Matt McDaniel

7 minute read

Earlier this week, the Des Moines Register released a poll of likely Republican caucus-goers in the First in the Nation caucus state (poll here, story here). With three weeks left until Iowans head to their caucuses, the poll, largely, showed what we’ve come to expect in Iowa: Cruz leading Trump but within the margin of error (25-22) and Rubio in third (12).

Yes, Trump might be statistically behind Cruz, but why would this poll premise a title of an article speculating about Trump’s downfall? After all a close second-place finish for Trump wouldn’t end his campaign and he’d just go on to capture New Hampshire, right?

Ok. The foregoing is the current (if grudging from some pundits) “conventional” analysis. While I would agree that this scenario (Cruz over Trump in Iowa, Trump generally unaffected goes on to win in New Hampshire) is most likely at this stage, let’s take a look at a growing seam that the Des Moines Register opened in its poll.

Like many of the polls that are currently being conducted, the Register asked voters to select a “second choice” in addition to their preferred candidate. Combining “first and second” choices had Ted Cruz at 48, Trump at 33 and Rubio at 28. Now, we have to realize that, among all survey participants, about 42% said that they had firmly decided who they intended to support in the caucus. Cruz supporters were at 51% and Trump supporters were at 64%.

Here’s where things get interesting. When people who did not choose Cruz/Trump/Rubio as their first or second choice were asked if they ever could be persuaded to support Cruz/Trump/Rubio, Cruz received 33% “maybe” and Rubio got 46% “maybe.” Trump lagged behind with 24% saying “maybe” and 41% saying that they would never support him.

So, why does all of this matter? Because Iowa is a caucus state and not a primary state. When Iowans gather, their preferences can change based on presentations by candidates’ supporters. This isn’t a situation where people just go behind a curtain, push a button on a screen and then go home. Rather, Iowa is a state where someone’s second choice can easily become the person they wind up supporting during the final caucus tally.

There’s an important person that we’ve left out of consideration so far in this analysis: Dr. Ben Carson. Now, you might (rightly) be saying that Carson’s whole political apparatus has decayed to the point that most folks would be surprised if he is able to make it to Iowa. However, in the Register poll, Carson’s combined “first and second” number is 19%. Granted, this number is not as high as Cruz/Trump/Rubio, but it is more than enough to put him strongly outside of the “also rans.” Carson’s direct amount of support in the poll (first choice) is 11%.

Now, we know from the past two election cycles that Iowans have selected the most conservative evangelical running (Huckabee in 08 and Santorum in 12) without regard to the “Buckley Rule” (select the most conservative candidate who would also be able to win). However, let’s assume that Carson’s 11% is soft. I think it’s safe to say that Carson is one of the most well-liked of the GOP candidates. However, his supporters could easily defect to Cruz (as many already have) on caucus night. We’d assume the discussion would look something like this: “Gee Bob, I know you’re supporting Dr. Carson, heck, I was a Carson supporter too. The problem is we really need to unite around a campaign that can go the distance. That’s why I decided to go with Cruz instead.” That’s going to be a fairly compelling argument to Carson’s remaining supporters (especially given the fact that Cruz is a solid second choice with high favorables among Iowans.

It’s not clear exactly how well Donald Trump will be able to turn out his caucus-goers. Even the billionaire has admitted that he’s bringing a lot of people who are new to caucusing on board for the first time. This is a fact that several commentators have run with to show that Trump’s numbers are inflated. Without any hard numbers, though, we assume that Trump’s 22% is a fairly hard floor.

Now, getting to the title of the post: how does Trump’s 22% equal a downfall? Because 22% would have put Trump in 3rd place in both of the previous caucuses. In 2008, it was Huckabee with 34 and Romney with 25; in 2012, it was Santorum and Romney both with 25. Notice something about both of those races? The conservative-evangelical vote coalesced around one candidate (Michele Bachmann was polling far higher than her 5% Iowa finish in 2012) and the “establishment” candidate came in second. In 2012, Dr. Ron Paul deflated the overall numbers of both candidate by drawing in about 22% support.

What does that tell us about Iowa of 2016? If Donald Trump’s support remains constant at 22%, there’s a case to be made that Marco Rubio’s “combined” number of 28% could be a more likely outcome than his 12% projection. Other “establishment” candidates (Bush, Kasich, Christie) are polling at a combined 9%. There are also 11% of voters who are either not sure or uncommitted. On caucus night, when other establishment voters find that their candidates’ support is soft, you can expect them to support Rubio. Likewise, given the hard support we are seeing among Trump supporters, it’s not unlikely that the “uncommitted” or unsure voters are wary of supporting Trump. Now, the big question is: will those unsure voters wind up caucusing for Rubio, or will they break in several different directions (or go for the conservative-evangelical apparent choice in Cruz)?

Here are the caucus-night scenarios:

  1. The polling is dead-on. Cruz ekes out a narrow victory over Trump. Trump loses a little bit of steam but is able to continue strongly into New Hampshire. Rubio and Carson are nearly tied in a statistically distant third and Rubio has a tough time separating himself from the pack in New Hampshire.
  2. Trump pulls out a slim victory over Cruz. This catapults him in the polls and he is able to push the story that he has a mandate and goes into New Hampshire as the frontrunner. Cruz will almost certainly skip New Hampshire and try to wager everything on South Carolina and Super Tuesday.
  3. Trump’s support stays at 22% and he loses to Cruz and Rubio. This pushes Rubio into the spotlight as the establishment choice and he quickly becomes the favorite in New Hampshire.
  4. Something totally unpredictable (like what Rand Paul is saying: all the polls are wrong and people want to get on board the RandWagon).

As we discussed at the outset, scenario one seems to be the most likely, but given the apparent stagnation of Trump and the potential upward momentum of Rubio, there’s a growing possibility that a Rubio 2nd place finish in Iowa would completely doom the Trump campaign.

How would it doom Trump? As noted in scenario three, Trump is not only a loser after Iowa, he loses to an establishment figure. Rubio, then, has the political capital to start to force his was to the front of the establishment pack and call for the others to drop out of the race and endorse him. Frankly, Rubio’s beating Trump has less to do with Trump than it does with Rubio clearing the establishment lane. As we’ve noted before, the combination of establishment support in New Hampshire is higher than Trump’s polling. However, because all of the establishment candidates are clustered together, none of them want to drop out as each sees himself being the establishment candidate. A strong Rubio finish in Iowa (especially beating Trump), gives him a clear shot to stop Trump in New Hampshire.

Losing the first two contests would not likely cause Trump to drop out (or worse, choose a third party run). Rather, he may fight all the way through the proportional stage of delegate allocation to see if he can make back his losses. Unfortunately, without any good news coming out of Iowa or New Hampshire, Trump will have a tough time convincing the delegate-rich South that he’s a better candidate than Cruz. In the end, if Trump isn’t able to buoy himself in South Carolina (where both Cruz and Rubio have strong endorsements), Trump will deflate to a distant third place in the delegate counts.

Will what we’ve talked about happen? It’s not the smart bet right now, but it’s a scenario that can be supported with the data we are seeing.