This is part forty-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.
Election type: primary
Date: May 24
Number of delegates: 44 [11 at-large, 30 congressional district, 3 automatic]
Allocation method: proportional (with winner-take-all trigger for majority support at the
congressional district level)
Threshold to qualify for delegates: 20%
2012: nonbinding caucus
Changes since 2012
Washington state joins Idaho and Missouri on the list of states that shifted from 2012 caucuses to 2016 primaries. Each has its own story, but the tale in the Evergreen state is a long one. Historically, the primary has been little used in Washington since a voter initiative added after the 1988 cycle. Washington Democrats have never utilized the election as a means of allocating national convention delegates in the three competitive cycles the party has had since 1988 and that carried over into the 2016 cycle as well.1
Evergreen state Republicans have only been marginally more active in using the primary election over the same time period. Traditionally, the party has allocated about half of its delegates through the primary while the other half were awarded through a caucus/convention process.
The story of the Washington primary, then, is one of a contest that has traditionally been underemployed by the state parties. Given that reality, in less competitive cycles, there has existed at least some motivation to cancel the primary in order to save the state money. That was the case in 2004 when only Democrats had an active nomination race and were not going to use the primary anyway, and again in 2012 when the combination of a budget shortfall and the Washington State Republican Party only half using the primary led to cancelations.
But those cancelations were always only temporary, lasting just the one cycle. That brought the primary back in the subsequent presidential election years. And that all set the baseline to which a number of changes the WSRP has instituted for 2016 can be compared.
Yes, the 2016 primary replaced the 2012 caucuses, but that came about because Washington Republicans narrowly held the state Senate after the 2014 elections. That allowed them to stop any efforts to cancel the primary (which they did). However, holding only the state Senate meant Republicans in the legislature faced a partisan roadblock in moving the primary to an earlier date. Plans to for a coordinated March 8 primary with Idaho were scuttled by Democrats on several occasions.
Once the primary scheduling had finally resolved itself in August 2015, Washington Republicans also opted to use the primary to allocate all of their convention delegates, breaking with the half and half approach that had been the custom. All told, Washington Republicans shifted from a non-binding, early March caucus in 2012 to a binding, late May primary in 2016.
In trading the non-binding caucuses for a binding primary, the WSRP adopted a proportional method of allocation that split the awarding of delegates across both congressional districts and the state. Part of that allocation plan entails a qualifying threshold. To win any delegates in the Washington primary a candidate must win at least 20 percent of the vote -- the maximum allowable threshold under RNC rules -- within a congressional district or statewide. Those who fail to reach that threshold fail to win any delegates. This is not unlike a number of the states that held contests in the proportionality window; particularly those that were part of the SEC primary on March 1.
Additionally, there is a winner-take-all threshold that applies. However, that only applies in the case of the allocation of congressional district delegates. The candidate who wins a majority of the delegates within any of the ten congressional districts in Washington state, wins all three delegates from that district.
But again, this threshold is only applicable in the congressional districts. There is no winner-take-all threshold for the 14 at-large and automatic delegates. As such, that pool of delegates is always proportionally allocated to candidates with more than 20 percent of the vote statewide. This is is a setup unique to Washington. In most cases where there is a winner-take-all trigger, it separately affects both the at-large and congressional district delegate allocation. In other instances, a statewide majority supersedes the separate allocation, awarding all delegates, regardless of distinction, to the majority winner.
Delegate allocation (at-large and automatic delegates)
If the unevenly applied winner-take-all trigger was not enough, there are additional wrinkles to the allocation of the at-large and automatic delegates. It is impossible under the WSRP delegate allocation plan for a candidate to win all 14 at-large and automatic delegates. Not only is there no winner-take-all trigger, but a candidate cannot even take a backdoor route to the allocation of all of those delegates by virtue of being the only qualifying candidate.
There is no backdoor winner-take-all trigger in Washington and that has everything to to do with the underlying allocation equation the Washington Republican Party is using. Rather than dividing a candidate's vote share by the total qualifying vote -- just the votes for the candidates with 20 percent or more of the statewide vote -- the WSRP is dividing by the total number of votes cast for both qualifiers and non-qualifiers alike.
This is atypical as the modal activity on the state level in 2016 has been to use the qualifying vote instead of the total vote as the denominator in the allocation equation.
Basically, the answer lies in efficiency. To use the qualifying vote is to streamline the process; to eliminate the possibility of the obvious follow up question. What do we do with these unallocated delegates? This is less of an issue when there is a small group of candidates who are qualifying. But that is a pretty limited sweet spot. In scenarios where there is a large field of candidates (and only a limited few are barely clearing the qualifying threshold) or a smaller field of candidates with a clear frontrunner (but the laggards are still winning votes), the number of unallocated delegates tends to increase.
Those are situations in which non-qualifiers are still winning a sizable chunk of the vote; leaving a corresponding bloc of delegates unallocated in the process.
Let's use the results in last week's Oregon primary to illustrate this point. Donald Trump won 67 percent of the vote in the Beaver state while Cruz and Kasich respectively took 17 percent and 16 percent of the statewide vote. If that outcome were to repeat itself statewide in the Washington primary, it would create the following byproducts: 1) Cruz and Kasich would not qualify for any of the 14 delegates but 2) Trump would not win all of them as the only qualifier.
Trump would take two-thirds of the delegates, but the remaining one-third would be unaccounted for. Of course, "unaccounted for" in this case means that those delegates are unallocated. And in Washington, the state party is treating those delegates as unbound. That one-third is about five unbound delegates.
All of this is compounded by the rounding method the WSRP is using. The basic rounding scheme is simple enough. The party is rounding to the nearest whole number: Qualifying candidates with a fractional delegate of .5 or above round up while any fraction below .5 rounds down. The complication comes in how the party is dealing with the contingencies in which rounding leads to either an over- or under-allocation of at-large and automatic delegates.
Consistent with some other states, Washington Republicans are using a furthest from the rounding threshold method in cases where superfluous delegates are allocated due to rounding. An over-allocated delegate is subtracted from the total of the candidate with a fractional delegate furthest from the rounding threshold.
If, however, there are any under-allocated delegates due to rounding, then those delegates remain unbound as opposed to being allocated to the candidate, for example, closest to the rounding threshold (as is the case in some other states).
Delegate allocation (congressional district delegates)
The picture of how the 30 congressional district delegates are allocated is a bit more defined. There is no potential for unbound delegates, but there are a number of possibilities that the WSRP has helpfully laid out for their allocation in Rule 37B of the party bylaws:
- There is a winner-take-all trigger on the congressional district level. If a candidate wins a majority of the vote in any of the 10 Washington congressional districts, then that candidate receives all three delegates from that district. The remaining scenarios are dependent on whether/if a candidate hits the 20 percent qualifying threshold.
- If no candidate hits 20 percent, then the top three votegetters in that district each are awarded one delegate.
- There technically is also a backdoor winner-take-all trigger on the congressional district level. If only one candidate clears the 20 percent threshold in one of the congressional districts, then that one qualifying candidate receives all three delegates from that district.
- If two candidate surpass 20 percent of the vote, then the allocation resembles, say, the Georgia congressional district allocation: the district winner gets two delegates and the runner up, the remainder..
- Should three candidates all qualify for delegates, then each qualifier is allocated one delegate as in Oklahoma.
- Finally, if four candidates clear 20 percent, then the top three finishers receive one delegate each.
While the 2012 caucuses in Washington were described as non-binding above, that glosses over some of the intricacies of how the Republican Party in the Evergreen state handled its delegate selection process. Delegates were bound but based not on the results of the early March caucuses in 2012. Instead delegates were bound as they have been in Colorado, the Virgin Islands or Wyoming in 2016: based on which presidential candidate the delegate candidates (selected at district and state conventions) affiliated with in filing to run for national convention delegation openings. Four years ago, Washington Republican delegates were bound for the first ballot at the national convention to the candidate they preferred when filing to run.
The first ballot is still the limit of the bond in 2016 as well. However, the selection of delegates does not have to follow the results of the primary. As is the case in most other states, the selection process is completely divorced from the allocation process. Whereas 2012 delegates from Washington were locked into to the candidate they preferred when filing to run, that is not the case in 2016. The delegates have already been selected in Washington, and 40 of the 41 selected were aligned with Ted Cruz. Rather than supporting the Texas senator -- as would have been the case four years ago -- those delegates will likely be bound to Donald Trump based on the primary results on the first and likely only ballot at the national convention.
State allocation rules are archived here.
1 Those cycles include 1992, 2000 and 2008. Yes, 2004 was a nomination that was contested on the Democratic side, but the Democratic-controlled Washington legislature postponed the primary that year.
Dayton Signs Minnesota Presidential Primary Bill
Colorado Presidential Primary Bill Dies in Committee
2016 Republican Delegate Allocation: OREGON