Three takeaways from Louisiana's U.S. Senate runoff
1. The New Normal: Results in the U.S. Senate result look a lot like what we now see in elections for state office in Louisiana.
John Kennedy’s and Foster Campbell’s vote shares in the runoff for U.S. Senate are very close to the average received by Republicans and Democrats respectively across all Louisiana statewide races over the past decade. Democrats tend to get 40 percent of the vote in elections that involve at least one Democrat and at least one Republican. Republicans average 58 percent.
That's pretty much where this one ended: 39 percent for the Democrat and 61 percent for the Republican. Even the map of how the state's 64 parishes broke in the runoff looks remarkably like the maps for the 2012 and 2016 presidential contests.
What is even more interesting is how close the results are to the state’s average for both federal office and state office. Indeed, the outcome of the 2016 U.S. Senate runoff are even closer to the average for state offices than for federal offices. After Democrat John Bel Edwards won the Louisiana governor’s office, there has been quite a bit of talk about how state politics are so very different than national politics. I have argued against this view. Here is further evidence that our state politics are not as distinct from national politics as we may have thought.
2. The Democratic Challenge: Campbell’s problem was with white voters, not black turnout.
As expected, turnout in the senate race dropped dramatically from the primary to the runoff. In fact, the size of the drop was historically unprecedented. This is because – for the first time under the state’s ‘jungle primary’ system – the primary was held in conjunction with a presidential election and the runoff was not. From adoption of this election system in the 1970s through the 1996 U.S. Senate race, the primary was held earlier and the runoff fell on the day of the presidential election. In 2004 there was no runoff in the U.S. Senate race. In 2008, the state did not use the ‘jungle primary’ system for federal elections.
Without a presidential election on the ballot, turnout plummeted in the runoff this year. Overall, the number of ballots in the runoff was just 46 percent of the number in November.
However, the drop was less in the parishes carried by Campbell. In those nine parishes, runoff turnout was 54 percent of primary turnout. These parishes also tend to have a higher share of black voters in the primary. In contrast, in Kennedy’s parishes the number of runoff votes was only 43 percent of the number of primary votes.
The takeaway: Turnout fell less in parishes that have larger shares of black voters, who tend to vote for Democrats. So, it was not a lack of black turnout that doomed Campbell. More likely, he was done in by the increasing difficulty Democratic candidates have with white voters in Louisiana.
3. Strategy in the Jungle? There probably wasn't much strategic voting in the primary.
One of the things that struck me in the primary results was that the Democratic candidates for U.S. Senate under-performed relative to Hillary Clinton. In a lot of parishes it appears that voters cast ballots for Clinton in the presidential election and a Republican in the U.S. Senate race.
One of the explanations I heard for this was that a significant share of Democrats were voting strategically – that is, fearing a Democrat could not win a runoff with a Republican, they voted for Charles Boustany, whom they saw as the most palatable Republican. The logic is similar to what some prominent Democratic supporters suggested early in the 2015 gubernatorial campaign when their was some skepticism about whether a Democrat could defeat David Vitter.
The primary results suggested that there may have been some strategic voting: The Republican candidates’ vote share in the U.S. Senate primary was especially likely to over-performed Donald Trump’s vote share in Boustany’s strongest parishes. The chart below plots the combined vote share for all Republican candidates on the primary ballot for U.S. Senate against Trump’s share of the vote in each parish. Parishes are color-coded to indicate which Republican did best among all Republican candidates.
Notice that all the parishes where Boustany led the Republican field fall on or above the line. This was not so for Fleming and Kennedy. In short, Republican senate candidates did especially well relative to Trump in places where Boustany did best. But was this the result of strategically minded Democrats backing their ‘next best’ option?
If there was indeed a bunch of strategically-minded Democrats driving up Boustany’s vote in the primary, then we should see Kennedy’s share of the runoff vote fall closer to Trump’s vote in those parishes where Boustany did especially well. But, we don’t…at least not convincingly. The results are kind of a mixed bag. In a few of these parishes the Kennedy runoff share is close to the Trump share, but in many Kennedy still outperformed Trump.
The test is not conclusive, but it leads me to suspect that the gap between the Trump vote and the Republican primary vote came from Republicans that could not quite stomach voting for Trump rather than from Democrats strategically voting for Boustany.