Roanoke is about to have something it’s never had before: A contested mayor’s race that coincides with a presidential election.
Mayor Sherman Lea hasn’t announced his intentions yet but is widely expected to seek a second term — as a Democrat. Former Mayor David Bowers, previously elected as a Democrat, is running as an independent.
Roanoke’s decision to move municipal elections from May to November — and in a presidential year — directly benefits Democrats. Roanoke generally voted Democratic in May elections (no Republican has won since 2000 but some independents have), but it’s voted even more overwhelmingly Democratic in November elections. In the 2018 mid-terms, Roanoke voted 64% for Democrat Tim Kaine for U.S. Senate and 63% for Jennifer Lewis for the House of Representatives.
We’ve looked before at how a Lea vs. Bowers race would play out and concluded that, based on previous election results, Lea would have a distinct advantage in a two-way race in May. A November election would help Lea even more — he’d presumably benefit from straight-ticket voting from voters going to the polls to vote for whoever the Democratic candidate for president is. Today let’s look at a different question: Is there room here for a third candidate here — specifically, a Republican? Could Lea and Bowers split the Democratic vote enough for a Republican to squeak through? Let’s do some math!
First, let’s recap what we said last fall when the prospect of a Bowers vs. Lea race first surfaced. Bowers’ vote totals have generally declined over the course of his political career, perhaps a consequence of him having an older political base that has the inconvenient tendency to die off. In Bowers’ first race for mayor in 1992, he polled 10,261 votes. In his last two runs for mayor, in 2008 and 2012, his vote totals fell to 5,968 and 4,827 — but he still won. Every election generates its own dynamic — different issues, different candidates — but we can say this with certainty: Since 2000, Bowers has only topped 5,000 votes once.
Lea ran unopposed in 2012 — he faced only a write-in candidate — but still polled 5,315 votes. In his three successful (and contested) runs for city council before that, he polled between 6,364 votes and 6,889 votes.
It sure looks like Lea has a bigger base than Bowers does. Plus, many of Bowers’ votes have historically come from the city’s African American precincts. Lea is the city’s second black mayor. It seems unlikely that Bowers would be able to count on those votes in a contest with Lea, further putting him at a disadvantage.
On the other hand, Lea’s council elections were multi-candidate fields where voters got to cast three votes. He’s never had a head-to-head race. We ask this question gingerly but as a matter of political reality: In a two-way race against Bowers, how would Lea fare with the city’s white voters? Let’s look to the past: In Lea’s most contested council elections, he’s always run strong with both black and white voters. The 2012 council race involved only four candidates for three seats, so it’s not a good measure. The better measures are 2008 (six candidates for three seats) and 2004 (seven candidates for three seats). It’s notable that in 2008 Lea finished first in some of the city’s most affluent white precincts — Fishburn Park, Lee-Hi, Raleigh Court No. 1 — and second in both South Roanoke precincts. Those precincts have never been part of Bowers’ base. He lost many of those by wide margins when he won re-election in 2012. That’s another reason to believe that in a head-to-head race, Lea would have an advantage over Bowers. Lea has proven he can win in both the city’s northwest precincts and the southwest precincts, which is historically a winning coalition.
What would happen, though, if there were a Republican in the race? Would a Republican take those southwest votes away from Lea — either enabling a Bowers win or perhaps winning outright? We don’t have much data to go on because there have been so few Republicans to run in city council elections over the past two decades. The last Republicans to win were in 2000, when Ralph Smith took the mayorship in a four-way race and Bill Carder won a council seat. Smith took 5,368 votes; Carder an astonishing 8,468. Four years later, though, Carder’s total fell to 4,296 and he lost his seat. In 2014, when there were two Republicans running for council, the strongest managed only 2,302 votes. A more optimistic measure for Republicans might be Mark Lucas, who took 4,479 votes in his 2012 race against Bowers for mayor — good enough for 48% of the vote, but not good enough to actually win. Lucas won nine of the 11 precincts that then existed in southwest Roanoke, losing only Wasena and Raleigh Court No. 1 by two votes apiece. He also won several precincts in southeast Roanoke. Lucas lost because Bowers ran up landslide numbers in the city’s black precincts and then padded those numbers with smaller margins along Williamson Road. If a Republican could win Raleigh Court/South Roanoke/southeast Roanoke, while Lea held onto northwest and Bowers took Williamson Road, would that be enough to win? If you took Bowers numbers from 2012 — 4,827 — and split them in two, yes, then a Republican could win. That’s an academic exercise, though, not a real-life campaign where voters may decide to cast a ballot for their second-favorite candidate to block their least-favorite candidate.
Here are more relevant numbers to consider. Let’s assume for math’s sake that 2020 looks like 2016. That year Hillary Clinton took 22,286 votes in Roanoke to Donald Trump’s 14,789 and 2,391 for other candidates. If every one of those people voted in the mayoral race, then a Republican can count on 14,789 votes. How will Bowers and Lea split those 22,286 Democratic votes? Let’s assume Lea gets the bulk of those as the Democratic nominee. Could Bowers peel off enough votes — 7,497 in the scenario above — to tip the election to the Republican? And what kind of Republican might that be? Lucas, who did so well in 2012, cut a far more centrist figure than Trump does. Would a centrist Republican get those Trump votes in the mayor’s race? Would there be some anti-establishment voters who might vote for both Trump and Bowers, boosting him but undercutting a centrist Republican for mayor? Or is it possible that a Trump-backing Republican could get elected mayor in a three-way race, just like the quite conservative Smith won a multi-candidate race in 2000? The math is inconclusive, but is it intriguing enough for a Republican?