2016 Republicans, Poll Watch, Trump, Uncategorized

Poll Watch: Episode 18 (Dancing on the Ceiling)

December 11, 2015

Hope you enjoyed the Week of Trump.  To say he dominated the discourse (perhaps a kind way to describe recent chatter) is an understatement inside a dilution inside a downplay.

He’s both leading the GOP polls and threatening to run as an independent.  His numbers are both climbing (voters picking him as current first choice) and falling (overall favorability rating).

Republican insiders, including party chair Reince Preibus are already discussing convention procedures in case The Donald has enough delegates to cause a problem but not enough to wrap up the nomination on the first ballot.

Earlier in the week we examined the difference between likely voter polling, which shows Trump trailing Ted Cruz in Iowa and ahead in South Carolina, but with Cruz, Marco Rubio and Ben Carson in hailing distance, and registered voter polling which has The Donald sky high.

In order to have a chance to get nominated, Trump needs the broader voting base polls to more accurately reflect the truth.  To try to make that so, he does things like suggest banning Muslims from the U.S.

Let’s say it works.  Let’s assume he can continue polarizing the electorate and controlling the conversation for another couple months, managing to finish ahead of Cruz in Iowa, a state tougher for Trump than almost any other.

Let’s also assume the field consolidates fairly quickly, with most candidates out by South Carolina, at the absolute latest after March 1.

Previously I’d figured Trump had a decent shot of at least causing a brokered convention in a 4-way race, would be an underdog with 3 contenders and was virtually assured to lose a 2-way contest.

Post-Muslim ban, how do the numbers look?  If he can get a fair amount of Trumpists to the polls, is his ceiling still in the 25-30% range?

The first factor is the spread of his support, the gap between Trump’s national numbers and performance in a given state. Chris Christie is at up to 12% in New Hampshire, 14% in New Jersey, but scores in the low single digits overall.

He runs 10 points and about 5x better in his best states.  Cruz and Rubio each do about twice as well in their best states as their worst, with the national average in between.  Rubio has a little less variation than Cruz, but it’s similar.

You get the idea.  Jeb Bush and Ben Carson are about the same as Cruz and Rubio.  John Kasich is a less-effective version of Christie.  Nobody else is polling above 5% on a state or national basis.

Trump is the outlier.  While the style of poll matters greatly, the location does not.  If he can’t turn his supporters out, especially in Iowa, we have nothing to discuss.  His best recent result with tight sampling is 27% (national, South Carolina), his worst is 19% (Iowa).

Less of a range than his opponents and not enough to win, especially when you consider he’s not a frequent second choice.  Say hello to a Cruz v. Rubio contest if this happens.

Under the other scenario, Trump is incredibly consistent and at a higher level.  Just since Monday, he’s posted the following results:

Iowa (CNN) 33%

North Carolina (PPP) 33%

New Hampshire (CNN) 32%

National (CBS/NYT) 35%

New Jersey (Rutgers) 30%

South Carolina (CNN) 35%

Very consistent.  If his voters show up.  There is a drawback to being this consistent.  From March 15 forward, delegates are mostly winner-take-all.  While he would compete everywhere, finishing in the low-mid 30s does not guarantee victory in a 3-person race.

However, it is great for momentum in early states and would lead to winning all or most of them.  So a ceiling there would keep him in contention after March 1, but guarantee little.

The thing is, that’s not his ceiling.  It’s higher.  To begin with, close to 100% of voters pick someone at the polls, but less than that number pick someone while being surveyed.  Anywhere from a few to  more than ten percent are undecided, depending on how much the pollster wants to push.

In the CBS/NYT poll, Trump is at 35% out of 90% who chose someone.  Unless you want to assume undecided voters are completely unlikely to choose Trump, his actual support share is just shy of 39%.

The CNN South Carolina poll adjusts to 38%.  Others have smaller or similar adjustments.  If you figure The Donald could pick up even a few voters from candidates dropping out, or voters deciding Carson has little chance, he quickly gets to 38-42% in his average state.

There is a tremendous difference between 32 and 39 percent in a race with 3 real contenders and a couple others draining a few votes.  At 32, he trades off sneaking out a narrow win, losing to Cruz, losing to Rubio.

At 39 he almost always wins, with Rubio and Cruz taking turns making it close.  With Trump already incredibly momentum dependent, it’s all the difference in the world.

Should he start off by winning 16 of 20 states through March 15, he will get even more of his voters to the polls while mainstream Republicans make apoplexy a household word for the first time in a century.

There is at least a 50% chance the conservative turnout models are way more correct, making this moot.  If not, Trump’s ceiling is already high enough to get nominated unless all but one opponent are out before the end of March.  Even at that, it’s way too close for RNC comfort.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s