Republicans are Out Performing Democrats in Florida early voting and out pacing 2012 Republican Early Voting

In 2012 Time Magazine and Politico both reported Democrats had an edge on early voting in Florida:

• Politico
• Time

According to Politico the results were 43% Democrat, 40% Republican and a total of 4.3M ballots were cast or mailed in through November 5^th.  ^{Note that by election day a few more ballots came in, but I cannot find a reference to final number of absentee/early votes in 2012.}

Doing the math, we have the following results for 2012 early voting:

According to the Florida Division of Elections website, the actual results for 2016 (as of November 2^nd) are:

^{*Note the first line in the above table, these are ballots that have not yet been returned and are not included in the “Total Votes” row}

Comparing to the 2012 numbers as a percent of the total, Republicans are flat and Democrats are down 3%.  The total vote count shows that Democrats are lagging behind their 2012 totals (still could catch up to that number) while Republicans have exceeded the 2012 figure.

The numbers look good for Donald Trump, but by themselves do not prove anything, however, they offer some insight into the polls and can prove that the methodology (at least in Florida) the pollsters are using does not match the actual voter turnout, specifically they are over sampling Democrats.

The most recent CNN and Quinnipiac polls show a tight race, Clinton is +2 and +1 respectively:

The methodology the two polls are similar:

And the breakdown of who’s voting for who is similar between the two polls, most Dems are voting for Clinton, most Republicans are voting for Trump and Trump has a slight edge among Independent voters:

The internals of the Quinnipiac poll show that 42% of the people surveyed already voted and that 48% voted for Clinton while only 42% voted for Donald Trump.  How is that possible if:

• According to Quinnipiac’s own polling data 86% of Democrats are voting for Clinton and 88% of Republicans are voting for Trump
• The beginning of this post showed Republican’s actually have a slight lead in early voting

The answer is that the pollsters are using 2012 turnout models, but voters are not turning out according to their models.  Using Quinnipiac’s own party to candidate break down, the early/absentee voting results to-date should actually show Trump as +3, not the 6 point lead it claims for Clinton:

Essentially what this means is that the methodology of +/-6 for Democrats is bogus, according to early voting data they should be using +0.  In 2012, Obama won early voting by 5%, but carried the state by less than 1% due to more Republicans/Romney voters turning out on Election Day.  If that trend continues, then the pollsters are way off and Trump will win Florida by more than 5%.