Two Polls Show Tight Ohio Race

We’ve already adjusted our model to account for a “Clinton ceiling,” and it’s looking like it may be the right projection.  New polls from Quinnipiac and Public Policy Polling show Sanders closing the gap in Ohio, Illinois, and Missouri.  Our model baked in a Missouri upset (we have him winning there by nearly 10 points on the assumptions late breakers go for him).  The potential softening in Ohio, where PPP has Clinton 46-41 is a significant shift from our model’s 55-45 split (as, under our rule, this would translate to a 9-point Sanders win).  PPP’s 48-45 split in Illinois is not too far from the 51.5-48.5 we’re projecting.

The GOP numbers are equally interesting in Ohio, showing a tied race there. If Kasich fails to win there, it’s hard to see a credible pathNew @ppppolls and @QuinnipiacPoll numbers show tight Dem race in OH, MO, IL with potential #Sanders upset.  #Kasich #Trump tight. for him and Rubio to stay in (and for the GOP to block Trump).

A not-unreasonable path to “victory” for #NeverTrump

Donald Trump has a good chance to effectively wrap things up on March 15.  Polls show him with a 20-point lead in Florida (where early voting opened as Marco Rubio was tanking), putting him on track to take all of its 99 delegates in the winner-take-all primary.  If Rubio fails to win Florida, it’s difficult to see him staying in the race.  Here’s how our model sees that race playing out:

Trump sweeps Super Tuesday part 2, Rubio drops out, Trump wraps it up in May.
Trump sweeps Super Tuesday part 2, Rubio drops out, Trump wraps it up in May.

Change four variables, though, and the race takes on a different flavor:

  1. Rubio drops out after Florida.  His supporters go 75% to Kasich, 20% to Cruz, and 5% to Trump (yes, that’s a total guess).
  2. Kasich wins Ohio.  This isn’t a huge stretch; he leads Trump in current polling averages, and the 6-7% Rubio numbers may heed his strategic-voting advice.
  3. Trump’s poll numbers are his ceiling.  Until our last update, our model split the “undecided” votes (i.e. the difference between 100 and the sum of the candidates’ support) proportionally.  We’ve added a number of different scenarios.  The one that tracks most closely to reality is that Trump’s polls are his ceiling; the biggest difference in his favor, according to the excellent fivethirtyeight.com poll average, was the 3.3% miss in Massachusetts.  On average, he comes in below his polls 1-2%.
  4. Cruz wins Illinois.  He trails Trump by 6 points in the late fivethirtyeight.com average, but when we apply the “Trump ceiling” rule, he comes within a single point.
Assume four things, Trump needs a second ballot.
Assume four things, Trump needs a second ballot.

Based on that, Trump misses. the target by about 20 votes. Still not an ideal outcome for the #NeverTrump camp, but it doesn’t require too many logical leaps.

Stretching a bit further into supposition, Rubio’s dropout could broaden the gap:

  1. Kasich or Cruz win Pennsylvania.  Rubio dropping out would put the three remaining candidates within a point or two of each other (possibly favoring Kasich, as Rubio still polls strongly here).  Kasich may end up off the ballot, though.
  2. Kasich or Cruz win Maryland.  There, they really would all get a third under our polling split.
Trump falls over 100 short if he loses OH, PA, IL, MD, and CA.
Trump falls over 100 short if he loses OH, PA, IL, MD, and CA.

This setup would already have Cruz winning Wisconsin and Kasich winning California, based on Rubio’s support pushing them over the edge. That’s about as close to a NeverTrump dream scenario that the numbers currently bear out.  It would have the weird effect of an all-out fight in California (!), but perhaps March 15 will shed a bit more light on Rubio’s future, and on Trump’s ceiling.

Super Tuesday quick take: Great for Cruz, Par for Trump, and Holding for Rubio

The delegate math for the Republicans is difficult, with each state having its own rules down to the Congressional district level.  More than 30 hours after polls closed, the majority of the states are still figuring out how to apportion the delegates.  We have updated our projections in a simple way:  the leading candidate in the state gets the balance of the delegates still on the table.  This rough math generally favors Trump, but also gives Cruz a boost in Texas.

Projected values before Super Tuesday
Delegates expected before Super Tuesday

Based on those numbers, Trump fell short of our delegate projection by 71 (we saw him winning 358, but now see that coming out to 277). Cruz overshot: we see him winning 237 against the 165 we projected. While the news cycle for Rubio is certainly negative, he slightly overshot based on stronger-than-projected showings in Minnesota and Virginia.

Actual delegates won on Super Tuesday
Actual Delegates Won on Super Tuesday

The biggest deltas in our projection are Oklahoma, Minnesota, and Alaska.  The latter two are simply because there weren’t any good polls, so the model defaulted to the national polling numbers, which benefited Trump. Hard to call those an upset.  Oklahoma, though, is a big win for Cruz. The bigger prize for him, though, is probably the Carson dropout, which should boost him 5-10 points in states with strong evangelical populations (too bad most of those have already voted).

Does this change the outcome? Right now, not according to our model.  We’ll have another post before the Saturday votes, but playing with the numbers still shows a high likelihood that Trump wins outright in the May/June timeframe.  Florida and Ohio will be absolutely critical to the “block Trump” faction of the party…and, of course, whatever it is that Romney has to say may still have an impact.

GOP Insiders Hurting Rubio, Selves

At a dinner following the South Carolina primaries, a GOP insider close to Jeb Bush said party leaders were looking for ways to get Cruz and Rubio to better cooperate in order to stall Trump’s march to the nomination.  The insider rolled off a number of options, including possibly stacking up a POTUS-VPOTUS ticket in exchange for a drop out.  A plan B, they said, was to get sufficient cooperation to force a brokered convention, at which point they could move the delegates in the right direction.

With Bush and Graham saying the party is “batshit crazy,” you’d think the establishment would be looking to save the party.  You’d be wrong. While Jeb will eventually endorse Rubio, GOP insiders say he still seethes at both Rubio’s decision to run and his personal attacks against Bush; Bush won’t offer that endorsement until the last minute.

The situation invites the question as to whether they consider Rubio a true “establishment” candidate. They shouldn’t, and it appears they don’t. Insiders are not actively pushing Kasich to drop out, reportedly in hopes that he can become the nominee in a brokered convention.

These insiders have better data than we do, but unless their data is wildly different than ours, they’re suffering from a bout of wishful thinking.  No one will drop out before Super Tuesday, and Rubio and Kasich will probably stay in through March 15, when Florida and Ohio vote in winner-take-all primaries.  The problem is, Trump is leading polls in those states, and will have a yuuuge delegate lead before March 15.

GOP delegate projection thru March 10 as of Feb 28
GOP delegate projection thru March 10 as of Feb 28

The chart above is based on state-level polls with state-level delegate allocation rules. After Super Tuesday, Trump should have around 440 delegates, Cruz 192, and Rubio 112.  Those projections would have Trump on pace to wrap up the nomination before May 1, with Cruz and Rubio getting just over 300 delegates a piece.

538 delegate projections run February 28
538 delegate projections run February 28

We’ve run dozens of projections based on current polling data[1]Of course, none of them involve Cruz or Rubio dropping out. We’ll need to check the 2nd preference votes for both of those candidates.  We haven’t seen good data since before Iowa.  At that time, nearly 30% of Rubio’s supporters went to Cruz and 10% to Trump.  Cruz dropping out meant 30% went to Trump and around 20% each to Rubio and Carson..  The only scenario that produces a brokered convention involves Kasich dropping out before Super Tuesday and 85% of his supporters breaking for Rubio.  Even then, Trump falls only 21 delegates short.

Run February 28 based on fivethirtyeight polls, assuming 85/10/5 Rubio/Cruz/Trump reallocation
Run February 28 based on fivethirtyeight polls, assuming 85/10/5 Rubio/Cruz/Trump reallocation

So let’s assume no one drops out until after Ohio and Florida vote.  What happens then? No surprise, Trump wins, needing only around 300 delegates more after March 15:

Kasich drops out after (losing) Ohio (to Trump). Trump wins the nomination.
Kasich drops out after (losing) Ohio (to Trump). Trump wins the nomination.

So where does this put us? Barring massive shifts in polling numbers[2]Or, Ohio and Florida’s winner-take-alls bouncing to another candidate, GOP insiders will be out of time to affect the outcome if they wait until after Florida to shape the race.  They’re probably too late already; their unwillingness (or inability) to do so before Super Tuesday probably cemented the deal.

Footnotes   [ + ]

1. Of course, none of them involve Cruz or Rubio dropping out. We’ll need to check the 2nd preference votes for both of those candidates.  We haven’t seen good data since before Iowa.  At that time, nearly 30% of Rubio’s supporters went to Cruz and 10% to Trump.  Cruz dropping out meant 30% went to Trump and around 20% each to Rubio and Carson.
2. Or, Ohio and Florida’s winner-take-alls bouncing to another candidate