A new Emerson Poll finds Alabama incumbent Senator Luther Strange trailing former Judge Roy Moore by 14 percentage points with about two weeks until the Special Election on September 26. Moore polled at 40% while Strange was at 26%. Thirty-four percent (34%) of voters were undecided, and the survey had a margin of error of 5.2 percentage points.
Voters in Alabama gave President Trump a 52% job approval and 36% disapproval, but that popularity has not transferred to his endorsed candidate Strange, as voters who approve of the President are breaking for Moore 51% to 32%.
Supporters of Republican Congressman Mo Brooks who grabbed 20% in the GOP primary in August appear to be splitting their vote between Moore and Strange, with about 1/3rd breaking for each candidates and the last third undecided.
Regardless of who wins the divisive GOP nomination it appears Democrat Doug Jones will start in a competitive position for the General election as he trails Strange 40% to 43%, and Moore 40% to 44%, well within the polls 4.8% margin of error.
The GOP primary may have left supporters of both candidates unwilling at this time to support their party rival. Of those voting for Moore in the primary, 25% said they will vote for Jones and 49% said they would vote for Strange in the General. Similarly, Strange supporters found 31% voting for Jones and 34% voting for Moore. The GOP will need to find a way to unite during the 11 weeks until the General Election, or face the prospect of Jones pulling off an upset. If Jones were to win, Alabama could send their first Democrat to the US Senate in over 20 years.
The Emerson College Polling Society, under the supervision of Professor Spencer Kimball presents these findings based on a poll of Likely Voters in Alabama, for the Republican Special Election September 26, 2017 and of all likely voters for the December 12 General Election. All respondents interviewed in this study were part of a fully representative sample using an area probabilistic sample of registered voters from Aristotle International. The overall sample size was N= 416 with a margin of error of +/- 4.8 percentage points in 19 of 20 cases. In the Republican Primary the sample was n= 355 with a MOE of +/- 5.2 percentage points. It is important to remember that subsets based on gender, age and party breakdowns carry with them higher margins of error, as the sample size is reduced. Data was collected using an Interactive Voice Response (IVR) system of landlines only and was conducted September 8-9, 2017.
This survey uses statistical weighting procedures to account for deviations in the survey sample from known population characteristics, which helps correct for differential survey participation and random variation in samples. The overall adult sample is weighted to recent Census and voting behavior data using a sample balancing procedure to match the demographic makeup with age and gender. Margins of sampling error for this survey are not adjusted for design effect.