White voter turnout in Florida from 2012 to 2016 jumped by nearly 4 percentage points, from 73.1% of active voters casting ballots in 2012, to 77.1% of active voters casting ballots in 2016.
Black turnout, which was a major concern for Democrats with President Obama no longer on the ballot, was down by 3.3 percentage points, from 72.3% to 69.0% of active voters.
Hispanic turnout, which saw a dramatic increase from 2012 to 2016 during the early voting period in Florida, was up overall, from 63.1% of active voters casting ballots in 2012, to 68.9% in 2016.
Overall in 2016, Florida’s electorate was less white in 2016 than four years ago. Whites comprised 68.4% of the electorate in 2012; in 2016, they comprised 66.8% of voters. The difference, not surprisingly, is due to Hispanics making up a greater share of the electorate, from just 12.5% in 2012 to 14.8% in 2016.
It is erroneous to conclude that higher Hispanic turnout in Florida led to greater support for Trump. This is a classic example of an ecological inference fallacy, as others have shown at the precinct level in Miami-Dade County, and as my colleagues and I will be investigating more thoroughly statewide (and by method of vote cast) in the coming days.
The probability a district elects a black lawmaker in the Deep South (left panel) versus the Rim South (right panel) depends on the size of a district’s black population. Across the three decades for the given election periods, it is clear that black legislators are elected with smaller black populations in the Rim South relative to the Deep South. In 1993-1995, for example, the probability that a Deep South district elects a black lawmaker reaches 0.5 (even odds) when the black population is between 54 and 55 percent. In that same period, the probability a district in the Rim South elects a black legislator reaches 0.5 when the black population is between 49 and 50 percent. This 5 percentage-point difference nearly doubles in 2003-2005 (52 to 53 percent for the Deep South versus 43 to 44 percent for the Rim South) and in 2013-2015 (48 to 49 percent for the Deep South versus 40 to 41 percent for the Rim South).
Beyond the pressing normative views regarding the broader political and representational implications of the relationship between majority-minority districts and black representation, our empirical analysis indicates an inexorable dynamic in party politics. Our findings leave no doubt that a considerable reduction in majority-minority state legislative district populations can be accomplished while ensuring black descriptive representation. In light of the Supreme Court’s 2013 decision in Shelby County v. Holder, which scrapped the federal enforcement of the Section 5 preclearance provision of the Voting Rights Act, we expect in the next decennial round of redistricting most Democrats will push for a reduction in the size of minority populations in majority-minority districts, while almost every Republican will continue to insist that majority-black districts should remain as is, or better yet, contain even higher African-American populations.
Ahead of Election Day, 2.56m Democrats have cast ballots, 2.47m Republicans have cast ballots, 1.24m No Party Affiliates have cast ballots, and 154k voters registered with 3rd Parties have cast ballots.
3.52m women and 2.76m men have voted, with another 139k votes cast by voters whose gender is not reported.
And by race/ethnicity, 4.23m whites, 980k Hispanics, 841k blacks, and 375k voters of mixed, other or unknown race have cast ballots.
By gender, what follows are tables with the share of votes cast across party registration for each racial/ethnic group.