This must be brief, but I’ve been able to merge the registration and voter history files in Florida from 2010 with the early voting records from 2014.  Here are some first-cut comparisons at the individual level.  (Remember, another quarter million Floridians are still yet to vote early in this cycle.)

The thing that jumps out at me is that Hispanic and Black early voters in 2014 tend to be drawn more from non-voters in 2010 than whites.  In addition, there is evidence that among voters in 2010, Black and Hispanic Election Day voters are more likely to be voting early than white Election Day voters.

The other thing is that among those not  registered with a party, the early voters were disproportionately non-voters in 2010, compared to registered partisans.

Here are the notes I’ve made:

Among early voters thus far, how did they vote in 2010?

15.4% did not vote
5.9% voted absentee
45.0% voted early
33.5% voted Election Day

Among blacks, 19.3% did not vote, 4.0% absentee, 39.2% early, 37.2% election day
Among hispanics, 24.7% did not vote,6.0% absentee, 38.8% early, 30.4% election day
Among whites, 13.3% did not vote, 6.3% absentee, 47.1% early, 33.2% election day

Among Dems: 16.9% did not vote, 4.5% absentee, 42.3% early, 36.2% election day
Among Reps: 10.5% did not vote, 7.8% absentee, 50.4% early, 31.2% election day
Among NPAs: 24.7% did not vote, 4.7% absentee, 37.8% early, 32.6% election day
Among early voters in 2010, have they voted already in 2014?

Overall, 33.6% have already voted early
Among whites, 34.1% have already voted early
Among blacks, 35.0% have already voted early
Among hispanics, 27.8% have already voted early

Among Dems: 34.1% have already voted early
Among Reps: 34.3% have already voted early
Among NPAs: 29.8% have already voted early