rich Wingo

Rubio (1)
[5-1] Dean Brandon

Note that this does not include the RNC members here; I would guess that Trump 19 & 20 and Cruz 9 are actually not delegates, but the AL GOP may have some leeway here.

Useful Links
AL GOP 2016 Presidential Preference Primary Resolution
The Green Papers: AL
Frontloading HQ: AL
Delegate Ballot Results
Delegate Candidates

[Erc]

The recent revisions to the Iowa caucus results gives a slight boost to Sanders, as a victory in Marion County is now more plausible.

This updates the table in the post above to...

p	Chance of Sanders Victory
50%	<0.1%
60%	0.6%
65%	2.0%
70%	5.1%
75%	11.9%
80%	22.8%
85%	38.7%
90%	58.2%
95%	76.2%
100%	89.2%

This increases Sanders' chances if he's already getting a vast majority of the O'Malley delegates.  I honestly think Clinton will win this delegate, but rethinking my priors here, I'm not more than 90% certain that she will, especially given the possibility for differential turnout at the County Conventions, etc.

I'm reversing my call on the 44th Iowa pledged delegate; it will stay in the Uncommitted column until March 12 at least.

[defe07]

Erc, do you have any news from the North Dakota GOP district caucuses?

[Erc]

Erc, do you have any news from the North Dakota GOP district caucuses?

Not really; this is the only article I've been able to find, and it doesn't mention the Presidential race at all.  Seems like more of the attention is on local politics, presidential preference may not be playing a huge role.

Some fun pictures on the ND GOP's twitter feed, though.

[jimrtex]

More Iowa Analysis

Black County

County: Sanders 222 - Clinton 196 - O'Malley 1 - Uncommitted 1
State: Sanders 36 - Clinton 32 - Unallocated 1

Clinton could pick up that last delegate if both the O'Malley and Uncommitted delegates agree to support her.

Presumably this is Black Hawk County

[Erc]

More Iowa Analysis

Black County

County: Sanders 222 - Clinton 196 - O'Malley 1 - Uncommitted 1
State: Sanders 36 - Clinton 32 - Unallocated 1

Clinton could pick up that last delegate if both the O'Malley and Uncommitted delegates agree to support her.

Presumably this is Black Hawk County

Indeed.  For text parsing reasons, I cut all two-word counties down to one word in my spreadsheet, and forgot to restore some of them.  At least I didn't report results from "Des County"

[Erc]

New Hampshire Delegate Outlook

Democrats

24 delegates are up for grabs: groups of 5 and 3 At Large, and 8 in each CD.

Assuming Sanders wins, he gets 5 out of the 8 At-Large delegates.

In each CD, it depends on the margin of victory.  If he wins 56.25% of the two-way vote in a CD, he wins 5 out of 8 delegates; otherwise he and Clinton split them evenly.

Republicans

20 delegates to allocate, with the threshold at 10%.

% Vote	Delegates
0-10%	0
10-12.5%	2
12.5-17.5%	3
17.5%-22.5%	4
22.5%-27.5%	5

Anyone winning more than 27.5% is likely to win the state, given the state of the field.  The winner not only gets his own share of the delegates given his vote share, but also wins any delegates left unallocated.  Essentially, he's treated as if he won the votes that went to the sub-10% candidates.

The delegate result will depend a lot on whether the massive field of candidates all breaks 10% or whether many of them are just under 10%.  In the former case, the winner (presumably Trump) does not get any appreciable bonus, whereas in the latter he may win half the delegates from the state.  This happened in 2012, where Romney won 60% of the delegation on a 39% vote share since both Santorum and Gingrich barely missed the threshold.

[Vosem]

Are 20 or 23 delegates being allocated by the NH primary? If it is just 20, then what happens to the 3 RNC members? Are they simply free to support whomever they wish?

[Erc]

Are 20 or 23 delegates being allocated by the NH primary? If it is just 20, then what happens to the 3 RNC members? Are they simply free to support whomever they wish?

So, here's the situation as I see it.

New Hampshire state law provides the rules for the allocation of delegates (10% threshold, the rounding procedure, etc.).  They also provide for the explicit choosing of delegates after the primary (candidates choose which delegates they want from the list they filed earlier).

On the Democratic side, this law is in conflict with the national Democratic party rules for delegate allocation, and is overridden by them.

On the Republican side, they are not, so the rules are enjoining.  What does this mean in practice for the 3 automatic delegates?

The New Hampshire GOP Bylaws mention them, but really say nothing except that they are supposed to stay neutral during the Primary.

The 3 automatic delegates are indeed automatic, so the procedure of candidates selecting them from their list after the primary clearly doesn't apply here, though the binding might.

The national GOP rules seem to imply that the entire delegation should be bound.  This has been confirmed by an FHQ enquiry to the national GOP, but may still be open for interpretation.

Is there any good precedent here?  In 2008 and 2012, New Hampshire, with a January primary, was penalized and forfeited its automatic party delegates.  In 2000 and previous contests, these delegates didn't exist, so the only precedent we really have is the uncontested 2004 contest.  In that contest, it seems they were not allocated automatically to Bush, but nobody cared about the minutiae then of course.  Of course, the language for binding the entire delegation did not exist in 2004, so this isn't great precedent either.

TL;DR: My reading of New Hampshire State Law and the national GOP rules would seem to imply that all 23 delegates will be bound as a result of the primary.  

In the end, I'm likely to go with whatever the Secretary of State reports, unless the NH or national GOP says otherwise.

[Erc]

Update from FHQ:  They apparently had a chat with the NH GOP, who clarified that all 23 delegates are at stake tonight.  They will be allocated proportionally as one entire pool of delegates.  (How it's decided who gets which RNC member, I don't know).

[Erc]

New Hampshire Results

Democrats

I'd need to check the breakdown by CD to make absolutely sure, but it seems overwhelmingly likely that Sanders broke the 56.25% two-way vote barrier in both CDs.  The result therefore among pledged delegates is:

Sanders 15 - Clinton 9

Republicans

The percentages seem to have stabilized, so the delegate count seems pretty clear here.

Trump 11 - Kasich 4 - Cruz 3 - Bush 3 - Rubio 2

The decision to allocate the automatic delegates based on the results of the primary helped out Kasich, Cruz, and Bush.

[Gass3268]

Good work here! Thanks for taking on this daunting task.

[The Other Castro]

This should definitely be stickied btw.

[The Free North]

This should definitely be stickied btw.

Seconded.

[Gass3268]

This should definitely be stickied btw.

Seconded.

Thirded.

[PresidentTRUMP]

This should definitely be stickied btw.

Seconded.

Thirded.

Didn't Cruz win Iowa? How did he end up with one less delegate than Trump?

[Vosem]

So, is the breakdown of delegates Trump:Rubio 11:2 or 10:3? The race is very close for the last delegate. The way I've calculated it, and what you have in this thread, is 11:2. But most of the media is 10:3...

[LLR]

You should remove the Fiorina column...