How Doug Jones Beat Roy Moore
Zac McCrary (@zacmccrary) is a Democratic pollster at ALG Research.
Before last Tuesday, Democrats had lost every US Senate race in Alabama since 1992, most of them in blowout fashion. In fact, there had really only been one competitive Senate race in the interim (1996). To be sure, virtually everything had to break in Jones’ direction to pull off such a monumental upset. The list of developments that had to align for Jones is substantial: Governor Robert Bentley had to resign due to a sex scandal, Governor Ivey had to move the election from November 2018 to December 2017, Republicans had to nominate Roy Moore, Doug Jones had to run a strong campaign, numerous victims of Roy Moore’s predatory behavior had to courageously come forward, the national GOP infrastructure had to abandon Moore for most of the race, the Jones campaign had to have a dramatic resource advantage over Moore, the political winds had to be at Democrats’ back, Richard Shelby had to repeatedly emphasize his choice to write-in a Republican because he couldn’t support Roy Moore…remove just one of these factors and this race could have certainly broken the other way.
This map (created by Miles Coleman) tells much of the story. This compares Jones narrow 2017 win (a 1.5% margin of victory) with the narrow Democratic loss against Roy Moore in the state’s 2012 Chief Justice race (Moore won by 1.7%). Jones improved upon the Democrat’s 2012 numbers most dramatically in the metro counties of Jefferson County and Madison County, plus suburban-Birmingham Shelby County (historically very GOP turf). The two college counties of Tuscaloosa and Lee also performed dramatically better for Jones than the 2012 Democratic candidate. Additionally, Jones did well in the Alabama counties that comprise the Columbus, GA spill market (Barbour, Chambers, Lee, Russell) — presumably where Jones significant resource advantage was leveraged to an even greater extent than in the other relevant media markets. There was also slight Jones over-performance in a couple of the counties that would have been most impacted by the very small Dothan media market. Elsewhere, the state’s rural Black Belt, the Mobile metro area, the blue-collar “Shoals” area in Northwest Alabama also showed more muted, yet still critical, movement toward Jones.
Jones victory came together because of three crucial successes among the Alabama electorate — all of which were imperative to achieve the political upset of a generation.
Strongly solidifying the Democratic base
According to the exit polling, Doug Jones won 96% of African Americans and 98% of self-identified Democrats. With African Americans comprising in the high 20s as a share of the state electorate (on a good day), this base solidification is almost certainly a necessary component of a winning coalition. However, this is by no means a certainty in a competitive race.In what I largely believe is a political urban legend, there was a theory floating around that, due to his high-profile religious stands, Roy Moore could earn votes among African Americans that a more traditional Republican would not. I do not believe this was borne out in Moore’s previous races — and there is no sign at all that Moore made any inroads among African American voters on Election Day. In a race like this, there is indeed a major difference between earning 96% of the African American vote and 90%. In this case, that difference equates to approximately 1.7 points of statewide vote— greater than Jones’ ultimate victory margin. To put it another way, taking the exits at face value, if Jones African American support was 90% instead of 96%, Roy Moore would now be Senator-elect.
Higher turnout in Jones’ geographic strongholds than in Moore’s base
While Moore did well in his traditional base — even improving on his 2012 vote share in many largely white, blue-collar counties — the total turnout in many Moore base counties lagged that of Jones’ strongholds. Our polling indicated that Jones’ voters were more energized than Moore voters and, additionally, were more likely to feel this race was “more important” than most races.
For example, there were more votes cast in strongly Democratic, majority African-American Dallas County (home of Selma, AL) in the 2017 election (approx. 14,000) than in the 2014-midterm (roughly 12,000). However, in one of Moore’s strongest counties, Blount County, there were roughly 1,000 fewer votes in December 2017 than in the 2014 midterm. Data-crunching by Miles Coleman at the congressional district level, shows that the total turnout in Jones’ best district (the 7th) was 211K. In Moore’s best district (the 4th), the total voter turnout was the lowest in the state — just 165K voters. Montgomery Advertiser reporter Brian Lyman sleuthed out that the voter turnout percentage (40% statewide) was 42.25% in counties that Jones won, compared to 36.85% in counties Moore won.
The exit polls show African Americans comprising 29% of the electorate — which, if true, is a slightly higher share than the previous high-water mark of the 2008 elections (28%). If accurate, such an African American turnout is not only extraordinary, but historic. With that said, exit polling is not meant to be pinpoint accurate in such a fashion — and it will likely be several weeks until the individual-level voter data is available to determine if the African American share was indeed historic (29%+) or on the very high end of more “realistic” projections (27–28%). Regardless of the precise final number, it seems clear that the Jones’ campaign small-donor fundraising program allowed them to have the resources to build a very robust field / GOTV program. Simply put, the Jones campaign had the resources to pour gasoline on the fire that was already burning within the Democratic electorate.
Jones won Independents and peeled off enough white suburban voters to build a winning multi-racial coalition
The exits show Jones winning Independent voters 51% to 43%. In Alabama elections — as in many other red states — Independents have been behaving much more like Republicans than Democrats over most of the past decade. Given the state’s Republican DNA, it is of course a must for Alabama Democrats to win Independents by a healthy margin to be in a position to win statewide. While at this point, it’s difficult to disentangle how much of Jones’ success in the big four metro counties (Jefferson, Madison, Mobile, Montgomery) and some of the other counties where he performed very well (Lee, Tuscaloosa) is due to strong African American turnout versus white suburbanites moving toward the Democrat, it’s clear Jones did better among a type of white suburban voter than any Alabama Democrat has in a generation.
Based on the exit polling and our own data before the election, it’s clear Jones’ path to hitting his white “win number” ran through women, younger voters, those with college degrees, and those in urban/suburban areas. This is a voter profile that has moved Virginia from reliably Republican to leaning Democratic over the past decade — and has shifted North Carolina into a prototypical swing state in that time as well. Georgia is likely the Southern state next to join that list in the near future, but now there is at least a blueprint for Alabama Democrats of what a winning coalition looks like.
Base solidification, a Democratic turnout surge, and sufficient success with independent, white suburban voters were ALL absolutely necessary for the Jones’ campaign to thread the needle to win a Senate race in a state that Donald Trump carried with 62%. While some pundits argue that Democrats must focus disproportionately on base motivation or, conversely, should largely target voters who’ve moved toward Republicans over the past couple of cycles — the lesson that Senator-elect Doug Jones (D-AL) provides is clear. Democrats should (and likely must) employ an “all of the above” strategy to win tough states and districts in 2018 and beyond.
At the risk of self-promotion, my Election Day primer provides a pre-election deeper dive on many of these dynamics. With the benefit of hindsight, Jones’ margins in the state’s big four counties exceeded my expectations, turnout in the state’s heavily-Democratic Black Belt was indeed strong, Jones largely met his necessary vote goals in historically GOP, suburban areas, and Jones also ran well ahead of expectations in college towns of Tuscaloosa and (especially) Lee County. For those looking for a deeper drive, this pre-election piece can offer more substance.
Zac McCrary (@zacmccrary) is a Democratic pollster at ALG Research.