In person early voting concluded last night (mail in absentees can still be accepted up to the day before Election Day), and it broke some records. What did seven days of runoff early voting tell us?

Strong turnout

As of last night, 489.649 Louisianians either early voted by person or by mail in ballot (451,171 in person, and 38,478 mail in ballots). To put this number in perspective, this is the highest early voting turnout EVER for a non-Presidential election. Not only did it exceed the volume of in person early voting for the primary by 31% (or 115,459), but it was 91% (or 232,628) higher than day in person early voting for the 2015 runoff. The top five (by early voting volume) early voting parishes were East Baton Rouge (58,249 early/absentee votes), Orleans (39,260), St. Tammany (35,097), Jefferson (31,523), and Ouachita (20,854).

Democrats’ strong finish

Whatever the turnout levels are from election cycle to election cycle, black voters (who are almost unanimously Democratic) tend to show up in greater numbers on the last day. But very rarely do they turn out to the extent that they did yesterday: a whopping 40% of the last day early voters were black (blacks represent 31% of Louisiana’s registered voters). As further illustration of how unprecedented black participation this strong is, in the 116 days of available data for in person early voting going all the way back to 2008, only four other times has the black early vote as a percentage of the total vote ever hit 40% (and three of those times were in 2008, when Barack Obama was first elected).

This 40% black electorate was in addition to stronger (relative to the primary) black early voting for the previous six days which never dropped below 29%. This means that the final early vote is 31% black – a figure 6% higher (percentage-wise) than in the primary and 1% higher than in the 2015 runoff.

Does this 6% increase in the black share of the early voting electorate equate to a similar six point boost for Governor Edwards, though? Because at the same time that the black share of the electorate increased 6% relative to the primary, the Republican share only decreased 3% (it was 38%, compared to 41% in the primary). The remainder of the decrease came from lower turnout (percentage-wise) from white Democrats and white Independents. And since Governor Edwards has been running strongly among white Democrats (60-37% in the last poll JMC conducted and publicly released) and relatively well among white Independents (he trails 38-56% in the last poll), the incremental benefit to his campaign isn’t as much as it would seem, because the Republican percentage of the electorate remained relatively high, and among this group, Governor Edwards trails 17-79%.

What all of this analysis means in practical terms is that because of (1) the decrease in white Democrats/white Independents and (2) the relatively high Republican early vote, an Edwards poll lead of 50-47% over Rispone would become 50-46% Edwards/Rispone if the poll results were recalculated to reflect the demographic composition of the early vote.

JMC’s projections of early voting volume, overall turnout

Projecting turnout is a constantly moving target throughout early voting week, but since early voting has been in existence in Louisiana for over a decade, JMC has established (and continuously refined) benchmarks that can be used to project early voting and/or final turnout, even considering that this predictive model got slightly more complex starting in 2017 with mail in ballots’ starting to be automatically being mailed out to those 65 years old or over who opted to be on this list.

Given the above, as well as the fact that early voting twice shattered numerical records this election cycle for non-Presidential early voting, there are three “parallel universes” to JMC’s turnout projections (which consider the fact that an estimated 10K absentee ballots have yet to be mailed in and tallied before the Friday deadline), with one of those three being the likely outcome.

Parallel universe #1: In the October primary, 28% of the total vote was cast before election day (which itself was a record). If we want to assume that the higher early voting turnout relative to the October primary is indicative only of a surge in Election Day turnout, a 28% early vote equates to a 60% voter turnout, or a total vote of 1.78 million. Because this is a Presidential level of turnout (Louisiana’s Presidential turnout in the last three Presidential elections was 68%, 68%, and 67%). JMC does not see this as a realistic scenario at all.

Parallel universe #2: If instead we want to assume that the higher early voting turnout is solely indicative of “front loading” of the total vote with no increase in numerical turnout relative to the October 2019 primary (whose turnout was 1.36 million), an estimated 499K early vote (489K who already early voted plus an estimated 10K absentees yet to be mailed in) represents an early vote that is 37% of the total primary vote. It would be a logical stretch to think that early voting (which is slowly but surely becoming more popular) would jump from 28 to 37% in a month’s time. Again, not a realistic scenario at all.

Parallel universe #3: More realistically, there will likely be a combination of higher turnout and some increased interest in early voting. In other words, an early vote representing 33% of the total vote (33% being halfway between 28 and 37%), which calculates to a total turnout of 51%, or 1.51 million voters).

In fact, this scenario was what happened in the primary. JMC used similar reasoning to project a 48% turnout and an early vote of 27% of the total vote, and the actual numbers ended up being a turnout of 46%, with 28% of that vote being cast before election day.

In Conclusion

Why does early voting matter? When the Legislature essentially established “no fault” early voting more than a decade ago, you now have a noticeable constituency of people who prefer the convenience of early voting, and this constituency has for nine times in a row (the 2015 primary, 2015 runoff, 2016 Presidential elections, December 2016 runoff, October 2017 primary, November 2017 runoff, November 2018 primary, December 2018 runoff, and October 2019 primary) exceeded 20%. A politician would be foolish to ignore this many “up front” voters, especially in a closely contested race. Also, too, early voting numbers are the first ones that are typically reported after polls have closed at 8 PM.