After Gov.-elect Matt Bevin’s (R) strong victory in Kentucky last week — a nine-point win that surprised almost all political observers, including us — we’re again confronted with a difficult-to-handicap red state gubernatorial race, this time in Louisiana.
There, Sen. David Vitter (R) appears to be trailing state Rep. John Bel Edwards (D) despite the state’s strong Republican leanings. Vitter has been under fire for months for his 2007 admission that he used a prostitution service in Washington D.C. and appears weak in other ways. He won just 23% of the vote in the initial round of voting to edge out two Republican rivals and advance to a runoff with Edwards. While Republicans will try to link Edwards to President Obama — a surefire strategy in any red state — Democrats appear to be successfully tying Vitter to very unpopular outgoing Gov. Bobby Jindal (R). (Supporters of each candidate will quibble with these comparisons: Unlike Obama, Edwards is culturally conservative while Vitter and Jindal are longtime bitter rivals.)
Still, using Kentucky as a precedent suggests a clear outlook: Both states are very Republican at the presidential level, and polls showing Edwards with a solid lead over Vitter could very well be wrong, just like the polls showing Bevin trailing state Attorney General Jack Conway (D) were. The state’s underlying partisan fundamentals strongly favor Republican Vitter, just as they strongly favored Bevin.
Despite this, we’re downgrading Vitter’s chances of winning, moving the state from Leans Republican to Toss-up in advance of the state’s Saturday, Nov. 21 runoff.
How can that be? Regular readers know that we frequently stress the increasing down-the-ballot partisanship of voters, an observation bolstered recently by Michigan State University political scientist Corwin Smidt’s widely-cited new study about the decline of swing voters. We and others set aside this trend when assessing Kentucky, and we ended up missing the forest to focus on an odd-looking tree. Aren’t those who believe Vitter has a real chance of losing committing the same mistake?
Quite possibly. But the circumstances in the two races are substantially different.
It’s easy to generically lump Kentucky and Louisiana together. They both have the same number of electoral votes, eight, and they voted similarly in the 2012 election: Mitt Romney got 60.5% in the Bluegrass State and 57.8% in the Bayou State. Both states have two Republican senators, and the GOP also controls five of each state’s six House seats: The Democrats are limited to a single seat apiece in each state’s biggest city, Louisville and New Orleans.
But, demographically, the states are dissimilar. While we do not have exit poll information for either state’s most recent election, both had exit polls in their 2014 Senate races — that includes the initial jungle primary in Louisiana last year, not the eventual runoff won by now-Sen. Bill Cassidy (R). Louisiana’s electorate was far more diverse, 30% black, than Kentucky’s, which was 8% black. That does give Edwards a bigger base to work with, although the average white Kentuckian is more open to voting Democratic than the average white Louisianan. The level of black turnout is just as important to Edwards as is his ability to outperform former Sen. Mary Landrieu’s (D) abysmal performance with whites in 2014 (in the initial vote, she got just 18% of whites and almost certainly didn’t do much better in the runoff).
Perhaps helpful to Edwards in his pursuit of GOP votes is the endorsement of Lt. Gov. Jay Dardenne (R), who finished just behind Vitter in the recent gubernatorial jungle primary. It is always difficult to judge whether a cross-party endorsement like this one really makes much of a difference. No doubt some Dardenne backers will vote for Edwards and others will refuse to show up in the runoff to vote for Vitter. Yet Dardenne supporters are overwhelmingly Republican-leaning, and in this partisan era, it’s not easy to bend most voters’ party identification.
Then there are the polls.
For all of the problems with Kentucky’s surveys, they did generally get one part of the race right: Conway’s share of the vote. The final HuffPost Pollster average had Conway leading 43% to 41%, with 7% for independent Drew Curtis. Conway ended up winning 44% of the vote. Curtis only got about 3.5% — so polls overstated Curtis’ support, which is common for third party candidates — while Bevin’s total went up 12 points, effectively taking in almost all of the undecideds and some of the Curtis vote. In a Republican state with an environment that naturally favored a Republican — President Obama loomed over the race, and Kentucky strongly dislikes him — this makes a lot of sense in hindsight.
These polls remind us of two recent off-year races in our home state of Virginia. In the 2013 gubernatorial contest, the final HuffPost average showed now-Gov. Terry McAuliffe (D) leading then-Attorney General Ken Cuccinelli (R) 45%-38%, with a third party candidate taking 9%. The final result was a McAuliffe win, but only by a 48%-45% margin, with the third-party contender winning 7%. A year later, Sen. Mark Warner (D) was leading former Republican National Committee Chairman Ed Gillespie (R) 50%-39% in the average. Warner won, but only 49%-48%. In both races, the undecideds in the polls overwhelmingly flocked to the Republican in what were both Republican-leaning years overall. Virginia has had a long history of supporting a gubernatorial candidate from the party that did not control the White House, and of course 2014 was a big GOP year nationally. Virginia is a much more Democratic state than Kentucky and Louisiana nowadays, so McAuliffe and Warner won, but the polling trends are still instructive.
Perhaps the lesson is that when we look at the Louisiana polls, we should assume that Vitter, as the Republican running in a Republican state in a pro-Republican environment, will get the lion’s share of the undecideds in the polls, while Edwards will probably be stuck at about whatever his final polling average is.
What’s fascinating is that unlike the other Democrats mentioned here — Conway last week, Warner in 2014, and McAuliffe in 2013 — Edwards is above 50% in an average of the four polls conducted since the runoff: He leads 52%-38% over Vitter. The most recent poll, from a Republican firm, is the closest, showing Edwards up 49%-41%. We’re very interested in what happens to Edwards’ percentage in the remaining polls — if it starts to drop, that will be a clear indication that Vitter is coming back. We doubt Vitter will lead any neutral poll, but the trend is more important than the topline if you assume, as we are doing at the moment, that a large majority of the undecideds will vote for Vitter. So, to us, the current polls don’t show a 52%-38% edge for Edwards — they show something more like a 53%-47% or 52%-48% lead. Edwards is up but not overwhelmingly so, and Republicans have another week and a half to hit him with highly negative ads tying him to Obama in an effort to reclaim a few percentage points of voters who typically vote Republican.
We plan to weigh in again on Thursday, Nov. 19, two days before the runoff, with our final thoughts on the race. At the moment, though, it is premature to declare a favorite. Again, it’s the fundamentals vs. the campaign. In Kentucky, the fundamentals won. But that does not necessarily mean they will again in Louisiana.