Incumbency still matters in the age of partisan voting
One of the most intriguing story lines in the Louisiana governor’s race – where an incumbent Democrat is seeking reelection in a red state – is the contest between two ideas about what matters for election outcomes. One idea is that, whether it is a presidential contest or a gubernatorial election, the partisan leanings of a state matter a lot. Therefore, the candidate(s) who represent this partisan lean have the advantage. The other idea is that incumbents have the advantage because of the benefits of holding office (or because they are already proven winners).
The story line of incumbency versus party has already captured media attention beyond the Pelican State. Last week, Amber Phillips of the Washington Post rated Louisiana the second most intriguing gubernatorial contest of 2019 and 2020 because of this tension tension between incumbency and party (e.g., “Can a Democratic governor win reelection in a state that voted for Trump by 20 points?”).
How do partisanship and incumbency shape John Bel Edwards’s prospects for reelection?
The short answer is that being an incumbent helps despite being a Democrat in a red state.
Here are three points to know:
1. Governor elections look a lot more like presidential elections than they used to.
Partisanship matters a lot in gubernatorial elections, and their outcomes look a lot more like the outcomes of presidential voting in the same state than they used to.
The vast majority of victors in American governor elections today are from the same party as the presidential candidate who carried their state in the most recent presidential contest. It has not always been this way. The graph below shows how often these same-party-as-the-presidential-nominee gubernatorial candidates won over the past 40 years. I pool governor elections into four-year groups to account for the fact that in some years there are only two elections and in other years there are more than 30, but in each case the gubernatorial contest is compared to the most recent preceding presidential contest for that state (i.e., gubernatorial elections in 1980 are compared to presidential voting in 1976).
The consistency between gubernatorial and presidential outcomes rises steadily from below 40% in the early 1980s to approximately 75% today.
[For a more in-depth dive into this phenomenon, read The Increasingly United States: How and Why American Political Behavior Nationalized by University of Pennsylvania political scientist Dan Hopkins. Hopkins describes how our state and local elections are increasingly nationalized – that is, we appear to use the same partisan basis for our decisions about whom to vote for whether it’s a presidential contest or a mayoral race. In the third chapter of his book, he shows how Democrats’ vote share in a state’s presidential election has become a more accurate predictor of Democratic vote share in the state’s governor’s race over the past forty years.]
There is ample prima facie evidence of this alignment between national and state elections in Louisiana. The average Republican share of the statewide vote in presidential and U.S. Senate elections in Louisiana over the past 12 years is 57 percent (versus 41 percent for Democrats). What is the average Republican share of the vote in statewide elections for state government office over the same period? Nearly identical - 59 percent (versus 40 percent for Democrats).
There are exceptions, of course (e.g., Democratic governors in Louisiana and Montana; Republican governors in Massachusetts and Maryland). However, to treat these elections as if they are unrelated, as if state politics is somehow immune from the partisanship of our national elections, is simply foolish. Indeed, even our state legislative elections look a lot like presidential voting.
2. Incumbents are not immune to this trend.
Are incumbents an exception to this trend? There are a couple of reasons to think they might be. Historically, incumbents tend to have an advantage over their challengers. This advantage may arise from holding office – more media coverage, more opportunities to appeal to voters, more deference from voters for current office-holders. Or, it could simply indicate that incumbents are competitive candidates as evidenced by the fact that they have already won election. Either way, perhaps incumbency provides a source of support beyond the force of partisanship.
The evidence indicates that the electoral fortunes of incumbent governors are increasingly tied to how their party fared in the state’s presidential voting. Looking at how the share of the vote incumbent governor’s receive compares to the share of the vote their party’s nominee received in the most recent presidential contest in the state, the link stands out.
In the 1980s and 1990s, there was no relationship (or even a negative relationship) between how an incumbent governor did relative to his or her party had performed in the last presidential contest in the state. In the last 20 years, however, incumbents do better where their party’s presidential nominee did well and do worse where their presidential nominee did poorly.
3. Incumbency probably helps Edwards more than being a Democrat hurts him.
So, the partisan patterns of presidential voting are more predictive of gubernatorial voting than they used to be, and this trend is tying incumbents’ fortunes to how well their party does in the presidential contest. Is this all bad news for a Democrat like Edwards running in a red state? No.
The reason is because the party penalty is still not all that big for incumbent governors. It’s real, but it is relatively small…for now.
The average vote share in recent elections for incumbent governors in Edwards’s shoes is 51 percent of the vote. That is only five points less than the average vote share of incumbents who share the same party as the presidential nominee who carried the state (56 percent). It is eight points higher than the average vote share of non-incumbents from the party who lost the presidential contest in the state ( 43 percent).
Electoral fortunes of mismatched partisan incumbents are on the decline in recent years, but they have not lost their staying power.
Perhaps more important, the incumbents who are from the party that lost the state in the presidential election still average more than 50 of the vote. In other words, incumbents still do better than predicted based on their party alone.
Even if we confine ourselves to the south, incumbent governors from the party that lost the state in the most recent presidential election nevertheless won reelection the last three times they have tried – a Republican in Florida in 2014, a Democrat in Arkansas in 2010, and a Democrat in North Carolina in 2004. You have to go back 17 years to 2002 to find the last time incumbent Democrats sought and lost reelection in southern states that supported the Republican presidential nominee (in Alabama, South Carolina, and Georgia).
This is all relatively good news for Edwards.
Ultimately, the role of incumbency will be shaped by whether voters think things have gone well (or, at least, well enough) during the incumbent’s time in office. The candidates in Louisiana are fighting to shape voters’ answers to that question. Given the boost that incumbency can give to a Democrat in a red state, the election could depend a lot on which interpretation wins out in the minds of voters.