Tuesday, March 15, 2016

The Impact of Divided National Parties on Presidential Elections

The following is a guest post from Paul-Henri Gurian, associate professor in the Department of Political Science at the University of Georgia.

University of Georgia Political Science Professors Paul-Henri Gurian and Audrey Haynes, together with Drs. Nathan Burroughs, Lonna Atkeson and Damon Cann, have published a study that measures the impact of national party division on presidential elections. Their results are relevant to the current party divisions, especially within the Republican party. 

History shows that when one party is divided and the other party is united, the divided party almost always loses the presidential election. Consider, for example, the elections from 1964 through 1984; in each case the divided party lost. The study measures party division during the primaries and indicates how much the more divided party loses in the general election. 

The study shows that both national party division and divisive state primaries have significant influence on general election outcomes. But national party division has a greater and more widespread impact on the national results. A divisive state primary can have a small negative impact on that one state, but national party division can have large negative impact on the national vote. A divisive state primary rarely leads to more than a 1-2% decrease in the general election in that state. For example, Hillary Clinton received 71% of the Democratic vote in the Georgia primary while Donald Trump received 39% of the Republican vote. If nominated, Trump would lose almost 1% in the state of Georgia in the general election.

In comparison, national party division often leads to decreases of more than 3% nation-wide. For example, as of March 12, Trump has received 36.1% of the total national Republican primary vote while Clinton has received 60.0% of the Democratic vote. If these proportions remain steady for the remainder of the nomination campaign (and if these two candidates win the nominations), then Trump would lose 5.7% of the vote in the general election (compared to what he would otherwise be expected to receive). In a close election (such as 2000, 2004, 2012), 5-6% could change the outcome in terms of which party wins the presidency. 

The results of this study provide analysts with a way to anticipate the impact of each primary and, more importantly, the impact of the total primary vote on the general election results. Subtracting the percent of the Republican nominee’s total popular vote from that of the Democratic nominee and multiplying that by 0.237 indicates how much the Republican nominee is likely to lose in the November election (compared to what would otherwise be expected). The 5.7% figure calculated through March 12 can be updated as additional states hold their primaries. (The same can be done for individual state primaries by multiplying by 0.026.)  

The full study, “National Party Division and Divisive State Primaries in U.S. Presidential Elections, 1948-2012,” published by Political Behavior, is available online now (http://link.springer.com/article/10.1007/s11109-016-9332-1). Contact Dr. Gurian at PHGurian@uga.edu for further information, questions or clarification.


Monday, March 14, 2016

2016 Republican Delegate Allocation: NORTHERN MARIANA ISLANDS

This is part thirty-two of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

NORTHERN MARIANA ISLANDS

Election type: caucus
Date: March 15 
Number of delegates: 9 [6 at-large, 3 automatic]
Allocation method: winner-take-all
Threshold to qualify for delegates: n/a
2012: non-binding caucus

--
Changes since 2012
Unlike in the other nine delegate territories, the Republican Party in the Commonwealth of the Northern Mariana Islands has more clearly defined rules on the allocation and particularly the binding of its delegates to the Republican National Convention. Whereas the delegates in the American Samoa, Guam and the Virgin Islands have unbound delegations, the Republicans in the Northern Marianas will allocate their nine delegates to the winner of the territorial caucuses on March 15.1

That is not only a departure from the other territories but from how the process operated in 2012 in CNMI when the delegation headed to Tampa unbound.


Delegate allocation (at-large and automatic delegates)
There will be two votes that participants in the caucuses in the Northern Marianas will cast. The first is for presidential preference and the second to elect the six at-large delegates from the islands. That first vote binds the delegates chosen in the second vote to the winner of the caucuses.


Binding
Delegates will be bound to the winner of the Northern Marianas caucuses through the first ballot at the Republican National Convention in Cleveland. After that, the delegates are released. If the winner of the caucuses withdraws before the convention, then the delegates vote as a group on who to support as a nine member group on the first ballot. Should the first ballot be inconclusive, then the delegates are freed of their bond to the original winner or the replacement choice of the delegation and can support any candidate they prefer (as individuals).


--
State allocation rules are archived here.


--
1 Yes, the Virgin Islands would have had bound delegates had any delegates committed to candidate been elected. The rules there -- carried over from 2012 -- did not necessarily call for the binding of delegates. That is a function of an interpretation of the Virgin Islands rules by the RNC given the new rules on binding for 2016.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

2016 Republican Delegate Allocation: MISSOURI

This is part thirty-one of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

MISSOURI

Election type: primary
Date: March 15 
Number of delegates: 52 [25 at-large, 24 congressional district, 3 automatic]
Allocation method: winner-take-most/winner-take-all by congressional district (with majority winner-take-all trigger statewide for all of the delegates)
Threshold to qualify for delegates: n/a
2012: caucus

--
Changes since 2012
Change perhaps seemed inevitable in Missouri for 2016. Following a cycle in which the state legislature could not agree on legislation to shift back the date of the presidential primary in the Show-Me state, the Missouri presidential primary is back and back in compliance with national party rules on timing.

Five years ago, the Missouri legislature failed to move the primary out of February. While February primaries were compliant under both national parties' rules in 2008, they were not in 2012. This forced a lot of states to schedule later primaries. Missouri was not one of those states. In fact, the Republican-controlled Missouri legislature could not only not agree on moving the primary, it also could not agree on canceling it even after the Missouri Republican Party had opted out of the non-compliant, state-funded primary to hold mid-March caucuses to avoid losing delegates to the national convention.

But the Missouri legislature righted the primary ship in 2014, shifting the primary back to a point on the calendar comparable to the caucuses four years ago. The move pushed the primary far enough back on the calendar that the Missouri Republican Party was eligible revert to allocating all of the state's delegates to the statewide winner (as had been the case prior to 2012). However, the party opted to replace the winner-take-all plan with a winner-take-most delegate allocation plan.

However, that plan has a twist. Rather than a standard winner-take-all by congressional district plan, the Missouri Republican Party has reweighted the distribution of congressional district delegates. Most states allocate three delegates per congressional district. That is a number that is derived from the RNC apportionment formula. Each state is apportioned three delegates per congressional district, and most states that consider the congressional district vote in the allocation formula allocate those three delegates in various ways.

In Missouri, though, the winner of a congressional district will be allocated five delegates in 2016. That includes the regular three congressional district delegates, but also two of the at-large delegates. Basically, the party has raided the at-large pool to increase the power that each of the congressional districts carry in the overall allocation. That shifts 16 delegates from the at-large pool into the congressional district pool for the purposes of allocation.1

On the surface, all this means is that the congressional district wins carry a bit more weight in Missouri than in other states and that the at-large pool is decreased from 25 (plus the three automatic delegates) to 9 (plus the three automatic delegates).

Finally, the Missouri Republican Party has added as part of their plan a winner-take-all trigger. If any candidate wins a majority statewide, then that candidate is entitled to all 52 delegates at stake in the Show-Me state.


Delegate allocation (at-large and automatic delegates)
Win statewide and a candidate wins the 12 at-large and automatic delegates.


Delegate allocation (congressional district delegates)
A win in a congressional district means a candidate wins all five delegates from that district.

There is the winner-take-all trigger that could allocate all the delegates -- at-large, congressional district or otherwise -- to a majority statewide winner.


Binding
Delegates are bound to the candidates they have been allocated on the first ballot at the national convention or until released if a candidate withdraws before the the convention.

Missouri is not a state where candidates have clear input on who their delegates are. It requires organizing delegate candidates to stand for election in the caucus/convention process that culminates with congressional district conventions on April 30 and the state convention on May 20-21.



--
State allocation rules are archived here.


--
1 While two at-large delegates will be allocated in each of the congressional districts in Missouri, they will still be selected at the late May state convention with the other at-large delegates.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

2016 Republican Delegate Allocation: ILLINOIS

This is part thirty of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

ILLINOIS

Election type: primary
Date: March 15 
Number of delegates: 69 [12 at-large, 54 congressional district, 3 automatic]
Allocation method: winner-take-all (at-large/automatic), directly elected (congressional district)
Threshold to qualify for delegates: n/a
2012: loophole primary

--
Changes since 2012
The basic infrastructure of the Illinois Republican Party delegate allocation/selection process is the same in 2016 as it was in 2012. It is still a loophole primary, that primary is still on the third Tuesday in March, there are still 69 delegates at stake, and congressional district delegates are still elected directly on the presidential primary ballot.1

However, there are a few subtle and not-so-subtle differences. There are a couple of things that fall into that latter, not-so-subtle category. First, the at-large and automatic delegates -- 15 delegates total -- are allocated to the statewide winner of the primary. Those delegates were all unbound in 2012. Second, while the congressional district delegates continue to be directly elected, they are being considered bound to the candidate they aligned with when filing (or being filed) to run as a delegate candidate. In 2012, those delegates were considered unbound. And if a delegate has filed to run as an uncommitted delegate, then those delegates, if elected, would also be unbound at the 2016 convention. The affiliation with a candidate upon filing is the key.

One subtle difference between the 2012 and 2016 processes in Illinois is that for 2016 there are three delegates being elected in each of the 18 congressional districts across the Land of Lincoln. Four years ago, there were a handful of districts that elected just two congressional delegates and other, more populous districts that balanced that out by electing four congressional district delegates.2 The majority of districts still elected just three delegates in 2012, but that has been standardized for this current cycle.

The other small difference is an echo of the winner-take-all allocation of the at-large delegates discussed above. As the Illinois Republican Party suggests...


No, that is not some nod to the vote early, vote often maxim that was the hallmark of the bygone days of Chicago machine politics (though FHQ did chuckle at that segment upon reading it for that very reason). Instead, that statement is a function of how the process works in Illinois. Voters have traditionally voted for a presidential candidate and also congressional district delegates to go to the national convention on the presidential primary ballot. Two types of votes.

However, in the past, that presidential preference vote was largely meaningless. It has never really had a direct bearing on how the delegates would be selected at the state convention.3 This time it does. The allocation of delegate slots to candidates is a direct reflection of who has won the statewide vote in the primary. The results are binding.


Delegate allocation (at-large and automatic delegates)
At-large delegates will continue to be selected at the May state convention, but will be bound based on the statewide vote in the presidential primary. And those 15 delegates will be bound to the winner on top of that.

Nine at-large delegates and nine at-large alternates will be chosen at the Illinois state convention. Additionally, the national committeeman and national committeewoman will be elected at that convention. Participants in that state convention, then, will select 11 delegates and 9 alternates to fill the slots allocated to the statewide winner in the March 15 primary.

The state party chair position is not elected at the state convention, so the current chair will ultimately serve as an automatic delegate to the national convention. That is the only delegate not elected as part of the 2016 process. All three of the automatic delegates -- the state party chair, the national committeeman and the national committeewoman -- all serve as delegates with no alternates.


Delegate allocation (congressional district delegates)
The nature of the loophole primary -- the direct election of congressional district delegates -- as FHQ described in the context of the 2012 Illinois primary, is that the statewide winner usually ends up with a disproportionate share of the delegates. In other words, the loophole primary historically has been neither truly proportional nor truly winner-take-all. The allocation tends to end up somewhere in between with the winner taking a greater share of delegates than their share of the statewide vote.

That pattern may or may not hold in a more competitive, multi-candidate race. Trump supporters would theoretically vote for Trump and Trump delegates. All of the other voters in the Not Trump category may find it difficult to choose which other candidates delegates to support. Barring any clear direction there, the vote for Not Trump congressional district delegates will tend to be diluted as compared to Trump's. For example, Cruz supporters may not have as clear an indication that they need to support Kasich or Rubio delegates in a district where Cruz may be at a disadvantage. That is a long way of saying that there is an organization hurdle that the Not Trumps have to overcome in Illinois with which the Trump campaign is not faced.


Binding
The Illinois Republican Party rules bind the different types of delegates in a different manner. Since the congressional district delegates are directly elected (and bound to a candidate with whom they have aligned if they have aligned), those elected delegates are bound until released by the candidates. However, the at-large and automatic delegates are bound to the statewide winner through the first ballot at the national convention. If a candidate formally withdraws before the convention and has any at-large or automatic delegates, then those delegates would be released at the point of withdrawal.


--
State allocation rules are archived here.


--
1 These delegate candidates will continue to appear on the ballot with the name of the candidate to whom they have committed list alongside. Voters know that they are voting for a Trump delegate candidate or a Cruz delegate candidate, etc.

2 Look for the red check marks for an indication of the two and four delegate districts at that AP link.

3 It has helped in most past cycles that the race was decided by the time it got to Illinois or the winner of the primary went on to win the nomination.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Sunday, March 13, 2016

2016 Republican Delegate Allocation: FLORIDA

This is part twenty-nine of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

FLORIDA

Election type: primary
Date: March 15 
Number of delegates: 99 [15 at-large, 81 congressional district, 3 automatic]
Allocation method: winner-take-all
Threshold to qualify for delegates: n/a
2012: winner-take-all primary

--
Changes since 2012
The one change that has affected the delegate allocation in Florida the most for 2016 when compared to 2012 (or 2008 for that matter) is that the presidential primary is scheduled in a compliant position on the calendar. Unlike the last two cycles, the Florida presidential primary is in mid-March 2016 rather than in late January as was the case in 2008 and 2012. Those state-level decisions regarding the scheduling of the Florida contest five and nine years ago had consequences. On the Republican side, the January primary date carried with it a penalty cutting the size of the Florida delegation in half.

Florida was early then, but there was a price to pay for that action.

However, the one silver lining was always that, reduced delegation or not, Florida Republicans used a truly winner-take-all allocation of their delegates. Even at half strength in 2008 and 2012, the Florida Republican primary delivered and early at least a +50 in the delegate count to both John McCain and Mitt Romney. In each case, the winner's delegate haul coming out of Florida was basically the seed money for a delegate lead that neither McCain nor Romney ever relinquished.

Florida retains that winner-take-all allocation method in its party rules for the 2016 cycle, but the contest comes six weeks after Iowa rather than four as in 2008 and 2012. No, two extra weeks does not sound like a big difference. It is not. But -- and this oversimplifying things a bit -- it is the difference between Florida being the fourth nominating contest in 2012 and being tied for/in the 29th position in 2016.

Florida, then, is a much later delegate boost for candidates in 2016 than in the previous two cycles.


Delegate allocation (at-large, congressional district and automatic delegates)
This one is pretty simple; a nice departure from the varied rounding rules that have dominated the discussions of states within the early proportionality window in the Republican process. If a candidate wins the (statewide) Florida primary -- even if by just one vote -- then that candidate receives all 99 delegates from the Sunshine state.

There is no rounding and no split of the delegates. A candidate very simply leaves Florida with a little more than eight percent of the 1237 delegates necessary to clinch the 2016 Republican nomination.


Binding
The 99 delegates allocated to the winner of the March 15 presidential primary are bound to that candidate through the first three ballots at the Republican National Committee according the Republican Party of Florida rules. That is the longest hold on the delegates at the convention of any state to have held a nominating contest thus far in 2016.

The selection process for the delegates to fill those allocated slots is also of note. Within a week of the Florida primary candidates -- all of them and not just the winner -- are to have submitted to the state party a list of potential delegates and alternates. It is from those lists that the delegates are chosen at a meeting of the party called by the state party chair. Congressional caucuses comprised of the state committeemen and committeewomen and county party chairs from that district elect the three congressional district delegates while the remaining at-large delegates are elected by the Executive Board of the Republican Party of Florida.

That process does not guarantee that the winner will have its supporters filling all 99 slots. And the motivation of the party in that instance is not necessary to help or hurt the winner. Rather, the party is often motivated to insert party regulars (including elected officials)and/or those who donate time and money to the state party in those delegate slots as a reward for that service. But again, regardless of their background, those delegates are bound to the winner of the Florida primary for three ballots at the national convention.


--
State allocation rules are archived here.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Saturday, March 12, 2016

2016 Republican Delegate Allocation: WASHINGTON, DISTRICT OF COLUMBIA

This is part twenty-eight of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

WASHINGTON, DISTRICT OF COLUMBIA

Election type: convention
Date: March 12 
Number of delegates: 19 [16 at-large, 3 automatic]
Allocation method: proportional (with majority winner-take-all trigger)
Threshold to qualify for delegates: 15% (districtwide)
2012: winner-take-all primary

--
Changes since 2012
Quite a bit has changed with the way Republicans in the District of Columbia will allocate delegates in 2016 as compared to 2012. First, the District government shifted back the primary from April to June. The second Tuesday in June on which the primary falls in 2016 is outside of the window in which nominating contests can occur under Republican National Committee rules.

Faced with sanctions from the national party, DC Republicans opted out of the June primary election, replacing it with an earlier convention to select, allocate and bind delegates to the national convention. Curiously, the party scheduled that convention just inside the proportionality window on the tail end. That normally is of little consequence, but in this case it meant the DCGOP giving up its traditional winner-take-all method of delegate allocation. Instead of the party allocating all 19 delegates to the winner of the primary as in the past, in 2016, those delegates will be proportionally allocated. That cuts into the already small power and influence of the delegation. 19 delegates bound as a bloc is more beneficial to a candidate and the party is more meaningful than a proportional split of those 19 delegates.


Thresholds
To qualify for a proportional share of those 19 delegates, under DCGOP rules, a candidate must first win at least 15 percent in the presidential preference vote at the district convention. Should no candidate reach the 15 percent barrier, then the threshold is lowered to ten percent. There is an additional contingency in place if no candidate reaches ten percent -- dropping the threshold to eight percent -- but with a winnowed field, it seems likely that those alternate thresholds are little more than insurance.

Additionally, there is nothing in the rules of the Washington, DC Republican Party delegate selection plan prohibiting a backdoor winner-take-all outcome if only one candidate surpasses whatever defined threshold above is being used. Yet, a winnowed field makes it less likely that any candidate will take a backdoor to all 19 delegates.

And while the timing of the convention forced the party to abandon its past winner-take-all allocation method, such an option -- some candidate winning all 19 delegates -- is still on the table, but only if a candidate wins a majority of the vote districtwide. As the field of candidates shrinks, the odds of only one candidate reaching the 15 percent threshold decreases, but the chances of the winner-take-all trigger being tripped increase.


Delegate allocation (at-large and automatic delegates)
The allocation calculation the Republican Party is using in the district divides the candidate share of the vote by the total qualifying vote; just those over 15 percent (or whatever threshold is being used). The smaller the field is, the less likely it is that any candidate will receive anything more than a proportional share of the 19 delegates. But if any candidate or candidates fall(s) below that threshold, the delegate share for the qualifying candidates increases beyond a simple proportional share. All that really means is that as the unqualified share of the vote increases, the share of the delegates allocated to those above the threshold increases as well.

Candidates have to win at least 15 percent of the vote to qualify. One cannot round up to that threshold from 14.8 percent, for example.

Fractional delegates are rounded to the nearest whole number. If the allocation results in an overallocation of delegates, then the superfluous delegate is removed from the total of the candidate furthest from the rounding threshold. In the event of an under-allocation, an extra delegate -- one to get the total to 19 delegates allocated -- is awarded to the candidate closest to the round threshold. Compared to other states, these rounding rules do not by default favor those at the top of the vote order in the preference vote.


Binding
The binding rules hold delegates in place -- bound to a particular candidate -- through the first ballot at the national convention. There are only a couple of exceptions to that rule. For starters, when candidates withdraw from the race, any delegates allocated to that candidate are released and immediately unbound for the convention. However, if only one candidate's name is placed in nomination at the national convention, then the DC delegates are bound to vote as a bloc for that candidate.

Delegates will not only be allocated and bound at the March 12 convention, but they will be selected as well. In a vote similar to the one called for in the Virgin Islands rules, the top 16 votegetters become national convention delegates and the next 16 in the order are the alternates. Candidates for delegate file to run on their own, but can accept the endorsement the candidates and their campaigns along the way. The delegate candidate either affirms that endorsement or does not. In the case of the former, the delegate is listed on the convention ballot with the candidate they are supporting listed with the delegate.

Even if an endorsement is not accepted by the delegate candidate or is not made in the first place, a delegate still has to sign an affidavit that he or she will support the candidate to whom they are bound by the results of the convention. The endorsement part of that would tend to help the candidates to handpick delegates or at the very least indirectly influence who their delegates are. But in the end, the delegates would have accept the endorsement and be elected in the delegate preference vote. That means that there could emerge from the DCGOP convention delegates who prefer another candidate, but are bound to another.



--
State allocation rules are archived here.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Friday, March 11, 2016

2016 vs. 2012 and the Republican Delegate Count

Now that the Republican presidential primary process has entered March and hit hyperdrive, many are beginning to more closely examine the rules changes the Republican National Committee (and the Republican National Convention in Tampa) made for the 2016 cycle. At the core of that is a simple question: How have the rules impacted the progress of the race?

The answer to that simple question, however, is not so simple. Some have blamed the proportionality requirement at the beginning of the March part of the primary calendar. Others have pointed the finger at the newly compressed 2016 primary calendar for the results in the contests to date. The problem is that both of those explanations miss the mark by failing to take a deeper look at what has really changed with the rules for 2016 and how that has affected the actual delegate count. Both arguments basically hide behind the complexity of both changes without really offering an adequate answer to the original question.

There are at least two other explanations that better explain the differences in the delegate counts at similar points in the 2012 and 2016 cycles.

1) Texas
The Texas primary -- and its 155 delegates -- were in May in 2012. Think about that. That is 155 delegates that were virtually at the end of the 2012 process. In fact, it was those Texas delegates that pushed Mitt Romney past the 1144 delegates he needed to clinch the Republican nomination four years ago.

But the Texas primary was only scheduled for May because a redistricting dispute in the courts forced its delay. Originally, it was planned -- by state law -- for the first Tuesday in March. Just like this cycle (and every other one from 2008 back to 1988).1

With Texas back to normal in March for the 2016 cycle, those 155 delegates -- 12.5% of the number of delegates required to clinch the nomination -- ended up at a considerably earlier point on the calendar than had been the case since the 1988 Southern Super Tuesday. Not only was the Texas primary earlier for 2016, but the Lone Star state had a favorite son vying for the nomination. Without those 104 delegates, that favorite son -- Ted Cruz -- would not be in nearly the favorable position in the delegate count as he is at this point in early March on the calendar.

Cruz would lose 56 delegates to Donald Trump. That is the surplus he had over the real estate tycoon in the Texas delegate count. Furthermore, he would lose the 101 delegate advantage he had over Marco Rubio leaving Texas. Without an early Texas primary win, Cruz would not be as close to Trump in the delegate count and would be closer to Rubio and third place in the delegate count than Trump in first. In other words, Cruz would look a lot more like Gingrich and Santorum did relative to Romney in 2012.

It just cannot be understated how important that Texas win was for Cruz. And no, the position of the Texas primary on the 2016 calendar had nothing to do with the Republican rules changes in Tampa and thereafter.


2) Unbound delegates
While the Texas factor is not a rules-based change, there is one rules change that to this point has taken a back seat to other explanations; those arguing that the course of the 2016 Republican presidential nomination race is a function of proportional rules and/or a compressed calendar. The focus on those two changes is mostly misguided as the national party rules changes -- particularly with regard to proportional delegate allocation -- did not really yield that much change in the state-level rules. It has not to this point anyway.

The one thing that many are missing that has actually more directly affected (made things appear more competitive) the current delegate count relative to the one four years ago, is the new binding requirement the RNC instituted for the 2016 cycle. Gone is the fraction of automatic delegates who were more like superdelegates four years ago. Gone are the fantasy delegates from all those non-binding caucuses. Those delegates are mostly bound or will be bound in 2016. Iowa, Minnesota and Maine (and eventually Washington and Missouri) were all non-binding in 2012. Not in 2016. In the case of those first three, the delegates were allocated proportionally with either no or a very low threshold. That is making the delegate count more competitive. Delegates were allocated to a larger number of candidates in 2016 rather than being unbound as they were in 2012.

That is not necessarily affecting the gap between candidates in the current delegate count, but it is providing more delegates to more candidates instead of no candidates.

--
What effect have those two changes had on the delegate count cycle over cycle from 2012 to 2016?

If one backs out the Texas delegates and the unbound delegates (based on the formerly non-binding caucus state that have conducted caucuses at this time) from the 2016 count, 2016 looks even more like 2012. Cruz is, perhaps, a stronger version of Santorum (each won/has won multiple states) and Rubio is a weaker Gingrich (still each has/had two wins). Kasich stands in as Ron Paul; not winning contests, but winning delegates. Here is what that comparison looks like:

2016 (25 states)
Trump: 391
Cruz: 221
Rubio: 126
Kasich: 51
Unbound: 17


2012 (26 binding states, 32 total)
Romney: 454 R
Santorum: 172 S
Gingrich: 138 G
Paul: 27 P
Unbound/Unpledged: 2922

Again, there has been no accounting for the calendar changes -- other than a non-rules-based shift of the Texas primary -- or proportionality rules in this. In this exercise, the Texas delegates have been removed from the 2016 total as have the formerly unbound delegates in now-binding Iowa, Maine and Minnesota. The picture that leaves is one where the delegate leaders are within about 50 delegates of each other. Perhaps that is attributable to a newly compressed calendar and/or (an admittedly smaller) proportionality window in 2016. But a strong argument could also be made that the differences in the leaders' totals at similar points on the calendars in 2012 and 2016 is explained by a weaker frontrunner in 2016 than in 2012. That seems to be demonstrated by what looks like a basically 50 delegate shift between first and second place in 2012 versus 2016.

This is a surface level exercise, but it is quite suggestive.

Blame the rules?

Not really. The rules in 2016 -- as has been the case for the rules of the Republican process in the past -- are still designed to aid frontrunners/winners; to ease the way to a presumptive nominee. That is still happening.

The bottom line is that the rules changes have had an effect, but not in the way that many think. Maybe the finger should be pointed not at the compressed calendar and the proportional rules, but somewhere else instead.

--
1 Texas did shift from the second week in March to the first week in March in its law ahead of the 2004 cycle, but that change was delayed until 2008. From 1988-2004, then, Texas was on the second week in March on the calendar.

2 That this number is so much higher is a function of there being six more contests in 2012 than in 2016. Those non-binding contests drove up the total number of delegates.


Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Thursday, March 10, 2016

2016 Republican Delegate Allocation: VIRGIN ISLANDS

This is part twenty-seven of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

VIRGIN ISLANDS

Election type: caucus
Date: March 10 
Number of delegates: 9 [6 at-large, 3 automatic]
Allocation method: delegates directly elected
Threshold to qualify for delegates: none
2012: caucus

--
Changes since 2012
Literally none.

However, there is a story behind that. The process book the Republican National Committee released on October 2 -- the day after the deadline for states and territories to submit their plans to the national party -- indicated that the Republican Party in the Virgin Islands would allocate all nine of its delegates in a winner-take-all fashion based on the vote in March 19 caucuses.

But none of that was official (...at least not the part concerning the Virgin Islands). Instead the section on the Virgin Islands was based on a delegate selection plan that had yet to be submitted to the RNC. And, in fact, the VIGOP missed the October 1 submission deadline. In such situations, the state/territory party in violation is forced to use the rules that governed the selection/allocation process from the previous cycle.

That is why there are Thursday caucuses. The 2012 VIGOP delegate allocation rules called for a (Saturday) March 10 series of caucuses to directly elect six at-large delegates. The date remains the same, but the day is different. And that is essentially the only change to how Republicans on the Islands are electing delegates.


Thresholds
As the six at-large delegates are directly elected on the caucus ballot, there are no thresholds to qualify for delegates.


Delegate allocation (at-large delegates)
While there are no thresholds, candidates being "allocated" delegates depends on how many candidate-affiliated delegate candidates filed or were filed with the Virgin Islands Republican Party. In other words, a candidate cannot win all six at-large delegates unless there are six or more delegate candidates on the ballot who are aligned with a particular candidate.

Of the active candidates, only Ted Cruz has at least six delegates running on the caucus ballot. Marco Rubio and Donald Trump have only three delegate candidates running on their behalf while John Kasich has no delegate candidates pledged to him running on the ballot. What that means is that only Cruz can sweep the at-large delegates. The best Rubio or Trump can hope for is half the at-large delegates (and a third of the full nine member delegation). Kasich very simply is out of the running in the Virgin Islands.

It should also be noted that delegate candidates on the ballot can file and run as uncommitted to candidates. There are 20 such delegate candidates on the March 10 ballot. That remains something of a wildcard in all of this. Additionally, four of those uncommitted delegate candidates have had their residency on the islands called into question by the Department of Elections (not the territorial party) in the territory.

UPDATE: Those four uncommitted delegate candidates have cleared the residency hurdle for the time being.


Binding
If delegate candidates have affiliated with a candidate for the Republican nomination, then by rule of the VIGOP, the delegate, if elected, is bound to that candidate through the first ballot at the national convention. Should the candidate to whom those candidates are bound withdraw, then those delegates are released and treated as uncommitted.

As the 2012 rules have carried over to 2016, the three automatic/party delegates -- the territorial party chair, the national committeeman and national committeewoman -- are all unbound. Only the at-large delegates are bound and then only if they are aligned with a candidate. Yes, in other states, the automatic delegates have been treated as at-large delegates where the state/territory rules are not clear. However, the three party delegates will not appear on the ballot for electing delegates and there is not preference vote -- just the vote for the six at-large delegates -- to bind them.


--
State allocation rules are archived here.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Tuesday, March 8, 2016

2016 Republican Delegate Allocation: MISSISSIPPI

This is part twenty-six of a series of posts that will examine the Republican delegate allocation rules by state. The main goal of this exercise is to assess the rules for 2016 -- especially relative to 2012 -- in order to gauge the potential impact the changes to the rules along the winner-take-all/proportionality spectrum may have on the race for the Republican nomination. For this cycle the RNC recalibrated its rules, cutting the proportionality window in half (March 1-14), but tightening its definition of proportionality as well. While those alterations will trigger subtle changes in reaction at the state level, other rules changes -- particularly the new binding requirement placed on state parties -- will be more noticeable. 

MISSISSIPPI

Election type: primary
Date: March 8
Number of delegates: 40 [25 at-large, 12 congressional district, 3 automatic]
Allocation method: proportional
Threshold to qualify for delegates: 15% (statewide)1
2012: proportional primary

--
Changes since 2012
While there was an attempt to shift the Mississippi presidential primary up a week to join other southern states as part of the SEC primary, it remained just that: an attempt. Neighboring Alabama moved and left Mississippi as the lone southern state on the calendar between the two dates -- March 1 and March 15 -- with the most delegates at stake this cycle.

The Mississippi Republican Party, then, has the same primary date as 2012 -- second Tuesday in March -- but has also mostly carried over its rules from the 2012 cycle. The allocation is proportional and still split across congressional districts and statewide. There are some subtle changes, but those will be dealt with below.


Thresholds
The first of those subtle alterations has to do with the thresholds to qualify for delegates. Statewide, a candidate must reach 15 percent of the vote to receive any of the at-large and automatic delegates. A new wrinkle for 2016 is that the party has inserted a lower threshold of 10% as a fall-back option should no candidate surpass 15 percent of the vote. This is not an uncommon response on the state party level given how large the field of Republican candidates was when rules were being finalized in the late summer/early fall of 2015. But as a field winnows, the necessity of that fall-back, lower threshold decreases.

There is no winner-take-all trigger in Mississippi for a candidate who wins a majority of the statewide vote in the primary. However, there also is no prohibition on a backdoor winner-take-all scenario. If only one candidate receives more than 15% of the vote, then that candidate would claim all 28 at-large and automatic delegates. The usual winnowing caveats apply. As the field shrinks, so too do the odds that only one candidate will receive more than 15% of the vote.

At the congressional district level, there just one threshold; a winner-take-all trigger. Should a candidate receive more than 50% of the vote in any of Mississippi's four congressional district, then that candidate is entitled to all three of the delegates from that district.


Delegate allocation (at-large and automatic delegates)
There is not a lot of intrigue here. Again, candidates above 15% of the vote will receive a proportional share of the 28 at-large and automatic delegates based on the statewide vote in the primary. The allocation equation divides the candidates' shares of the statewide vote by the qualifying vote (the votes of just those over the threshold). If all candidates reach 15 percent, then the allocation is roughly proportional to the candidate's statewide share of the vote. Yet, as the share of the vote outside of the qualifying vote grows, the shares of the delegates for the qualifier increase as well.

For example, if one candidate misses the cut with 14 percent of the vote, that 14 percent of the delegates is distributed to the candidates who qualified. That would give them a share of the statewide delegates that is greater than their raw share of the statewide vote.

The rounding rules are fairly simplistic. Any fractional at-large and automatic delegates are rounded to the nearest whole number. If that results in an overallocation of delegates, then the superfluous delegates are subtracted from the last qualifier -- the one with the fewest votes -- to square the delegate distribution/count. In the event that the rounding results in an under-allocation, then those delegates would become unbound. This is counter to how a number of other states have handled similar, under-allocation situations. The norm is that when fewer delegates are allocated than a state has that the under-allocated delegate is added to the total of the top votegetter (see Michigan for example). That is not the case in Mississippi. That delegate (or delegates) would be unbound.


Delegate allocation (congressional district delegates)
Some states have had a number of ways of allocating their congressional district delegates. As FHQ has been fond of saying, though, there are only so many ways to allocate three delegates. The Mississippi Republican Party has kept the process pretty basic. If no one wins a majority of the vote in a congressional district, then the winner is allocated two delegates and the runner-up one. There is no threshold to qualify. There is, however, a winner-take-all threshold. Should a candidate win a majority, then that candidate would be allocated all three delegates from that district.

All that matters is whether someone wins a majority and, barring that, how the candidates place. Being in the top two is of the utmost importance. Third place (or lower) on the congressional district level is no place to be because such candidate would be left out of the delegates.


Binding
The Mississippi Republican Party requires delegate candidates to file with the party and affiliate with a candidate in the process. In addition, the candidates (or their representatives) have some input at the state convention over who ends up filling their allocated delegate slots. The campaigns have more influence in getting "their guys" through to the national convention as compared to other states.  Due to that connection -- based on the filing requirements and the candidates' say -- the Mississippi delegates are bound until released. If no one drops out, then the delegates remain bound. There is no limit to how long this bond lasts in terms of a number of ballots. It depends entirely on whether the candidate or their campaign releases the delegates from the binding.


--
State allocation rules are archived here.


--
1 If no candidate reaches the 15 percent threshold, then it is lowered to 10 percent.


--
Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.

Monday, March 7, 2016

A Delegate Count Reconciliation Quickie

FHQ saw Eric Ostermeier's post on the varying delegate counts this morning, and since I have looked at these numbers until I ended up cross-eyed each night for the last week, I will make a few comments about what is driving the differences across the major trackers.

First, it should be said that the combination of simpler rules and clearer results drove the agreement across all outlets. In most of those states -- six out of nine -- the delegates were pooled rather than split across congressional districts. Where they were not -- South Carolina, Alabama and Minnesota -- the results were clear. That clarity means that the rounding rules were never all that consequential or at the very least were easier to decipher in allocating delegates.

That suggests that in the states where there is disagreement among the various delegate trackers partially boils down to the results on the congressional district level. But there are some tricky rounding and other rules that are likely the culprits for the differences.

Let's look at the problem states one by one.

New Hampshire
Look, this one is a done deal for now. It may come back up at the convention, but the New Hampshire Republican Party has already certified the results and the delegate count. There is some question about whether the NHGOP or the secretary of state got the rounding wrong, but again this is a settled issue until the convention. CBS needs to catch up with everyone else. It will eventually go the other way if Rubio is still in the race because the rounding rules were not followed properly.

Nevada
Again, CBS is wrong on this one. It looks as if they are treating Nevada as a state with a qualifying threshold that it does not have. Nobody under 20% was allocated any delegates by CBS. That includes Ben Carson, who won two delegates also.

Arkansas
The Natural state is the first to demonstrate the congressional district discrepancy problem. AR-3 is a district that looks too close to call. Some of these outlets are granting Cruz the runner-up congressional district there. Others are not. The fight for second place there is really close between Cruz and Rubio. Some have called it. Others have not. FHQ is in the latter category.

Georgia
Here is another congressional district issue. Second place is close between Cruz and Rubio in both GA-4 and GA-13. Some have called that Cruz while other outlets have not. The Trump issue, given the range in the various counts, has to include some issues with whether or not some outlets are granting the winner all three automatic delegates -- as called for Georgia Republican Party rules -- or if they are allocating them as part of the at-large pool of delegates.

Oklahoma
This is an easy one. The congressional district results clearly gave one delegate to Cruz, Rubio and Trump in each of the five districts. That means the issue was with the at-large delegate allocation. Some outlets picked up on the fact that the allocation equation divides by the total number of votes instead of the qualifying candidates' share of the vote. The former leaves three uncommitted delegates  that some have missed out on. Any outlet that has Cruz over 15 delegates missed this. Period.

Tennessee
TN-9 is the issue here. Some have called this one for Trump (2 delegates) with Rubio second (1 delegate). Others are still waiting on results to clarify in that district.

Texas
This one is not that hard. The Texas Secretary of State has both the results -- with 100% of precincts reporting -- statewide and by congressional district. There really should not be any issues in the Lone Star state. As FHQ has pointed out, there does seem to be a big difference between the TXSoS count and the AP count in TX-33. Either Cruz got two delegates and Trump one (TXSoS) or Rubio won all three (AP). FHQ has deferred to the state of Texas on this one. But that one district accounts for most of the differences. in the counts.

Vermont
The AP has updated their count and is consistent with the others with an 8 to 8 divide between Trump and Kasich. CBS probably needs to just update because 100% of returns are in and Rubio is still under the 20% threshold. He would have claimed four delegates had he risen above that level, but that does not look to be on the horizon.

Kansas
This is another easy one. The outlets that do not have Cruz at 24 delegates missed that all fractional delegates round up sequentially from top of the voting order to the last qualifier in Kansas; even fractions below .5. Cruz rounded up to 24. That left one less delegate for Kasich when the rounding process got to him as the last qualifier.

Kentucky
In Kentucky, the differences are based on how one reads the rounding rules. Under Kentucky Republican Party rules, if there is an overallocation of delegates -- as there was Saturday night -- then the candidate furthest from the rounding threshold loses a delegate. Normally, that threshold would be .5, but all four qualifying candidates had remainders above .5. They all rounded up triggering an overallocation. To take one delegate away from the candidate furthest from a .5 threshold would mean Trump would lose a delegate as he was more than .3 above .5. However, since all four candidates were over .5, that made .00 the threshold for any of the candidates to round up. That took the delegate from Rubio. Those taking Kasich down from seven to six made the mistake of doing the allocation sequentially and leaving Kasich with the leftovers. That is wrong.

Louisiana
The results in the Pelican state seem to settled. The LAGOP released the delegate count there late Sunday night. Trump edged Cruz in the at-large count, but Cruz equalized by winning a majority in LA-4. In the remaining five districts, Trump, Cruz and Rubio all received one delegate apiece. That leaves Trump 18, Cruz 18 and Rubio 5 in Louisiana. That Rubio received no delegates in some counts for Louisiana was entirely due to unclear rules regarding the allocation of congressional district delegates. It was unclear if Rubio needed 16.67% of the vote to round up to one delegate or whether it was enough for him to simply place third there.


The FHQ delegate count can be found here.

Happy counting, everyone.


Follow FHQ on TwitterGoogle+ and Facebook or subscribe by Email.