There is an interesting bit of analysis in the Washington Post on the Louisiana primary. As I pointed out earlier today, the Louisiana polls understated Ted Cruz by an average of 11 points. The Washington Post set out to try to find out why, and they present a plausible explanation.
One of two things happened in Louisiana. We know that the margins between the top three candidates in the state shifted dramatically between votes cast by absentee ballot and those cast on Saturday, the day of the election. That means that either that: 1) A candidate had a very strong get-out-the-vote effort, or 2) There was a broad shift in attitudes about the candidates.
Actually, they had three options, the third being the polls were simply wrong for any number of methodological reasons, but, being the media they chose no to go there because polls sell newspapers.
What they did was look at absentee vote tallies and then compare those to the final vote. This is what they found:
Keep in mind that the graphic represents changes in vote percentage on election day versus absentee votes, it has nothing to do with winning/losing a parish.
What you see is that Ted Cruz made huge gains over the absentee tallies while Rubio and Trump both saw their vote percentage reduced. When you look at Beauregard Parish, election day reduced Trump's percentage of the votes by over 30%.
They conclude something else was at work:
This suggests that the shift probably wasn't a function of Ted Cruz's (clearly strong) get-out-the-vote effort. Field efforts like that result in relatively limited swings, and it's hard to see how they could have run a hugely successful turnout effort throughout the state uniformly.
Instead, this looks like the state of Louisiana bailed on Marco Rubio in favor of Ted Cruz. Which could explain why Cruz is targeting Florida all of a sudden. On Saturday night, Donald Trump called for Rubio to drop out of the race. If he can repeat what he did in Louisiana in Florida in just over a week, Cruz will take Rubio out himself.
This may or may not be true. The fact that election day votes aided Cruz and hurt both Trump and Rubio indicates something was afoot. The failure to consider the simple fact that the polls were wrong can account for the horrible numbers but it cannot explain the election day enthusiasm advantage for Cruz.