Get it while it’s hot! “Election Daze: Voting Modes and Voter Preferences in the 2016 Presidential Election” (with Seth McKee and Trey Hood)

Daniel A. Smith, Seth C. McKee, and M. V. “Trey” Hood III, “Election Daze: Voting Modes and Voter Preferences in the 2016 Presidential Election,” Florida Political Chronicle 25:2 (2018), 123-141.

ABSTRACT: To say that the outcome of the 2016 Presidential Election was a surprise to many political experts would easily qualify as an understatement for the ages. Nonetheless, in defense of the political handicappers, there is notable evidence that the dynamics of voter choice in the days leading up to the last day of voting were differentiable from preferences registered on Election Day. That is, in some states it would seem that Hillary Clinton (Democrat) was advantaged by early voting and Donald Trump (Republican) was favored by voters who came to the polls on Election Day. This paper draws on aggregate- and individual-level data from Florida to examine voting across racial/ethnic groups, distinguishing between votes cast on Election Day with those cast early in-person and by mail in the 2016 Presidential Election. The paper also compares variation across modes of voting in 2016 with 2012 county-level Presidential Election returns. By leveraging original datasets that merge the modes of voting for different groups with aggregate presidential results, as well as using 2016 Cooperative Congressional Election Study (CCES) individual-level data, we are able to determine the extent to which the story of Trump’s historic Presidential victory hinged on the support he garnered from voters who showed up on the final day of voting.

Available here.

An ElectionSmith Exclusive: Support for Trump in Florida across Predominantly White Precincts, by Method of Vote

Still trying to figure out how to best visualize Trump’s overwhelming victory on Election Day in Florida. The following panel of figures draw on precinct-level data that our Election Sciences team at UF has helped me collect. The data presented here show only the precincts in the state in which white voters made up at least 75% of the votes cast for each method of voting (Election Day, Early In-Person, and Vote-by-Mail), and that had at least 100 ballots cast for each method.  In other words, these are heavily white precincts where whites cast at least 75% of all ballots cast for each method of voting.

Running along the X-Axis is the proportion of the overall vote that Trump won (so, from 0% to 100%). The height of each vertical bar (there are a total of 60 bars) constitutes the fraction of the total votes Trump received — that is, add them all up and they equal 100% of Trump’s total votes by each method). The red vertical line in each panel is set at .5, or precincts in which Trump won 50% of the total votes cast.

As should be clear, Trump crushed it Election Days in these predominantly white precincts. The peak of the very tight normal distribution curve for Election Day voters in precincts where three-fourths of the votes cast were by white voters is greater than 65%.  Trump did not do nearly as well (relatively speaking) among vote-by-mail voters in predominantly white precincts; he fared better with predominantly white precincts whose voters cast early in-person ballots.

Stay tuned for similar graphs with Trump’s support among predominantly Hispanic precincts, broken down by method of vote.

Florida 2016 Presidential Vote by Precinct, by Method of Voting

(In each graph, the X-Axis is a precinct’s Election Day or Vote-By-Mail or Early In-Person votes cast out of total votes cast in that precinct by all three methods. The Y-Axis is Trump’s and Clinton’s share of total votes cast out of votes for all presidential candidates for each of the three methods.)

White Democratic Election Day (and Early In-Person & Vote-by-Mail) Support for Trump in Florida

In his detailed article today, New York Times Upshot’s Nate Cohn asks the following:

“Is it possible that the registered Democrats who turned out were Trump supporters, and that the Democrats who stayed home were likelier to be supporters of Mrs. Clinton? Perhaps, but our polling suggests the opposite. In our pre-election Upshot/Siena polls, voters were likelier than nonvoters to support their party’s nominee.”

My preliminary analysis from Florida suggests otherwise. Along with my UF Election Sciences team, we painstakingly collected precinct-level election results for Trump and Clinton by method of voting (early in-person, vote-by-mail, and Election Day) from the nearly 6,000 precincts in Florida’s 67 counties, parsed them and cleaned them up, and then merged the presidential results with precinct information gleaned from the Florida voter file. Since we know which method of voting Florida voters used to cast their ballots, we’re able to make lots of different kinds of precise estimates. One goes to the heart of Nate’s question, at least for Florida.

The plot below shows the proportion of total votes cast on Election Day by White Democrats in each precinct, and the vote share for Trump and Clinton won on Election Day.  The horizontal (X-axis) reveals that the range of White Democratic Election Day voters making up a precinct’s total Election Day votes ranges from 0% (e.g., heavily Republican, or Hispanic and Black precincts) to a handful that top 40%.

The precinct dots (weighted by Election Day turnout) and the accompanying lowess fit lines suggest a telling pattern: Trump generally won a greater share of the Election Day vote in precincts where the share of White Democrats exceeded 5% of the total Election Day votes cast on Election day. The greater the share of White Democratic Election Day voters, the greater Trump’s share of the vote.

(This is the identical plot, just with Trump’s precinct vote share overlaying Clinton’s precinct vote share.)

Interestingly, the pattern doesn’t exist for either Early In-Person or Vote-by-Mail votes cast in a precinct by White Democrats (see below), suggesting that some 11th hour surprise (hmmm, I wonder what that could have been, James Comey?), may have influenced the vote choice of White Democrats who waited until Election Day to cast their ballots.

The New York Time’s @Nate_Cohn finds “Turnout Wasn’t the Driver of Clinton’s Defeat,” but in FL, Trump crushed Clinton in precincts that had high turnout and in those where White’s cast greater share of ballots on Election Day

Nate asks, “So how much did turnout contribute to Mr. Trump’s victory? As the party registration numbers and turnout figures by race imply, just a bit. But Mr. Trump won the election by just a bit — by only 0.7 percentage points in Pennsylvania, for example.”

That seems reasonable.

Let’s take a closer look at the precinct results in Florida.

Trump crushed Clinton across Florida’s precincts that had high voter turnout, as the figure below reveals (with the size of the precincts weighted by the total votes cast).  Trump won a greater share of the total votes cast in the smattering of low turnout precincts, but more impressively, in precincts with turnout (of registered voters) that exceeded 75%. For example, he won more than 60% of the total votes cast (which includes write-ins and 3rd party votes) in precincts with greater than 90% turnout.

Trump also crushed Clinton in precincts that had high rates of Whites voting on Election Day. The figure below reveals that Trump won bigly in precincts in which Whites comprised more than 65% of those who waited until Election day to turn out (precinct size is weighted by Election Day Turnout).

So, it appears that that Trump did well in Florida not only in precincts with high voter turnout, but also in those where White voters turned out on Election Day.

More to come…