Trump's Delegate Lead is Overstated
UPDATE: I have launched a Poor Man's Stop-Trump Delegate Tracker based on the "Nightmare Scenario" laid out in this post. The media and narrative-driven, math-challenged insiders have largely set their sights on March 15th as the pivotal day, just as this post predicted. As I said when I first posted this, don't listen to 'em!
Somehow, in the last four days, a large portion of the GOP has gone from, "Trump is a joke who cannot possibly win; therefore we don't need to fight him," all the way to, "Trump is an inevitability who cannot possibly be fought; therefore we won't," with no stopover in the middle.
This is very irritating. As of today, 118 delegates have been awarded, out of a total of 2,472; we're only 5% of the way into the voting. Donald Trump needs 1,237 delegates to clinch the nomination. He actually has 81, which means he's just 7% of the way to the total he needs. Calling this game for Trump would be like calling the NCAA college basketball championship after the first minute and a half of gameplay.
Can we give it some time, people? After all, comebacks have occasionally happened before, in both sports and politics. I've never seen a basketball game where Louisville leaps out to an early lead over Duke -- say, 4 points to 2 points -- and Coach Krzyzewski says, "Oh, well, I guess it's over!" and forfeits! The color commentary never says, "Louisville now has a commanding lead... TWICE as many points as Duke! Duke can never come back from this!"
Yet when Donald Trump pulls out to an early lead of 4 points to 2 (which is about where the race stands right now), suddenly it's the end of the world. Unfortunately, the ol' early-primary panic & surrender trick is not new. The media hype about one candidate being in the lead feeds back on itself, sapping morale from the other candidates and encouraging defections and donations to the current leader, until the field psyches itself out and withdraws, usually by the time Ohio votes. Rick Santorum was by no means a dead man in 2012, but he dropped out after narrowly losing Michigan and Ohio. Four years earlier, the machine did its best to drive Mike Huckabee out of the race weeks before it became mathematically clear that he actually should. "The early leader always wins" is a self-fulfilling prophecy... and an unsettlingly effective one.
However, with the Republican party and the conservative movement mortally imperiled by the potential candidacy of Donald Trump (and let no one fool you into thinking the party could recognizably survive a Trump nomination), one hopes that the machinery will be a bit less hasty to drive candidates out of the race unnecessarily. One hopes that, this year, someone might actually bother to look at the delegate math and see where the race takes us, if our three leaders hang in there until the bitter end.
On that note, guess what I did last week?
That's right: I made a spreadsheet, which calculates where the race takes us if our three leaders hang in there until the bitter end. I spent hours on TheGreenPapers.com figuring out state-by-state delegate allocation rules so you don't have to. (I had no idea there were so many subtly different ways to round numbers. But it seems like every state Republican party has its own rounding rules! And now I know them all!) I'm not going to share the whole spreadsheet with you, because it's enormous. But I will share with you the basic results.
The assumption of my little wargame is that the race eventually contracts to a Rubio/Trump/Cruz three-way (with Rubio mostly picking up Kasich's delegates and Cruz mostly getting Carson's). Trump stays in the lead, getting 40% of the vote everywhere. Rubio and Cruz, in every single state, get 30% each, splitting the non-Trump vote evenly. Rubio never wins a state in this simulation. (Cruz wins only Iowa.) This isn't really realistic: because Americans are not a demographic monolith, and Rubio and Cruz do not evenly split the vote everwhere, both Rubio and Cruz are sure to win at least one or two states along the line. (They're also sure to do relatively worse in at least one or two states, for the same reason.) But I'm not Nate Silver, so I went with the simplest, most pessimistic assumption I could: neither candidate ever wins a primary after today. Trump never gets an actual majority to vote for him, but he still wins every single state from here on in.
Tell me, how long do you think it will take Trump to clinch the nomination while winning every state? March 1st? March 15th, after the big winner-take-all states? Maybe a few more weeks after that?
How about June 7th, the very last day of the primaries? Yeah, that's right: if people keep their heads and don't start flocking to the con man, this three-way contest stays alive for the next three months even if Ted Cruz and Marco Rubio stay locked in combat the whole way.
Here's how it all plays out:
Totals 2472 total delegates available 1237 needed to win 1236 needed to force a brokered convention (presumably blocking Trump) Trump wins 40% of the GOP's national popular vote. Cruz and Rubio each win 30%.
Today Trump 82, Cruz 17, Rubio 16, Kasich 6, Carson 5, Bush 4, Fiorina 1, Huckabee 1, Paul 1 Trump 87, Cruz 22, Rubio 24 (reallocating from doomed/dropped out candidates) Trump % progress to winning nomination: 7% Rivals % progress to forcing brokered convention: 4% Total delegates allocated so far: 133 (5.4%)
March 1st (Super Tuesday) Trump 377, Cruz 205, Rubio 146, Unbound 29 Trump % progress to winning nomination: 30% Rivals % progress to forcing brokered convention: 31% Total delegates allocated so far: 757 (31%) If Trump sweeps every state on Super Tuesday -- which is very unlikely, given Cruz's polling in Texas and the general pro-Rubio attitude of the Minnesota caucuses -- there will be universal despair. As we see here, though, that despair is fueled by media hype, not by math. The non-Trump contingent is still winning.
March 6th Trump 447, Cruz 260, Rubio 206, Unbound 29 Trump % progress to winning nomination: 36% Rivals % progress to forcing brokered convention: 39% Total delegates allocated so far: 913 (37%) Given how unbound candidates are selected, it is very likely they will represent the most establishment/activist wing of the party: unbound delegates are highly likely to be Rubio or perhaps Cruz backers, but only rarely Trump supporters.
March 8th Trump 510, Cruz 305, Rubio 242, Unbound 29 Trump % progress to winning nomination: 41% Rivals % progress to forcing brokered convention: 47% Total delegates allocated so far: 1086 (43%)
March 12th Trump 517, Cruz 311, Rubio 248, Unbound 38 Trump % progress to winning nomination: 42% Rivals % progress to forcing brokered convention: 48% Total delegates allocated so far: 1114 (45%)
March 15th Trump 769, Cruz 337, Rubio 280, Unbound 95 Trump % progress to winning nomination: 57% Rivals % progress to forcing brokered convention: 58% Total delegates allocated so far: 1415 (57%) Today sees six states vote, including the crucial first winner-take-all contests in Florida and Ohio. In our wargame, Trump wins both. The media says we should panic if that happens. Don't. Notice that block-Trump is still winning by a nose.
April 1 Trump 852, Cruz 349, Rubio 292, Unbound 132 Trump % progress to winning nomination: 69% Rivals % progress to forcing brokered convention: 62% Total delegates allocated so far: 1625 (66%) This marks the two-thirds point in the primary. By this point, in the real world, either Rubio or Cruz really need to be negotiating about who will drop out. But neither actually really needs to be out until the winner-take-all states in the Northeast vote on April 26.
April 19 Trump 945, Cruz 356, Rubio 329, Unbound 169 Trump % progress to winning nomination: 78% Rivals % progress to forcing brokered convention: 70% Total delegates allocated so far: 1843 (75%)
April 26 Trump 1036, Cruz 368, Rubio 344, Unbound 223 Trump % progress to winning nomination: 84% Rivals % progress to forcing brokered convention: 75% Total delegates allocated so far: 1971 (80%) At this point, Trump has become very dangerous indeed. If Cruz and Rubio are still in it, still splitting their vote, they could still survive by one dropping out and endorsing the other... but their math is very, very fragile, and a few defectors or unbound delegates going Trump could give Trump his clinch early. If we get here, and the race looks like this, it's time to panic. But notice: the panic date is still two months from now. The only way Trump becomes inevitable here in February or early March is if we panic prematurely. (Which we probably will. Because we always do. But now at least you've been warned.)
May 10 Trump 1151, Cruz 371, Rubio 353, Unbound 223 Trump % progress to winning nomination: 93% Rivals % progress to forcing brokered convention: 77% Total delegates allocated so far: 2098 (85%) By this point, if Rubio and Cruz are both still in the race, still splitting their vote against a 40% popular Trump, it's probably too late to stop Trump. Trump hasn't officially clinched yet, but a series of winner-take-all states have allowed him to pull ahead of the field, with several states yet to come.
May 24 Trump 1188, Cruz 383, Rubio 376, Unbound 223 Trump % progress to winning nomination: 96% Rivals % progress to forcing brokered convention: 80% Total delegates allocated so far: 2170 (88%) Washington votes in a winner-take-most primary today, in the last contest before the close of primary season on June 7th, aka Ultimate Tuesday, when six states will vote, including California's massive winner-take-most 172-delegate jackpot.
June 7 Trump 1420, Cruz 408, Rubio 422, Unbound 223 Trump % progress to winning nomination: 100% (clinched) Rivals % progress to forcing brokered convention: 85% (failed) Total delegates allocated so far: 2472 (100%) (done) Finally, Trump is able to close the deal on the very last day of the primary season, mere weeks before the national convention, with big wins in California and New Jersey. Cruz and Rubio, by staying in all the way to the bitter end, ultimately made it impossible for either of them to win, making Donald Trump the Republican nominee for president.
So that's the nightmare scenario. It's the nightmare we're all having right now. But look how long it took for the worst to come to pass, even in this worst-case wargame. It's not hard to imagine how, in the real world, things could go better.
Suppose Cruz stays in all the way to April, but finally drops out on April 2nd and endorses Rubio, throwing his delegates Rubio's way. Rubio consolidates -- not hugely, just a little -- and, while still losing the entire Northeast on the 26th, he is able to pull off wins in a few small, mostly midwestern winner-take-all states -- Indiana, Nebraska, Wisconsin, and Montana, and (in a narrow fight) Maryland. Boom, that's enough: Rubio has just enough to block Trump's nomination in the first ballot at the convention. And that's without considering any other effects consolidation would have on the various winner-take-most races in the final third of the nomination race.
Or suppose Kasich wins Ohio, then gives a big speech and withdraws a few days later, knowing he has no path to the nomination (and he doesn't). He consigns his delegates to Rubio (or Cruz, but I just don't see that happening), and they are mostly cooperative. This gives the non-Trumps even more breathing space late in the campaign.
Or... but you get the idea.
The point is, we still have a long race ahead of us, in which practically anything can happen. We're only 90 seconds in to a 40-minute game. The thuggish crowd may be jeering and telling you to go home, but, if you watch the clock (not the stands), you know better than to give up now. Consolidation does have to happen: the party needs a single anti-Trump. But it doesn't have to happen right this instant, and it can certainly wait until after Super Tuesday, and even past March 15th.
But that depends on us not collectively losing our heads and anointing Donald Trump King of the Party three months before he's got even a decent shot at earning the title. Don't panic. Keep calm and vote against Trump.
UPDATE 27 Feb 2016 1451: In the original post, I copy-pasted some numbers for May 24th wrong. I fixed those. Nothing else affected.
UPDATE 28 Feb 2016: I can't say enough good things about this delegate forecasting widget from The Upshot blog, which allows you to replicate not just the wargame I played out in this post, but a wide variety of other possible wargames, like...
What if Rubio drops out on March 17th, throws support to Cruz, Kasich and Carson rally to the Cruz banner, what happens? (Answer: depends how much of the Rubio base goes to Trump. If 10% go to Trump, Cruz can force a brokered convention with a healthy enough margin that he has a good shot. If 20% go to Trump, Trump wins.)
They've got a few scenarios already set up that you can play around with. Check it out!
UPDATE 3: The updates to this post were getting crowded, so I launched a Poor Man's Delegate Tracker to compare actual results to this scenario.