Alabama Senate Republican Primary Runoff Election
August 24-26, 2017

Electoral Environment


Candidate Characteristics: Voters were asked to identify the candidate characteristic that is “most important in determining your vote.” A plurality of likely voters say “strong support for President Trump” is the most important candidate characteristic (38%). “Conservative values” is the second most-cited quality (27%), followed by “an ability to get things done for Alabama” (22%). Having a candidate who demonstrates a “willingness to fight the Republican establishment in Washington” (11%) is less of a priority for voters. 

Q. Thinking about the Republican primary runoff for United States Senate, which of the following candidate characteristics is most important in determining your vote: Strong support for President Trump, Conservative values, A willingness to fight the Republican establishment in Washington, OR An ability to get things done for Alabama?


Trump Candidate: Following his endorsement from the President, a majority of likely voters choose Luther Strange as the candidate who is most supportive of President Trump, rather than Roy Moore (40% Moore-50% Strange).
-A full 60% of voters who identified support for Trump as the most important candidate characteristic say Strange is the most supportive of Trump. 
-Very Conservative voters are more likely to select Moore as more supportive of Trump (54%-37% Strange).

 
Q. Which of the following candidates for U.S. Senate do you consider to be the most supportive of President Trump: Roy Moore OR Luther Strange?


Trump Image: President Donald Trump is viewed favorably by an overwhelming 87% of likely Republican primary runoff voters. More than two-thirds of these voters hold a very favorable view of the President (69%). His image weakens with voters who want a candidate who will get things done for Alabama (62% favorable, 39% very favorable). Trump’s intensity of favorability is far higher among Very Conservatives (85% very favorable) than among Somewhat Conservatives (56%).  

Senate Candidate Images


Luther Strange (60% favorable/24% unfavorable) and Roy Moore (59%/26%) have nearly identical favorability and name identification among likely voters. 
-Roy Moore has a stronger image among men (63%/28%) than he does among women (55%/25%) while Luther Strange has a stronger image with women than with men (64%/18%; men: 54%/30%). 
-Moore’s image improves among Very Conservatives (72%/14%) and worsens somewhat among Somewhat Conservatives (54%/33%). Strange maintains his image among Very Conservatives (59%/24%) and improves with Somewhat Conservatives (64%/18%). 
-Moore is viewed net-unfavorably by voters who want a candidate who can get things done for Alabama (39%/45%). 

Senate Ballot Test


Roy Moore holds a 2% advantage over Luther Strange in the Republican primary runoff election for U.S. Senate (47% Moore-45% Strange). Moore and Strange voters exhibit identical levels of intensity of support for their candidate—84% of Moore voters say they are definitely voting for him, as do 84% of Strange voters. Strange has an advantage among the 8% of voters who are undecided—30% of them say they lean towards Strange (6% Lean Moore, 63% truly undecided). 

Q. If the Republican primary runoff election for Senate were held today, which of the following candidates would you vote for: Roy Moore OR Luther Strange?


(Asked only of Moore voters)
Q. And would you say you are definitely or probably going to vote for Roy Moore?


(Asked only of Strange voters)
Q. And would you say you are definitely or probably going to vote for Luther Strange?



(Asked only of Undecided Voters)
Q. And would you say you lean towards Roy Moore or Luther Strange?


The following chart shows the ballot by various key demographics. As the chart demonstrates, this is a close race.

Additionally, the race is tight in all of the key media markets:
-Birmingham: 46% Moore-46% Strange
-Huntsville-Decatur: 47-43%
-Mobile-Pensacola: 48-44%
-Montgomery-Selma: 44-47%

Campaign Communications

Likely voters are hearing slightly more about Luther Strange (61% yes) than they are about Roy Moore (56%). The flow of information about both candidates is a net positive, with Moore holding somewhat of an advantage (Moore: 49% more favorable/26% unfavorable, Strange: 42%/37%). 

Q. Have you seen, read or heard anything recently about Roy Moore?


(Asked only of ‘Yes’ respondents from previous question)
Q. Has that information given you a more favorable or a more unfavorable opinion of Roy Moore?


Q. Have you seen, read or heard anything recently about Luther Strange?


(Asked only of ‘Yes’ respondents from previous question)
Q. Has that information given you a more favorable or a more unfavorable opinion of Luther Strange?


The following chart shows the favorable-unfavorable information flow number for each candidate among several demographics. Strange has an advantage with Somewhat Conservatives, support for Trump voters, and get things done voters. Moore has the information flow advantage with Very Conservatives, Conservative values voters, women, men, those who have seen, read or heard about both candidates, and those who know the date of the election.

Undecided Voter Profile

The following table compares various results with the overall sample and among undecided voters.

 
- Undecided voters have a positive view of President Trump, but with somewhat lower intensity.
-They have comparable images of the two candidates, but Strange has higher name identification.
-These voters are less likely to know the date of the election.
-They are hearing less about Moore than they are about Strange, and Strange has an information flow advantage.
-Their top candidate characteristic is "an ability to get things done for Alabama."
-They are significantly more likely to be women.

Full Results & Methodology

METHODOLOGY:
The sample size for the survey is 600 likely Republican primary runoff election voters in Alabama and the margin of error is +/-4.0%. Reponses were gathered via landline interviews conducted with Interactive Voice Response (IVR) technology. The survey was conducted August 24 - 26, 2017 by Harper Polling. The total percentages for responses may not equal 100% due to rounding.

Harper Polling