CLIENT MEMO: Bannon Endorsement Hurts Republicans
In September, we teamed up with 0ptimus and found that Donald Trump’s endorsement had absolutely no impact on the Alabama Senate runoff. Some interpreted the results as a win for Steve Bannon, who supported the winner of the runoff. While we saw no evidence in our data from Alabama that showed Bannon’s actions changing the trajectory of the race, we wanted to learn more about what voters thought about the Breitbart editor and former White House staffer. We interviewed 2,435 likely midterm voters in Florida, Wisconsin, Ohio, Pennsylvania, and Virginia to learn more.
Key takeaway: Steve Bannon is highly toxic among general election voters and not helpful with Republican voters.
- Nearly four in 10 voters (37.2%) say an endorsement by Bannon would make them less likely to support that candidate. Only one in 10 (10%) voters would be more likely to support a candidate who’s received Bannon’s endorsement. A combined 52.8% said they either do not know Bannon or that his endorsement would make no difference.
- An endorsement by Donald Trump would hurt a candidate with a similar number of voters (43.2%), but the potential upside is much more significant: 1 in 3 voters (33.1%) say they are more likely to support a candidate that Trump supports.
- Bannon’s endorsement does not help candidates with Republican voters. Fewer than one in 7 (13%) of Republican voters said they would be more likely to support a candidate that Bannon endorses, while the same number (13.9%) of Republicans say they would be less likely. In contrast, four in 10 (40.2%) of Republican voters said they would be more likely to support a candidate that Donald Trump endorses, while one in 7 (14.1%) say they are less likely.
Methodology: Between 10/27-10/28, we surveyed 2,435 modeled likely midterm voters in Florida (N = 699), Wisconsin (N = 330), Pennsylvania (N = 384), Virginia (N = 601), and Ohio (N = 421) via IVR, landline only. Likely voters were defined as anyone having voted in the 2010 or 2014 midterm elections, plus the 15% additional most likely to turnout based on in-house turnout score modeling. Margin of error varies by question and segment, but is generally +/- 2.2% for topline results. Sample was weighted by state, age, gender, and party based on 2014 midterm turnout in the latest L2 voter file for each state. Results were then re-balanced based on these cohorts.