Do you like math? I do. So does Sam Wang, a very smart numbers guy at Princeton who publishes the indispensable Princeton Election Consortium blog. His new, must-read post shows that unless something quite dramatic and, frankly, unexpected happens in the Republican primary race very soon, Donald Trump is almost certain to be the Republican nominee.
Why? The short version is that, given the (incredibly arcane and complicated) way in which Republican delegates are awarded, it’s quite possible for a candidate to amass enough delegates for the nomination yet never win a majority of the vote in any primary or caucus. Right now, Trump remains in the lead, polling in the low 30’s nationally, as he has been for a long time, and a bit better than that in South Carolina. There’s relatively little recent polling data from other states. And, of course, he has lots of momentum from his big NH win.
Wang ran a lot of numbers, and concluded:
The not-Trump scenario occurs if Republicans cull their field, fast. As far as I can tell, if Republicans want a candidate who is acceptable to most of their party to get most of the convention delegates, their deadlines are:
- Deadline 1 (February 29th): Get down to two alternatives to Donald Trump as a consequence of South Carolina and Nevada – and before voting starts on Super Tuesday, March 1st.
- Deadline 2 (March 14th): Settle on one alternative to Trump as a consequence of Super Tuesday and the March 5th-12th primaries.
… If these drop-dead dates aren’t met, Trump could still be stopped, but it would be difficult. First, it would require somebody other than Trump to take the popular lead in April. In a three-way race, that is hard to imagine. Even in a two-way race, it is not at all clear that Trump will lose, since for now, he picks up enough “Establishment” support in head-to-head matchups to get a majority. Consistent with this, exit polls in New Hampshire show that some Republicans of all stripes like Trump.
In other words, the GOP field needs to be reduced to three by Feb. 29, or two by March 14, for there to be a good chance to deny Trump the nomination. That seems very unlikely to me, for several reasons.
- Unlike most Republican primaries and caucuses, which award delegates in a way that Wang calls pseudo-proportional (it varies state by state), the delegate-rich contests in Ohio and Florida are winner-take-all. That creates an enormous incentive for candidates with home-field advantage in those states – Kasich, Rubio, and Jeb! – to stay in until then. And those contests (plus three other states) are on March 15, after Super Tuesday on March 1, and 8 other states plus DC and PR on March 5-12. By the time the results are in on March 15, almost 60% of Republican convention delegates will have been awarded.
- Kasich of course did unexpectedly well in NH and so has some wind at his back. He has little expectation of doing well in the south, where most of the upcoming contests are, so bad results there won’t push him out. Rather, he’ll try to hold on until states where he stands a better shot are in play, which is later in March and April. In addition, he just picked up a mega-donor and fundraiser in the person of Kenneth Langone, who formerly backed Chris Christie. That could give him just the financial boost he needs to survive poor showings on Super Tuesday.
- Bush and Rubio really don’t like each other. The notion of one of them conceding their home state of Florida to the other before the voting on March 15 seems implausible to me. Plus, while before NH Rubio had some momentum and Bush seemed to be fading, the NH results were awful for Rubio and not-great-but-sort-of-encouraging for Bush, so at this point they seem roughly evenly situated, and of course, both of them have wealthy SuperPACs behind them.
- Ted Cruz isn’t going anywhere. He’s had two solid showings, and he seems likely to do reasonably well in SC (he’ll probably finish second) and on Super Tuesday. And he’s so universally despised that, at this stage, it seems incredibly unlikely that the others would cede the field to him in hopes of stopping Trump. It’s not clear to me that they’d see a Cruz candidacy as any better for the party; anecdotal reports suggest that some establishment types would actually see it as worse.
Of course, things could change. Trump could finally self-destruct, though it’s increasingly difficult to see what he could do that would be outrageous enough to affect his support. His poll numbers could be wildly inflated, though NH certainly suggests that that is not the case at least in primary states. And support for say, Rubio or Bush could drastically surge, giving us a new frontrunner.
But it’s hard to see why any of that would happen any time soon.
It’s Yuge.
…is Trump with a majority of delegates, at least not necessarily. If he doesn’t go to convention with a majority already in the bag I can see convention finding another candidate.
But the point is that if he stays where he is in the polls, he can still amass a majority of delegates before the convention unless the GOP field is winnowed quickly. Do you disagree with that? Wang’s math seems pretty compelling to me.
Nullifying the primary results would be an entirely unprecedented move in the modern era, if they did that they should just pick Jeb to continue the Bush tradition of nullifying popular votes.
eom
The convention can easily justify coalescing around an establishment candidate after the first ballot if the remaining establishment candidates combined have more delegates than the highest single delegate recipient.
…or do you mean plurality? I think if the GOP field is not winnowed he could go in with the most votes, but not a majority. If it IS so winnowed then he or someone else has a much better chance of going in with an actual majority.
Again, the point of the simulations is that Trump can win a majority of delegates (and hence the nomination) by winning only a plurality of votes in statewide contests. You should read Wang’s posts. They are very interesting.
Or Bloomberg on a third party ticket. I think he would kill in a HRC v Trump battle.
Ever see physics demonstrations of the two-body problem? You get predictable results, even keeled, steady changes. Then you go to the three-body problem and it’s chaos. Anarchy. Fire, maelstroms, protesters in the streets.
I think a Trump-HRC-Bloom three way battle royale would be like that. Impossible to predict from iteration to the next. Chaotic and fascinating, but entirely unpredictable.
Every 2012 Romney state would go to Trump, I think that’s a safe assumption. Some blue places like Maine, with its LePage plurality and split electors would probably go to Trump with Bloomberg on the ballot, maybe more than that. Even carrying just his home state would be enough for Bloomberg to throw this into the House whose Republican majority would throw the presidency to Trump.
No way that’s a safe assumption.
I actually think Bloomberg would fare as well as Perot. His stance on firearms means he’s not gaining traction with a big chunk of voters in lots of Romney states, to be sure. His pro-business stance might well pull off suburban and more upscale GOP voters in plenty of places though. Trump’s ability to say crazy things could offend citizens of a state/region/religion at any time, and a 3-person race could well push “marginal” Romney states to HRC.
Missouri, Kentucky, Arizona, West Virginia, Tennessee, North Carolina, Nebraska CD, Arkansas, Louisiana, Virginia, Montana, and Georgia all come to mind as places that could go HRC in the aforementioned three person race. Hell, William Clinton and Barack Obama collectively won those places at least once.
over Bloomberg.
Lots of folks in the GOP won’t be returning for the next term, so could safely ‘take one for the team’ and vote against Trump, to deny Trump the White House. And if the choice was Trump or Bloomberg, want to bet a lot of Dems in the House would vote for Bloomberg?
30-40 Republicans, half of whom probably wouldn’t be returning, with the Democratic Caucus makes Bloomberg the President.
None of this is to say I think it’s likely. If Bloomberg wins, I doubt he wins more than a state or two.
If it goes to the House, do they HAVE to choose one of the candidates on the ballot?
I just read the relevant section of the constitution, and it’s not entirely clear to me.
I’m sure Mitt Romney’s best lawyers are already on it, though. /snark
The 12th amendment calls for the House to choose from among the top three recipients of electoral votes (down from five in Article II).
…if it goes to the House it’s one state, one vote so that has to factor into the math.
I get the theoretical but I have lived the real. I have never seen a candidate who is so hated (sometimes unjustly) for so long as HRC. I have never seen a lunatic like Trump treated seriously. I fear that Bloomberg becomes the NYT/USA/CNN choice.
If Trump is loathsome to enough Republican backers I could see them flip over to Bloomberg. Though what a wierd race where three New York City residents with tremendous connections to Wall Street and one another are running. More reason to vote for the outer borough socialist from Vermont I guess.
He is known nationally as a northeastern liberal who wants to ban soda. He is too conservative for the left, and WAAAAY too big intrusive government that wants to control what kind of car you drive and what you eat. There are so very many sound bites of him ruminating on this stuff that I cannot imagine he has appeal to anyone.
insert “for the middle and right” after eat
Iowa had one result, New Hampshire another, reshuflling the deck. I think the Iowa strong second place finisher (or winner) will try to pressure the others to drop “for the good of the party.” Or, more accurately, they won’t have to because donors will pressure the others.
If it’s Jeb v. Rubio, they might hate each other enough to slug it out a little. But if it’s, say, Kasich vs. Cruz with Trump still ahead, there will be a lot of pressure on Cruz to play nice. And Cruz if he’s strong might get a VP shot, but I don’t think they want to nominate him.
to our enemies.
a hold out on Trump’s inevitability. I think he was so concerned with following the data and not the hype, that he overcompensated in not considering Trump’s viability.
Trump is a losing candidate. People are embarrassed by him. He’ll lose unenrolled voters. He’ll lose minority voters even more. There are a lot of people who hate HRC, but Trump is scary to many people, Republican and unenrolled. He has no ground game, no GOTV. That stuff matters in the general election. People may not vote for Hilary, but they will sit out the election.
No one knows who Michael Bloomberg is outside of New Yorkers and some in the GOP establishment. He’s too moderate for the base and Trump supporters. The GOP has big problems.
I think he made the mistake of veering into punditry, when his real value to political discourse has been as a dispassionate number-cruncher. He personally couldn’t believe that Trump was a viable candidate, so he refused to believe the data. Which, ironically, is basically the same mistake all his detractors made in 2012.
I keep going back to the less engaged and unenrolled voters when I think of a Bloomberg candidacy. He can buy millions of eyeballs (ad time on every content provider) but do it much more targeted these days because you can do it by local cable provider and social media.
No need to spend money across the nation when you can blow off the ad costs of states you know you cannot win (I am thinking Bloomberg won’t play well in Texas or many southern states given his position on guns (and oh yeah, he wears a yarmulke on Friday night).
Candidates get a preferred ad rate compared to super PACS according to federal law. That means self-funded Trump and Bloomberg campaigns will get more eyeballs than the same amount of Hillary’s super PAC money
take more than media. They take boots on the ground.
Trump can work with no organization because he’s Trump and the GOP base is batshit crazy. For Trump, I think there are going to be organizational problems in the general election.
Bloomberg has no organization. No media profile. No charisma. He’s from New York, not an automatic attraction for many Southern voters. He’s moderate. He’s in favor of gun control to the point of starting an organization. He’d be a great candidate for the 1950s GOP.
Check out this WSJ piece about his gun club.
He’s either lying about being anti-gun, or he’s even more nuts than Trump. He endorsed several moderately anti-gun control Republican incumbents over Democratic challengers who were pro-gun control. He stayed out of competitive races where we had an opportunity to replace NRA zealots with pro-gun control people. Even from a general pro-incumbent standpoint it makes no sense because he left out vulnerable Democratic incumbents with a strong record on gun control.
Given that there hasn’t been a modern election like 2016, I think there’s good chance that Silver’s data just isn’t able to pick up the anti-establishment feel of the populace.
For example, one of the big weights in Silver’s data is endorsements, because he recognizes that endorsements means money and ground troops.
In an anti-establishment year, though, the endorsements may be as likely to weigh against a candidate for having them as it would help.
Silver predicted a Hillary win in NH, despite the polls which suggested otherwise, for a very, very long time, because of all those other variables in his modelling. And, by all means, it seems like he’d have been right… basd on past years. Heck, Marty Walsh freaking trucked in 1000 people to NH to help Hillary win. Every Massachusetts pol and their mother was in MA, with at least a car-full or two of allies, knocking on doors for Hillary. This is a big part of how Hillary beat the polls in NH in 2008, for such a stunning upset.
But those predictions are only good if this election is similar to elections in the past. So far, it doesn’t seem to be.
So… I just think there’s a lot of reason to suspect that Nate’s predictive modelling isn’t going to be nearly as spot on as it has been in elections past, and that’s why Nate’s been caught a bit flat footed in his early predictions of how all this would play out.
The endorsements are why I’m not voting for her. If every criminal in town endorsed a candidate for District Attorney, wouldn’t you be suspicious? Same deal here.
538 just bumped Nevada to a push, based on a new poll showing Sanders and Clinton tied.
Popcorn time.
The 538 model weighs polls based on timeliness and reliability. As of now, the only poll that’s timely enough is the most recent one, by TargetPoint. It’s the first time TargetPoint has polled the Nevada primary this go-round, so we have no idea how accurate they are.
All we know is that a single pollster with a broad poll (1k+) came up with a tie. If you ask the folks at 538 they’ll tell you that sure, their model is forecasting a tie — but that the error bars are really really wide on this one.
Once (if?) another poll comes out, we’ll have a much better feel.
Elsewhere Wang characterizes his analysis as follows:
I won’t that that can’t happen, but it’s not a slam dunk (especially with Cruz in the race). The likely alternative however is a messy fight all the way to the convention.
but other than Carson, who’s going to drop out?
The earliest Kasich drops out is after the Ohio results, Mar 16. At that point, 59.87% of the GOP elected delegates will already be allocated. John Bush and Marco Rubio will wait until the same day (FL results).
And here’s the tricky part — the winner take all (or most) states between now and then:
SC, AL, AR, GA, MN, OK, TN, TX, VT, ME, PR, ID, DC, FL, IL, MO, N.Mar., OH.
Proportional (of sorts):
NV, AK, MA, VA, WY, KS, KY, LA, HI, MI, MS, Guam, NC,.
So long as 4 (or 5) other candidates are in the race, Donald Trump can win quite a few of those “winner take all/most” events with his 30-40%. By the time Ohio and Florida have come and gone, whoever is in 2nd place may have too much ground to make up. The upshot is that nearly all GOP events post-Mar 15 are winner take all or winner take most, which means that if the non-Trump catches fire he could gain ground quickly. And, of course, superdelegates.
So yeah, it’s not a slam dunk, but it does look likely that Trump will have a substantial lead in elected delegates as of March 16. If that happens, the longer other candidates stay in the race, the better chance Mr. Trump has of consistently winning with a plurality.
…at least not nearly to the extent the Democrats do.