2016 Democrats, 2016 Republicans, Predictions, Uncategorized

Forecasting Kansas: Final Prediction

March 5, 2016

Welcome to Super Saturday. Between today and tomorrow, Kansas, Kentucky, Nebraska, Louisiana, Maine, and Puerto Rico weigh in. Kentucky and Puerto Rico have nothing to say about the Democrats yet. Nebraska is holding off on the GOP.

Still, we have plenty to find out. Will Bernie Sanders continue winning in many places with fewer African American voters, while getting destroyed where there are more? Can Ted Cruz begin closing the gap with Donald Trump?

Will Marco Rubio or John Kasich step forward? We begin searching for these answers in Kansas, which caucuses today.


Bernie Sanders 53.8%

We only have one poll to work with. It’s 10 days old and shows half the voters as undecided. Hillary led 33/23 among those who picked someone. If this was Alabama, I’d figure the undecided voters would choose her.

It’s not. Oklahoma had the next largest amount of unsure voters (side note: polling methods impact this too.) The vast majority of them supported Bernie on Super Tuesday or  chose none of the above.

Somehow the two candidates combined for closer to 90% than 100% without any other available choices. Until proven otherwise, I’m going to overreact to this one data point, and assume we’ll wind up a bit short of 100% this time too.

The Iowa caucus is unique. Hillary began organizing in full a year ahead. Turnout is smaller than a primary state, higher (as a percentage of registered voters) than normal caucus states.

As soon as she defeats Bernie in a mostly white, lightly attended caucus, I’ll start picking her to win one. Extrapolating old polling would have placed her ahead of him in Minnesota and Colorado. He won both rather easily.

Hillary Clinton 42.9%

Hillary’s campaign team is already focusing on how she will win by more in Louisiana than Sanders does in Kansas and Nebraska, meaning she will collect more delegates on the day even if he wins more often.

You can call this spin, but it’s also true.



Ted Cruz 36.9%

When I started playing with delegate estimators 10-12 weeks ago, this was a state I usually gave to Ted. It’s located near a bunch of other places I thought he could/would win.

Unlike Georgia, Alabama, et al, where Trump was already ahead of him in polls, we didn’t have any evidence one way or the other here.

The first survey came out before Super Tuesday and showed Trump well ahead of Cruz and Rubio, with Kasich an asterisk. Only one poll, and from Fort Hays State University, meaning it was from a source that only surveyed Kansas.

No ability to check previous polls in the state, or their work in other states to see if Cruz had improved or declined, if they tend to favor Trump or not. Taken at face value, Ted had work to do, with plenty of undecided voters remaining.

Then he exceeded expectations on Super Tuesday. Another poll was taken by Trafalgar Group. This one had Trump narrowly leading Cruz 35/29 and Rubio a few points up on Kasich 17/13.

Trafalgar is surveying other states voting on March 5 and 8, and seem to have no problems finding large leads for Donald Trump. As such, I’m inclined to take the narrow one here seriously.

The debate happened after the survey period (technically very late in the evening on the second and final day). It’s my belief Cruz was helped. He may have picked up some remaining undecided voters as well as stealing a few Trump leaners.

By itself, that was likely enough to close the gap. But there are a couple other factors. This is a closed caucus. Trump does better in open primaries. Cruz is the reverse. Ted virtually always out-performs his polls. The Donald only does in Trump-friendly states.

I think Ted has this one pretty well in hand. I suspect if I’m wrong, it’s in understating the margin between the top two.

Donald Trump 31.1% 

He skipped out on his speech at CPAC in favor of large rallies in Kansas and Florida. Some have suggested The Donald didn’t want to deal with conservative opposition at the conference, but he could have chosen to go anywhere to hold an event.

Florida is obvious. He wants to finish off Little Marco on March 15. Kansas is just one of several places about to vote. This is a sign he thinks the state is competitive too. Holding a mega rally is his version of a ground game.

When Trump only polls a little ahead of Cruz in the final poll somewhere, Ted tends to win. His victories are when he has a huge edge. At this point, 30% probably represents his floor, except in places like Minnesota that are distinctly unfriendly to him, or like Texas, have a candidate in the race.

John Kasich 16.4%

If you’d asked me three days ago, I would have given him a number in single digits. He climbed to 13% in the pre-debate poll. While I don’t expect his avoidance of the carnage to send him immediately rocketing up the ladder, it should score him a couple/few more percent.

There are only so many moderate votes to go around, though Kansas has more of them than you might think. Rubio will get his share, so there’s a cap on what we might expect from Kasich.

If everything breaks right, maybe he gets to about 20%, but I’m more comfortable here. He hasn’t regularly surpassed his polling. The Vermont surprise was without current data.

Marco Rubio 15.5%

This number is a problem for the Rubio campaign. As you can see from the close positioning, I’m figuring Rubio and Kasich are a coin flip today. That in itself signals danger. He can’t catch Cruz and Trump if he’s a rough equivalent to a supposed also-ran.

Trafalgar had him at 17%. The poll was taken after his disappointing Super Tuesday. Dropping him much below this isn’t logical. In a state like Kansas, Marco is dependent on a decent amount of well-off, educated, somewhat moderate voters.

I think a few of them likely migrated to Kasich after the debate. In some places, he actively challenges Cruz with evangelicals and strongly conservative voters. The Trafalgar cross-tabs don’t indicate this here.

Cruz probably had the stronger debate, so there’s no reason for them to migrate his way now, especially when Ted is seeming more viable than he did a week ago.

To show you how variable the Rubio experience is, when I did my preview a couple days back, I thought he could win the state. Now I’m officially picking him last.

If I’m right, this is a huge blow to his hopes.


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