In Which I Am Right: Shadow Primary plus Kasich
In my recent piece on convention delegates, I speculated that, in the scrum to win the loyalty of as many delegates as possible, Ted Cruz would have the edge. Cruz has a strong organization, has been gearing up for precisely this fight for over a year, and his base is both a lot larger than Kasich's base and a lot more active within the GOP than Trump's. This might allow Cruz to secure the loyalty of delegates bound to him, as well as capture the loyalty of delegates bound to others.
Early indicators suggest this might be right.
A story from Politico yesterday discusses South Dakota's delegation, which Cruz apparently dominates. But this is not too surprising, because South Dakota is prime Cruz Country: a Midwestern conservative plains state with an especially arcane delegate election method (delegates are chosen in convention before the primary on June 7th). I would be surprised if Cruz did not win the state outright, ending up with 29 bound delegates who are also totally loyal to him.
More intriguing is word from Louisiana, reported in last night's Wall Street Journal. Although Donald Trump won the popular vote in Louisiana, by 3.6 points, and the two candidates tied in the bound delegate count (18-18), Cruz has apparently captured all of Louisiana's 10 unbound delegates (5 of which used to belong to Rubio), giving Cruz a 28-18 delegate lead in Louisiana. Unsurprisingly, the Cruz-controlled delegation then elected fellow Cruz supporters to fill various RNC committee positions, taking 5 of those 6 slots as well. (We'll discuss the importance of those committees in upcoming entries of Conventional Chaos.) This is all good news for Cruz, because Louisiana is a southern state where (as the Journal reports), the Trump campaign not only won a strong plurality (41% of the popular vote), but also made significant inroads with Republican party officials, who normally wield outsized influence in the delegate selection process. If Cruz's delegate-acquisition program is stronger than Trump's in Louisiana, it's probably stronger than Trump's everywhere, or just about.
That said, I don't want to read too much into this story, because all the Journal reported is that Cruz did very well picking up unbound delegates. If Cruz is to have any chance at all of denying Trump 1,237 first-ballot votes, he is going to have to pick up virtually all of the unbound delegates nationwide anyway. (Fortunately, there aren't very many in the South. Louisiana had 10, which Cruz has picked up, Georgia has 16 now that Rubio dropped out, and Oklahoma has 2. Cruz is very well positioned to get all of them, which is good because it's not yet clear that he can afford to lose any of them.) These are good, necessary pickups, but they aren't what I'm really interested in.
What I really want to know is whether Cruz managed to get any Cruz supporters elected to Trump's delegate slots (or vice versa). This would begin to tell us how well Cruz (and Trump) are going to be at acquiring "supporters-in-name-only," or what John Yob called SINOs, who could wreak havoc on a convention by rendering the bound delegate count moot on the second ballot... or the first. If Cruz's delegate-acquisition organization is good at picking up unbound slots, it's probably good at picking off Trump slots as well... but we don't know, and won't know, until a newspaper starts asking.
Unfortunately, the Journal had absolutely nothing to say about this.
This is not terribly surprising, because (segue to new topic!) the Journal's reporting and opining on the delegate race has been abysmal, especially in their fawning and downright irrational coverage of Gov. John Kasich's campaign. In the story we're discussing, for example, the Journal reports, falsely:
The Texan must win 85% of the remaining delegates to win outright, a highly unlikely scenario with many states awarding delegates proportionally.
In fact, of the 19 states yet to vote, only 3 are awarding delegates proportionally. All the rest are winner-take-all or winner-take-most (or unbound, which are more likely than not to be winner-take-all in effect). This distinction is a very big deal, because proportional allocations can favor multi-candidate races, while winner-take-most allocations strongly favor the plurality winner and harshly punish candidates who split their votes.
Unfortunately, based partially on this error, the editorial board recently made the insane argument that Kasich should stay in the race in order to win the Northeast:
Mr. Cruz finished back in the pack in New England outside Maine, and the states coming up include Connecticut, Rhode Island, New Jersey, New York, Delaware and Maryland. Mr. Trump wins the nomination if he wins most of those states.
For one, it is totally false that Trump wins the nomination if he wins most of the Northeastern states. In fact, my personal spreadsheet (hey, everyone builds one after a certain point) has Trump winning every single delegate in the Northeast. As long as Trump is denied delegates in Wisconsin, Indiana, and most of California, and Cruz takes the states he's already expected to take (South Dakota, Nebraska, Montana, etc.), Trump will not have 1237 bound delegates going into the convention... even with the whole Northeast under his belt.
Unfortunately, Kasich's presence in the race makes winning those key states away from Trump exponentially more difficult, since every single one of them allocates delegates winner-take-all or winner-take-most. I am not alone in my analysis: the best analysts, both on the Left and on the Right, agree that Trump will likely win most or all of the Northeast but can still lose the nomination, and every serious observer agrees that, by staying in the race, Kasich is either a stubborn fool or running to be Trump's vice president. Unfortunately, the Wall Street Journal editorial board is apparently doing the same -- or simply cannot bear the thought of a nominee who isn't popular in the Acela Corridor, and has become willfully blind to facts that get in the way.
For two, it is now obvious that Kasich can hardly make a dent in the Northeast. In delegate-rich New York, for example, Trump leads by at least twenty points in every poll, and by 50 points in one recent poll. Better still? Cruz, not Kasich, is in second place in most of those polls. In Maryland? Same story... which the Journal knows full well, since it's their own poll. Trump will win the entire (or nearly the entire) Northeast with or without Kasich in the race. Kasich might beat Cruz throughout the Northeast, but, without a radical change in the direction of the race, Trump will still win the night and the vast majority of the delegates.
You may recognize these two points as precisely the same two points I made to argue that Rubio should drop out before March 8th. You may further recognize that these two points turned out to be absolutely correct on both March 8th and March 15th. The Journal, inaugurating its practice of being totally oblivious to both delegate counts and delegate allocation methods, was on the wrong side of this question three weeks ago, too, arguing that Rubio should stay in the race days after it was clear he was helping Trump more than anyone. I was right then (and I was not alone). Same argument applies now.
For three, as the Journal itself appears to concede, the best hope anyone has of stopping Trump in a three-way Kasich-Cruz-Trump race is if Cruz supports Kasich in Kasich's strong states and Kasich does the same in Cruz's. But the Journal makes no mention of the fact that Kasich has deliberately embarked on a maximally destructive path, focusing his attention in obvious Cruz-favored states Utah and Wisconsin. To put this in perspective: in Utah, Kasich was hoping to get delegates (which would be nice for him) while denying Trump any (which is the absolute top strategic and tactical priority of any stop-Trump campaign bar none). Under Utah delegate allocation rules, the only way Kasich could both win delegates and deny delegates to Trump is by personally winning more than 50% of the popular vote in Utah. Based on the pre-caucus polling average in Utah, which Cruz led by huge margins, this means Kasich would have had to beat his polls by as much as Bernie Sanders did in Michigan. You may recall that people got pretty excited about Sanders' win there. That's because it was the second-largest primary election upset in history. So Kasich deployed resources that would either have no effect at all, or could conceivably help Kasich in a once-in-a-generation lucky break, or, in all other cases, would give Trump delegates. It was a totally irrational move. The New York Times reports that the Cruz campaign is furious about it. You're goddamn right he is, and has every right to be. (The Times also reports that Kasich is furious at Cruz for calling him a spoiler. Kasich can't handle the truth; too bad.)
Fortunately, Kasich's campaign is too hapless, and Utah's voters too intelligent (with a late assist from Mitt Romney), for this to work: Cruz won the state with nearly 70% of the popular vote, denying all delegates to Trump, despite Kasich's spoiler effect.
Unfortunately, Kasich is now pursuing a similar strategy in Wisconsin, where Cruz narrowly leads him, while himself narrowly losing to Trump, in a state that allocates delegates winner-take-most. You may recognize this as the mirror image of demographically similar Illinois, where Kasich handed some 20 delegates to Trump by spoiling the race. If Kasich were serious about actually stopping Trump, he would be stumping in the New York City GOP congressional districts, where he can pick off 39 delegates from Trump without too much difficulty. Instead, he is deliberately running in states where he will do far more to help Trump than hurt him. The Wall Street Journal says nothing about any of this. The Journal, either through better-Trump-than-Cruz malice, mathematical ignorance, or sheer arrogance (ironically, precisely the same blind arrogance that drove Trump's supporters from the Wall Street Journal wing of the party in the first place), is enabling Kasich's totally destructive, Trump-abetting campaign.
Hopefully we don't have to wait for another awful night like March 15th for the Journal -- and Gov. Kasich himself -- to realize that, though I am often wrong, I am absolutely right about this. As of today, a vote for Kasich is a vote for Trump. This is not a slogan, and it is not a joke. It is the fact that will likely determine the future of the Republican Party -- if the Grand Old Party has a future at all.
P.S. A word of reassurance: though all agree that the Kasich campaign is a net benefit to the Trump campaign -- hence, "a vote for Kasich is a vote for Trump" -- there is no need to panic about it... not yet.
If Kasich performs unexpectedly well in the Northeast, that would be a minor miracle, but it would also be great news: Trump's path to the nomination would be complicated, perhaps ruled out entirely.
If, as seems far more likely, Kasich is humiliated in the Northeast, he would be forced to drop out... and the Northeast votes before the most critical states in the Cruz stop-Trump strategy. Either way, Cruz still has a path, though a much more difficult path than one where Kasich drops out before Wisconsin.