…and turnout is basically everything where it comes to deciding whether John Bel Edwards or Eddie Rispone wins the race.
We have three basic pieces of data to work from over the past four days. The first was the early-vote totals that were compiled following the closing of the early voting period on Saturday.
There were 489,654 early votes cast during the seven days of early voting for Saturday’s runoff in the gubernatorial race. Of those, 31.1 percent were black voters, the product of a very vigorous “Souls to the polls” operation the Democrat Party put on for Edwards. That number sparked a good deal of consternation in various Republican camps, and jubilation in Edwards’. By voter registration, that 31 percent number is virtually identical to the percentage of black voters by voter registration.
But 38.2 percent of the electorate turning out to vote early were Republicans, which is a very high percentage as well given that Republican voters are just 31 percent of the state’s registered voters.
Typically, blacks vote in smaller numbers than they’re registered, and Republicans vote in larger numbers. The early-vote numbers indicate that black voters matched their usual number while Republicans exceeded theirs.
But early voting is no longer a measure of the direction of a race, because it’s becoming more and more common for folks to vote early. There were well more than 100,000 more early voters last week than in the early voting period for the October 12 primary, and considerably more than that margin compared to the races in 2015.
So what that means is both sides are likely cannibalizing their Election Day vote with early voting. And therefore it’s impossible to tell from early voting what the composition of the electorate will be.
Going by the results four years ago one would expect the Republicans will have a big advantage on Election Day. That’s usually the case in Louisiana, as it is in other states.
That 31.1 percent figure representing the share of the electorate which is African American is one to watch, because it’s extremely high. In fact, the high-water mark in recent Louisiana elections for the black share of turnout is 30.6 percent. That came during the 2008 presidential election when Barack Obama was first on the ballot. It would be very surprising were John Bel Edwards to produce a larger result in the black community than Barack Obama does. Very surprising, indeed. In the primary, black voters were a bit under 28 percent of the electorate; it might be safe to say they’ll end up somewhere between 28 and 29 percent of the total vote in the runoff when it’s all said and done.
Why do we believe this? Because Edwards’ get-out-the-vote specialist is an old-school Louisiana operative named Scott Arceneaux, who has been in Florida for the past couple of election cycles. Arceneaux’s results indicate a great deal of electoral cannibalism with early votes in that state.
Both Hillary Clinton in 2016 and Andrew Gillum in 2018 banked heavily on early voting in Florida, and posted impressive early voting numbers, which local media reported as an indicator they’d win that state. And in both cases the results on Election Day were ugly.
And last week’s early voting in Louisiana was an indication of much of the same. Of the 152,000 or so black early votes posted last week, only 26,000 of them didn’t vote in the primary. The vast majority of them were chronic voters. On the Republican side, the new-voter number was 22,000 or so out of 187,000 votes. You’d say that’s evidence the Republicans are cannibalizing more of their vote than Edwards’ camp is with black voters, but remember – Republicans turn out in much larger numbers relative to their voter registration than black voters do, which means a far higher proportion of them are chronic voters. To only be down 4,000 votes in a comparison of new voters from core voting groups after a major GOTV effort on the part of the Democrats isn’t a bad sign for Rispone’s electability. It puts him in a solid position to win should his GOTV effort on Election Day be worthwhile.
Let’s also remember that in the primary 37 percent of the vote was Republican, and Republican candidates (Rispone, Ralph Abraham and Patrick “Live Wire” Landry) combined for 52 percent of the vote. If the 38 percent of the early vote which was Republican should hold up in the runoff, Rispone has 53 percent of the vote to work with given that ratio. Most independent voters in Louisiana lean right of center and Republicans tend to capture a third to a half of white Democrats, who are people who identify as Republican voters though they’re still registered as Democrats. It doesn’t matter what party you’re registered with in Louisiana thanks to the jungle primary system.
So while Rispone’s camp probably wasn’t happy with the final early vote tally, it was close enough that he could easily do what Trump in 2016 and Ron DeSantis in 2018 did to Arceneaux in Florida.
The second point of data came Monday, in a Mason Dixon poll for the Gray Television stations around the state. That poll showed Edwards edging Rispone 48-46, with six percent undecided and a margin of error of four percent. That’s a statistical tie.
The key item in the Mason Dixon poll is that it sampled blacks at 31 percent of the vote, which was the number generated in the early voting period. Again, that number would be more or less a modern record; it would exceed, by a half of a percent, the number Barack Obama was able to generate. If the share of the electorate represented by black voters is 29 or 28 percent rather than 31, Edwards is tied or behind Rispone, with six percent or so of the vote undecided and traditionally those voters will break against the incumbent.
The Mason Dixon poll showed Rispone getting 85 percent of Ralph Abraham’s voters, with Edwards getting 10 percent and five percent still undecided. With President Trump coming back to Louisiana and holding a rally in Bossier City tomorrow night, it’s a decent bet Rispone’s performance with that 15 percent of Abraham’s voters still unsold could tick upward.
Rispone’s other problem as reflected in the Mason Dixon poll is that while in North Louisiana and in the south-central and southwestern part of the state he’s got a 20-point advantage over Edwards, in the Baton Rouge and New Orleans areas he’s getting hit by about 15 points. Rispone’s inability to move Jefferson Parish into his corner to date has been a significant failure of his campaign, and he’s going to have to mitigate that by Saturday. That’s why many Republican pols in the New Orleans area have been griping that Trump was needed a lot more in Kenner than in Bossier, and if Rispone doesn’t pull out the race that’s going to be something the critics will point to as a tactical error.
The campaign not taking political advantage of the holy mess that is the opening of the New Orleans airport’s new terminal, and particularly the poor ingress and egress into the facility thanks to Edwards not having built the flyover ramps from I-10 to the airport, is another somewhat inexplicable decision affecting his performance in the New Orleans area.
Still, the Mason Dixon poll shows more or less a dead heat, with Edwards under 50 in the week of the election. Generally speaking that would be a sign the challenger is likely to win.
The rule the Democrats like to use for winning statewide elections is that they need black voters to be 30 percent of the vote, and then they need 32 percent (it used to be 30 percent, and now they’re saying 32 percent) of the white vote to come in for them. The Mason Dixon poll says Rispone is beating Edwards 66-28 with white voters, with six percent undecided. It’s hard to see Edwards getting more than half of the undecideds; he’s the incumbent. And if we’re right about the black vote being cannibalized in the early voting, Edwards would come up short on both sides of his equation, with the black vote share under 30 and his share of the white vote being well under 32.
Then there was the poll released Tuesday by Edgewater Research and My People Vote, which was performed by pollsters Ed Chernevak of the University of New Orleans and Tony Licciardi of the My People Vote app, which had Edwards holding an infinitessimal 49.2-48.9 lead over Rispone. Those same pollsters had surveyed the race on Oct. 20 and found Edwards ahead 50.3-46.6, indicating Rispone with momentum.
There were some items in that poll which didn’t quite add up, though.
For one thing, the Edgewater poll says Edwards is getting 38 percent of the white vote but Rispone is getting 30 percent of the black vote. Those numbers would both be far off what we’d consider normal. Additionally, the poll says Edwards is only beating Rispone 50-48 with women, when every other poll shows a gender gap between the two.
The Edgewater poll shows Rispone winning every congressional district in the state save for Cedric Richmond’s District 2, the majority-minority district, which Edwards is winning by a 79-20 margin. Rispone is getting 59 percent of the vote in District 4 (Mike Johnson) and in District 5 (Abraham), he’s getting 57 percent in District 6 (Garret Graves), 53 percent in District 3 (Clay Higgins) and 52 percent in Steve Scalise’s District 1.
The poll has black voters as 28.5 percent of the electorate and actually has 45 percent of its respondents being Republican. The first number is probably close to correct, while the second one is not.
We’re not sure this poll is all that reliable, though its top-line result looks about right.
The conclusion to this is this race is far too close to call, and from a Republican perspective there really is no excuse not to vote – or even to let your friends and neighbors sit the race out. Every vote counts this time, for real.