Slight differences in turnout could change the outcome of tomorrow's special U.S. Senate election in Alabama. According to the Monmouth University Poll , a standard midterm turnout model gives Republican Roy Moore a slight advantage. A higher, although less likely, near-presidential election turnout would give Democrat Doug Jones a slim lead. An adjusted midterm estimate based on patterns seen in last month's Virginia gubernatorial race - i.e. relatively higher turnout in Democratic strongholds - puts Tuesday's election up for grabs.
Using the 2017-based model, Moore and Jones each garner 46% of the likely electorate's stated vote intent, with 2% opting for a write-in candidate and 6% still undecided. A historical midterm model, akin to Alabama's 2014 turnout, gives Moore a 48% to 44% edge. A model with higher overall turnout, where voter demographics look more like the 2016 election, gives Jones a slight 48% to 45% advantage. It's worth noting that all of these leads are within the poll's margin of error for each model.
The key difference between the 2017-based model and the standard midterm model is an upward adjustment in the statewide vote share coming from Jefferson County - home to the state's largest city, Birmingham - and twelve Democratic-leaning counties that form a belt across the lower portion of the state from Sumter to Russell, including Montgomery. These counties typically make up 24-25% of Alabama's electorate in any given election. The 2017-based adjustment increases that share to 27%. The even higher turnout model that puts Jones in the lead is significantly younger (33% under age 50) than the 2017-based model (24% under age 50).
"In a typical year, we would probably default to the historical model, which shows Moore ahead. It could still end up that way, but both 2016 and 2017 suggest that typical models may not apply. If we see a surge in Democratic turnout, especially in the Birmingham region, Jones has a chance. On the other hand, if turnout is significantly lower than a standard midterm election, Moore's prospects increase," said Patrick Murray, director of the independent Monmouth University Polling Institute.
Another wrinkle in assessing voter intent in this election is the undecided vote or soft support that has the potential to change, particularly given the potential reluctance of voters to say they are supporting Moore given reports of his past behavior. The Monmouth University Poll used responses to a series of questions - on party preference, candidate favorability ratings, and the impact of President Trump - to allocate undecided and "write-in" voters to either of the two major party candidates. This potential allocation works to the Republican's benefit.
Using the 2017-based turnout model, Moore could take a slim 3 point lead if undecided and write-in voters decide to break in the direction of their overall partisan tendencies. In the high turnout model, Moore would still trail but by an insignificant single point. On the other hand, he could expand his advantage to 7 points in the historical midterm model. Furthermore, some voters with a strong partisan identity say they currently intend to vote for the nominee from the other party. Reallocating these voters to the candidate that best matches their strong partisan inclinations could provide a further net gain for Moore.
"Basically, the various turnout and vote intent models suggest that a Moore victory is the more likely outcome, but there is still an opening for Jones. He needs to get relatively higher turnout in Democratic areas and keep GOP-leaning voters who are uncomfortable with Moore from ultimately choosing him once they get into the privacy of the voting booth," said Murray.
Polling this race is further complicated by the fact that Alabama is one of the remaining states that allows voters to choose a straight party ticket rather than having to select a candidate for each office. Based on information from the Alabama Secretary of State, approximately half of the electorate chose a straight party ticket in 2014 and 2016.
It is difficult to replicate this dual ballot option in a telephone poll, but Monmouth did ask about voters' likelihood to choose a party ticket in this election. Overall, 42% of likely voters said they are at least somewhat likely to choose the straight party ticket, which is somewhat lower than the historical average, although there is only one office on this year's ballot. Slightly more voters indicate they will choose the Republican (22%) versus the Democratic (18%) party option, which is in line with historical patterns.
In many cases, Alabama voters choose the party ticket and also select a candidate for the individual office. According to the Secretary of State, an individual candidate vote overrides a party ticket vote. Among voters who are still undecided or are only leaning toward a candidate, 19% say they may vote for the Republican ticket compared with 9% who may vote for the Democratic ticket. Moore's overall vote could be boosted by one or two percentage points if these voters end up choosing the party ticket option and leave the Senate choice blank.
"It's worth keeping in mind that one difficulty in polling Alabama's electorate is that very few, if any, pollsters have a track record there. Monmouth's only prior foray into the state was during the 2016 presidential primary season. This lack of familiarity is further compounded by the unpredictability of special elections, which is why we chose to describe the contours of this race rather than release a single estimate. Ultimately, that is what polling should be about anyway," said Murray. He added, "There's a lot we can learn from this election in terms of whether changes in voting patterns we have seen since November 2016 signal a realignment or are just a momentary blip. We'll be examining these results in more detail after the voter rolls are updated in the spring."