(None of this is an endorsement of a candidate or anyone’s ideas or policies)
So, last night Donald Trump underperformed our reduced expectations in Wisconsin. Cue the regular discussion of “is this the end of Trump” etc. Rather than address what Donald Trump would need to do to right his campaign and sail straight (we’ve already done that once, but the New York billionaire seems not to read this blog), let’s dive into what will be billed as the biggest, most complex, political spectacle in a generation (well, at least since 1976).
Here’s the scenario: June 7th comes around and the voters in the last GOP primary states go to the polls. It hits 8pm on the West Coast (11pm in DC) and polls close. “Breaking News” blares across every cable news station: “Donald Trump fails to get 1,237 delegates before the convention. Republicans to have the first open convention since 1976!”
Our current projection would have Trump finishing somewhere between 1,100 and 1,200 delegates by the time all of the voting wraps up in the states that first week of June. Certainly, we could be wrong. If Trump retools his campaign and smashes through the Northeast and wins the West Coast in a stunning fashion, the rest of this article is meaningless. However, if he does not, what happens?
Ok, if you’re generally familiar with the delegate selection process and bound versus unbound delegates, you can skip down to the chart below. Otherwise, let’s take a look at the GOP delegates and what’s really going on “behind the scenes.”
Each state makes its rules with how delegates are selected to go to the Republican National Convention (this year, in Cleveland). So far in our discussions on the blog, we’ve focused on how delegates are allocated based on vote totals. It’s the quick and dirty way to approximate how many people are going to show up in Cleveland to vote for particular candidates. You’ve read things like “winner take all,” “winner take most,” and “proportional” used to describe these methods. (Not to mention loopholes and differences between statewide and Congressional District allocations). Basically, the popular vote creates blank nametags, or empty chairs, that have to be filled in a way dictated by each state.
In Florida, for example, the 99 delegate chairs are all for Donald Trump. Per Florida’s rules, the Trump campaign had a week after the Florida primary to submit a list of 96 people who Mr. Trump wants to fill those chairs (the State GOP chairman as well as the two national committee members each get a spot). Between March 22nd and June 3rd, the names of the Trump recommendations (as well as lists submitted by every other candidate who appeared on the ballot) are circulated. Delegates are then elected off of these ballots by the Party to go to Cleveland.
Now, you might be saying (and Florida is a simpler case than other places) “whoa whoa whoa! Those rules mean that people who don’t like Trump could get to be delegates just as easily as a Trump supporter!” If that’s your thought, you’re, technically, right. In most elections, it doesn’t matter since by the time state conventions are held, the outright nominee is known. There’s one more check as well. In Florida, delegates are bound to vote (they must vote) for the candidate who won the “preference primary” (the chairs, above).
There is a caveat in the rules of nearly every state that explains a candidate who releases his delegates causes his delegates to become unbound. This is complicated by the fact that “suspending” a campaign has been generally held to be an insufficient release to unbind delegates. Unbound delegates are, with some caveats depending on the state, free to vote their conscience (even if that differs from the candidate they had been bound to on the first ballot).
Our example of Florida is the most basic example and the Floridian delegation stays bound the longest of any delegation at the convention: three ballots (With the exception of delegates from states without a clear unbinding provision. In those states, the argument may be able to be made that those delegates are permanently bound to the candidate).
Some states (like Maryland), have a two-tiered method of delegate selection: by Congressional District and statewide. In some states that have adopted this model (South Carolina), the voter preference (the “chairs”) are filled by Party members at Congressional District conventions (South Carolina’s have to be held before May 2nd) and then the at-large (statewide) delegates are selected at the State Convention (South Carolina’s is May 7th). Using the South Carolina example, Donald Trump has 50 “chairs.” Each delegate to Cleveland (again, these are folks selected by Party members at the Congressional District Convention and the Statewide Convention) must vote for Trump on the first ballot. After the first ballot, all of the South Carolina delegates become unbound.
Maryland is one of the more complex states. Maryland’s 38 delegates are apportioned “winner take all” by Congressional District and statewide (like South Carolina, above). However, Maryland lets the voters decide the Congressional District delegates to Cleveland (rather than having Congressional District conventions). For Maryland voters, delegates appear on the primary ballot with the name of the candidate they are supporting (or if they are unaffiliated). The remaining 11 (38 total minus 24 from Maryland’s 8 Congressional Districts and Maryland’s Chairman and two national committee members) are selected at-large at the Maryland State Convention on May 14th by members of the Maryland Republican Party. Adding to the complexity, Maryland’s delegates are bound for the first two ballots at the national convention unless the candidate that they are supporting receives less-than 35% on the first ballot.
What would Maryland’s delegation look like, in practice? Well, let’s look at our projection from the other day: Trump getting 29 of Maryland’s 38 delegates. This would mean that Trump “wins” Maryland’s 11 at-large delegates as well as 5 of Maryland’s Congressional Districts. Let’s assume Ted Cruz gets the other 9 delegates to keep it simple. That means that we will know on Election Day who the Congressional District delegates to the National Convention will be (n.b. there is a scenario not accounted for in the MDGOP rules if a delegate who established a preference for a candidate who did not win the majority in the district but was nonetheless selected would be bound to the vote or the preference. It seems obvious that the delegate should be bound to the vote). Then, the Trump Campaign (or its associates), as well as anyone else who gets the signature of three Republican Central Committee members from three different jurisdictions in the State, will submit nomination forms for the State Convention’s selection of 11 delegates. Then, at the state convention on May 14, the amassed State Central Committee members will vote on the slate of delegates.
Now, those 11 delegates, given the hypothetical above, will be bound for Donald Trump (or if Cruz gets a majority in the State, Ted Cruz). However, after two ballots, the 11 delegates can vote for whomever they choose. Hence, the State Convention on May 14 in Maryland (and similar state conventions with these nominating rules), become the shadow battleground for control of the national convention in the event of a brokered convention. If this election were like any other GOP Primary since 1976, the delegation is, really, just a formality. However, in the context of the growing likelihood of an open convention (especially because Maryland is holding its convention before June 7th when there will be some certainty as to whether Trump has achieved 1,237), you can expect that the State Convention will be raucous. (This isn’t just a Maryland thing– just using Maryland as the complex example, and because I live here– this is going to be happening, and has already happened, across the country).
The foregoing is the reason why campaigns and pundits are talking about “ground game” or “organization” in states. Clearly, as you can see from the Maryland example (and even moreso in South Carolina), a sizeable portion of the delegation could easily be against the candidate who “earned” the “chair.” That person would, grudgingly, vote for Trump/Cruz on the first /second ballot (third in the case of Florida), but then be free to vote for anyone under the sun on subsequent ballots.
So, what does this mean? Here’s a handy chart:
(You’ll notice that the delegations from Ohio, Guam, and Washington State are “unclear.” This is because their rules do not seem to specify how long delegates are required to stay bound.)
So, what’s the first thing that jumps out to you about the chart? The fact that (at least) 29 states of territories only require their delegates to be bound for the first ballot. After that point, the delegates are free to vote their conscience.
Let’s break it out a little bit more and you can see why this situation becomes increasingly dire for the candidacy of Donald Trump (unless he gets 1,237 pledged delegates). Of the 29 states that release their delegates after an inconclusive first ballot, Trump has a majority of the delegation in 15. Cruz has a majority of the delegation in 3. Rubio in 2. Nine states that release after the first ballot have not voted, but, of those states, at least 5 are projected to be Trump victories.
If we are talking about the actual, numerical, totals of delegates at play, we would have to conclude the following: after the first ballot, the projected range (because some states have yet to vote) of bound delegates for Trump would drop from 1,100-1,200 to 385-425. Cruz’s range would drop to something like 330-370. This would leave, after the first ballot 1,600-1,700 delegates nominally unbound.
Now, this is where organization is tested. If we were to go strictly by the numbers, we’d expect to see those 700ish unbound “Trump” delegates to remain faithful to “their guy” (i.e. the one who won the vote and got them their “chair”). In the “perfect” Trump world, his people would court the ~200-300 delegates who either didn’t support Trump or Cruz or arrived at the convention unbound in order to bolster the vote for the second ballot to push over 1,237.
But, we don’t live in the “perfect” world. The big question is: how many defections can Trump and Cruz expect at the convention?
Now, you might be saying “wait, I thought this was just an article detailing how Cruz will out-work Trump in the states in order to get the nomination on the second ballot.” You might be right about that. If Cruz can get enough of his supporters bound as Trump delegates and then keep Trump under 1,237, he may make a push for the nomination as delegates become unbound.
However, let’s look at this objectively. There is no sense in the likes of Jeb Bush, Scott Walker, and Lindsey Graham endorsing Ted Cruz if there wasn’t a plan to make sure that Ted Cruz was not the eventual nominee. While the concept of a “Trojan Horse” advanced by the Trump campaign is melodramatic, it has a tinge of realism: members of the Republican Party establishment would prefer that neither Trump nor Cruz get the nomination. Some of this comes from a purely “electability” standpoint while there is also a deeply Anti-Trump/Cruz sentiment pervading the opinion-shapers in the Party.
So, let’s talk about the Second Ballot. For the sake of the argument, we’ll give Trump 450 bound delegates and Cruz 350 (meaning legally they must still vote for their chosen candidate) (n.b. for states that have a “35%” second ballot threshold, that number for Cruz is 886). We are assuming that, at the convention, Trump will arrive with 1,150 delegates and Cruz will have a little under 800. Kasich and Rubio will have 150 each. Around 200 delegates will arrive in Cleveland nominally unbound. (Again, approximations to show the hypothetical).
After the first ballot, the number of unbound delegates spikes from 200 to between 1,600 and 1,700. Now, it’s probably safe to say that Cruz will have a high delegate retention rate. So, he may retain his 800 into the second ballot (basically only needing to pick up 500 of the people who supported him before). About 100 “Party Leader” delegates will be unbound at this point (each state or territory gets 3). They are very unlikely to support Trump and are a wild card. In addition, if the 200 nominally unbound delegates declined to support Trump on the first ballot to give him the win (this is probably unfair, we’d project Trump may get as many as 50 of them), they would not switch to support him on the second ballot. Then we have to see what Trump’s organization has put together. He will need to cobble over 700 delegates back to his side to retain his starting number. As we discussed, folks who have been selected to fill chairs by Party elites are not likely to continue supporting Trump. Given the shoddy apparent nature of Trump’s State Convention teams, if he can retain 60%, he’d be lucky. Two other wild cards are whether Ohio’s delegation needs to remain bound to John Kasich and whether Marco Rubio is entitled to keep his 170ish delegates.
So, the projection would be that, on the first ballot, the voting would be reminiscent of the polling thusfar: Trump ~1200 (1,150 + 50/200 unbound), Cruz ~900 (less than 800 plus over 100 unbound), Other (Even if Rule 40 stays in place, the delegates remain bound) ~270.
The second ballot will be interesting. The model we’ve discussed here would have Trump around 900, Cruz at nearly 1,000, and then nearly 600 delegates who we would otherwise classify as “establishment.” That bloc has one of four options: give Trump the nomination because he was only a few delegates shy on the first ballot, rally behind Cruz like they did on the campaign trail, unite behind a new/old establishment candidate, fracture into five or six factions behind several candidates.
Without more information, it’s impossible to speculate about further outcomes. However, what remains abundantly clear is that the odds of an open convention are increasing and Donald Trump’s nomination chances are not likely to fair well at such an event.
The regular caveats apply, of course. If any campaign would completely buck traditional logic, it’s going to be this one.