April 29, 2016
Rumors of Ted Cruz’s demise are somewhat exaggerated. At least in Indiana. His path forward from there is narrow and bumpy, with potential drops off a steep cliff. But for now, we’re concerning ourselves with Hoosier State polls. Cruz needs to survive this one before we worry about his chances elsewhere.
We still have limited data to work with. Real Clear Politics shows a total of four surveys. None were completed before New York voted. Only one was taken after the April 26 Trump tsunami. It’s a one day poll from Clout Research on the 27th.
The Carly announcement was later in the day, we don’t know exactly how many respondents were surveyed before they heard the latest news. Things have happened in the 36 hours between the poll and now. John Boehner called Cruz Lucifer. Indiana Governor Mike Pence endorsed Cruz.
Perhaps these items cancel out. Maybe Boehner’s scorn actually helps Cruz. He used it for fundraising, though those are his core supporters. I don’t want to speculate on whether he’s making progress or falling back until we see a couple more post-New York surveys.
However, we can determine Trump does not have this in the bag yet. John Kasich is in the upper teens. His RCP poll average is 19.3%. The Clout poll is the only one taken after he pulled back from advertising and campaigning in the state. He’s at 16% there. None of the other surveys have him above 22%.
While he has a definite floor, probably about 15%, Kasich is not likely to clear 20% on Tuesday. Even if his absence doesn’t hurt him, it won’t help. Even if none of his voters go to Cruz, he won’t pick up plenty of new ones. He also virtually never exceeds his final poll average. We shouldn’t expect many undecided voters to pick him.
If we’re mostly sure where he winds up, this gives us 80 to 85 percent to distribute among Trump and Cruz. This keeps it simple. If Trump can get near 45% he wins. If he winds up in the 38 to 41 range, Cruz can top him.
The pre-New York surveys had Trump at 37, 41 and 40 percent respectively. Right on the line. No margin of error for Cruz. Prior to New York and April 26th, Trump tended to retain his final week poll support on Election Day, but rarely added to it. Cruz regularly beat his numbers by a few points, as long as it was a state he was competitive in.
Of late, Trump has exceeded his final numbers. Maybe it’s a change in the race, maybe it’s due to him having more of an advantage in places that are already favorable to him. Maybe (and this is my guess) both. So we can’t assume a couple/few point Trump advantage equals a couple/few point Cruz win the way it did in places like Iowa and Oklahoma.
We also shouldn’t expect undecided and leaning Indiana voters to act the same way as their peers in Rhode Island. Among more balanced states, Wisconsin was a positive loop for Cruz, Pennsylvania for Trump.
The Scott Walker endorsement, combined with support from talk radio, and perhaps a desire to keep the nomination process open, helped Cruz pull the majority of late deciding voters in Wisconsin.
However, in Pennsylvania, Cruz lacked institutional support, and Trump was able to leverage the extra complicated state process and his pitch about a rigged primary system to push wavering voters in the direction of shutting things down and choosing Trump.
In both cases, there were polling cues ahead of the final week. In Wisconsin, Cruz extended his lead in during the 7 to 14 days out time, going from a very narrow advantage to a statistically significant one. In Pennsylvania, Trump went from having an edge to looking like a clear favorite. The final week magnified the previous movement in both places.
If the Clout survey is as accurate as the others, Cruz actually narrowed the gap slightly in the second to last week before the vote. They have him trailing 37/35. Not only is it the closest margin (the previous polls were Trump +6, +8, +5), but it ties Trump’s worst poll number and Cruz’s best. The other Cruz 35% result was from CBS/YouGov, which consistently gives him higher than average numbers.
While I’m always wary of reading too much into a single survey from a pollster that didn’t previously poll the state, if the April 26 results affected Hoosier voters it was to push them toward Cruz and extending the race, not Trump and ending it.
Trump was still ahead. He led each of the four surveys we have. After closing well recently, he’s definitely within range of the low-mid 40s result that would win him Indiana and virtually clinch the nomination. But Cruz is in range too, and there is no evidence yet that he’s losing any ground.
Clout asked about favorability too. They were kind enough to break this into very and somewhat favorable, very and somewhat unfavorable. I tend to assume that (at least when there are 3 or fewer candidates) a candidate will get almost everyone who feels very favorable, and almost nobody who doesn’t think at least somewhat favorably.
If we treat very favorable as a floor, we get:
This reinforces the idea Kasich is going to retain support somewhere in the mid-upper teens. It also shows Cruz has as strong a core base as Trump.
If we look at very + somewhat favorable as a ceiling, we get:
With Kasich a clear third place choice, he’s likely restricted to people who view him very favorably, or those who are somewhat favorable on him and unfavorable on the two others. Voters fitting that description may pass on participating though.
Again, Cruz and Trump aren’t separated by much. Depending on turnout and anything that may happen over the final few days, either of them can reach the safe zone, neither of them are guaranteed to. With Cruz focusing intensively on Indiana, his candidacy hanging in the balance, and polls indicating a close race, there’s no reason for strong supporters of either candidate to stay away thinking their vote won’t matter.
Clout’s results show 68% of voters are very firm about their choice, 24% are somewhat, and 8% are not at all. That last group is going to decide this.
I’m not sure how many polls we’ll see over the final few days, but keep your eyes peeled for anything that shows Trump in the 43 to 45% range, or Cruz breaking 40. Either of those would indicate a break for one candidate or the other.