So far this season, the one that has seemed to be most accurate in terms of being a predictor of where things were headed is PPP. But we'll see. Three IA polls in last week, two good for Cruz, one good for Trump. I'm betting IA is a lot closer with a very slight Trump lead.
Either way it's clear that TEMPORARILY Cruz has picked up almost every disaffected Carson supporter. On the one hand that confirms my view that Carson's support was almost entirely "church lady" from the get go, but refutes my view that Cruz and Trump would split these evenly.
In about a month, I will run my Iowa Simulation for 2016.
Essentially, I go county by county and look at the results from the General Elections of 2004, 2008, and 2012, and Caucus Turnout for each county, the percent of the GOP Vote in the county compared to what turned out in the Caucus, and make estimations of what the turnout will be in the Caucus and General Election.
IA has 99 counties and I’m about 55 counties through so still a long way to go. But I know enough of the counties and the make-up of the voters to know what constitutes and “Establish Voter”, “Evangelical Voter”, “Ron Paulard”, etc.
They say turnout will rise in 2016. How much is still in the air. Some say 140,000 (from 120,000 or so in 2008/2012). Right now, I’m looking at 135,000. Reason is the GOP has gotten a little smaller in IA from economic issues in the State and many in the West left for other states in the past decade.
At 135,000 people I’m pretty certain that:
(1) Given Ted Cruz’s endorsements, he’s headed for a finish somewhere between 35-40% IF nobody else leaves the race. If Carson drops out, it could go higher. But this gives him an estimated vote total between 47,250 and 54,000. This total will be higher than what Huckabee got in 2008.
(2) Trump is banking on new attendees, former Paul voters, and little slices of everyone else (Establishment, Evangelicals, etc). Ron Paul got around 20% with organization. Based on the numbers, Trump looks to be capped in IA at 27% or just shy of the number of votes Huckabee got when he won in 2008. I don’t see Trump getting more than 37,000 votes in IA and if the weather is bad (blizzard on 1 Feb?) that always effects turnout.
(3) Rubio (or whomever the Establishment choice is) seems locked in at 13% this is essentially where he was last time. If Jeb and Christie (less likely) get out, his numbers could go up, but he is probably capped at 20% in a best case scenario.
(4) I’ve seen what happened to Ben Carson happen to Michele Bachmann last time. She tumbled all the way to 5% after everyone consolidated around Perry, then Cain, then Santorum at the last minute. Ben’s floor is probably slightly higher but I don’t think he’ll break 10% on Caucus night, unless Huckabee and Santorum drop out.
(5) Bush, Carly, Rand Paul, Huckabee, Santorum are finished in IA.
The last time I ran the simulation in 2012, the numbers showed a razor thin result between Romney and Gingrich (less than 60 votes). I nailed Romney’s total, but Santorum switched spots in Gingrich’s place, Bachmann switched places with Santorum, and Perry/Gingrich split the middle that originally just Perry.
Bottom Line:
(1) The Des Moines Register Poll shows Cruz with a 10,000 vote advantage to Trump (39,000-29,000) based on a turnout of approximately 135,000. If we follow the percentages to the eventual finish in February, this margin seems about right, but Cruz’s lead could grow whereas I think Trump is capped at 37,000 votes.
(2) Rubio doesn’t have to much to stay in 3rd
(3) Carson’s numbers are turning into Bachmann’s and the problem is once they’re gone, they’re gone.
Either way it’s clear that TEMPORARILY Cruz has picked up almost every disaffected Carson supporter. On the one hand that confirms my view that Carson’s support was almost entirely “church lady” from the get go, but refutes my view that Cruz and Trump would split these evenly.