Ted Cruz's 51% path to first-ballot nomination

UPDATED: Added a new tab to the spreadsheet to game out an alternative scenario: Cruz 45, Trump 40, Kasich 15 in remaining contests.

Despite the near-sweep last night for Donald Trump, Ted Cruz still has a plausible path to arriving at the 2016 Republican National Convention in Cleveland with enough pledged delegates to win the nomination on the first ballot.

I've seen many hasty, sloppy delegate-count projections that make faulty assumptions. They often assume that future delegate allocations will be roughly proportional to future popular vote totals. That's easier than analyzing the rules of the states yet to vote.

Twenty-one states and American Samoa have yet to vote. I've looked at each state's rules (thanks to the venerable and detail-obsessed website The Green Papers) and applied one assumption, as a starting point for discussion: That the anti-Trump forces coalesce behind Ted Cruz, who manages a 51% win in every remaining contest. That's 51% of the vote statewide, 51% in each congressional district (if delegates are allocated by CD), 51% in caucuses and conventions.

51% is not a tall order. Trump did not manage to win a majority of the vote in any jurisdiction on March 15, with the exception of the Northern Marianas. Florida -- Trump's second home -- is the only state that he plausibly could have won in a head-to-head match with Cruz.

Many states remaining are plurality-take-all. A few are proportional, except that a 50% winner gets the grand prize of all the delegates.

There are three remaining caucus-and-convention states: Wyoming has to elect its 14 statewide delegates. American Samoa will pledge 6 (its RNC members will remain unpledged). Colorado will elect 34 delegates at its state convention: 14 will be chosen by the convention as a whole, 3 by each congressional district caucusing separately, and the ballots will show the presidential candidate to whom each delegate candidate is pledging his or her support. (Colorado will also send its RNC members unpledged.)

What people who have never been involved in caucuses and conventions fail to understand is that a majority of the delegates to a CD or state convention pick ALL the national convention delegates. Back in 1976, a majority of delegates to the Oklahoma Republican convention were Reagan supporters, which meant that they elected a unanimous slate of 36 Reagan delegates to the Republican National Convention in Kansas City. Based on the Wyoming county convention results, it seems likely that Cruz supporters will control the state convention. I would expect the Colorado and American Samoa conventions will go the same way -- if anti-Trump has majority support, and if Trump opponents coalesce behind Cruz, Cruz would win all of those delegates.

I've uploaded a spreadsheet to Google Docs showing the math. If Cruz has a bare majority of support in the remaining contests, he enters Cleveland with 1289 bound delegates, enough to win the nomination on the first ballot. That doesn't depend on any of the unbound delegates voting for Cruz. Trump would enter Cleveland with 734 bound delegates.

Notice too that I haven't allocated the Pennsylvania congressional district delegates. According to The Green Papers:

Rule 8.4 of the Republican State Committee of Pennsylvania's Rules states that all delegates elected by Congressional District "...shall run at large within the Districts and shall not be officially committed to any particular candidate on the ballot.".

Conceivably, the Cruz and Trump camps would endorse slates of delegates in each congressional district, which would allow us to apply the winner-take-all rule by CD, adding 54 more delegates to Cruz's total. (This table from the state election board will show you all the candidates who have filed to run for delegate in Pennsylvania.)

Varying the scenario a bit, suppose that Trump wins, say, 20 congressional districts, while still losing every statewide total. (I could imagine Trump winning heavily Democrat CDs in New York City with low GOP turnout, without making a dent in a Cruz win fueled by heavy upstate support.) Switch 20 CDs and the total going into Cleveland is Cruz 1239, Trump 794. Cruz would still have a majority of delegates bound to support him.

Even if Kasich stays in and pulls 15% of the vote, Cruz could win remaining contests with 45 to Trump's 40 and go into Cleveland with 1214 delegates to Trump's 800, just shy of what he needs for a first ballot win, but close enough to get over the top easily and to dispel any Trump claim to victory. That's because in most of the remaining contests, a plurality is enough to win all the delegates. If Cruz marshals his forces to get his delegates elected in the Pennsylvania CD contests, he could still enter Cleveland with a bound 1st ballot majority.

Trump can be stopped on the first ballot in Cleveland, if the majority of Republicans, who want to stop Trump, coalesce now behind Ted Cruz. Cruz can win the nomination outright, fair and square, based on the votes cast in primaries and caucuses, without changing the convention rules, without the appearance or reality of backroom deals, without risking a Trump-voter revolt over perceived unfairness. All that is necessary is for the #NeverTrump forces to coalesce today behind Ted Cruz.

AND ANOTHER THING: The first-ballot nomination scenario would have been even stronger for Cruz had Rubio and Kasich dropped out as late as last weekend. (It's evident now that the poll showing Rubio winning early voters in Florida was way, way off. The subsample size was 72, which has an MOE of 11.55%.) Cruz still might not have won Florida -- it's Trump's second homestate -- but he likely would have won the remaining Ides of March states. Claims by Rubio and Kasich supporters that a contested convention was the only way to stop Trump was self-serving and wishful thinking.

ONE MORE THING: By my count, there are 906 delegates that will be bound to one candidate or another by upcoming contests. In which of the remaining states can Kasich win a plurality? He's only done well in his home state and two very small New England states where he spent a lot of time. I could imagine Kasich winning a state like Delaware , Connecticut, or Rhode Island. Maybe Kasich takes a CD or two in New York; he only won six CD delegates in Illinois. So that's 70 delegates or so he might win. The remaining 836 are going to wind up either in Cruz's pile or Trump's, and in large blocks. Each contest will either put Cruz closer to a majority or Trump. Only a very precise and improbably result would produce a situation where Cruz and Trump are both short of a majority and close enough to each other that neither can claim a mandate. Choose ye this day....

WHAT ABOUT KASICH? There aren't enough delegates remaining to win for Kasich to win the nomination on the first ballot. He would have to sweep all remaining delegates and win some of those released by ex-candidates or otherwise unbound delegates to get the majority. He would have to prevent Trump and Cruz from winning a combined 1237 delegates; they already have 1114. A nomination that involves last-minute rules changes or other convention chicanery would be seen as illegitimate by a big chunk of the GOP electorate, dooming not only the GOP's presidential hopes, but likely ending congressional majorities as well. Also, Kasich's decision to skip the Utah debate after Trump punted it tells me that Kasich is playing a game, positioning himself for a job in a Trump administration and not serious about winning.

LINKS:

Ace links and explains one of the media's confusing oversimplifications:

One important thing to keep in mind is that a lot of times the media tells you this or that primary is "proportional."

In fact, they're usually not. Illinois and Missouri were supposedly "proportional." That was shorthand for "Winner take all in the statewide race, then winner take all per Congressional District."

In fact, these "proportional" races turn out to be Winner Take Most, because the statewide winner usually wins almost all -- or just all -- of the Congressional Districts too.

Thus such races can quickly become de facto Winner Take All races, despite the media calling the "proportional."

If someone wins statewide, they might have a massive margin in one CD and lose the rest, but it's more likely that support will be evenly distributed across CDs, barring some special local circumstances. If the race is close, like Oklahoma's 2008 primary, which was then winner-take-all by state and CD, it's more likely that the CD victories will be split. In '08, McCain narrowly won statewide, beating Huckabee by 36.6% to 33.4%, but Huckabee won the two congressional districts nearest Arkansas.

Listening to the Glenn Beck team do their back-of-the-envelope calculation on the March 16, 2016, show, I noticed (with much consternation) that they treated "Winner-Take-Most" (really Winner-Take-All by state and by congressional district) as if they were proportionally allocated.

Ace speculates that Cruz might have better odds of reaching a number shy of 1237, but high enough to be nearly tied. If Cruz and Trump split all 906 remaining delegates, they'd be tied at 1010. (Any Kasich wins of a state or CD would reduce that number accordingly.) 181 are currently bound to ex-candidates for at least the first ballot, unless released. 271 would be uncommitted and up for grabs. Exit question: What will the RNC do to stop Trump from buying delegates?