Virginia governors are only allowed to serve one term, which means that every four years, voters are treated to the turbulent highs and lows that usually accompany an open gubernatorial election.
This year’s contest is no exception: Lt. Governor Ralph Northam is competing with Republican political consultant Ed Gillespie in what has become November’s most hotly contested race.
Polls for the 2017 election have, for the most part, favored Northam. However, over the past month Gillespie has bested Northam in three polls and tied in another, and some Democrats are worried that the party might lose yet another governorship (there are 34 Republican governors and only 15 Democrats).
Former Congressman Tom Perriello, whom Northam bested in the primary, told Politico earlier this month that the mood within the party could be described as “anxious optimism.”
“People’s anxiety comes from the fact that the Democratic coalition doesn’t always show up in these off-year elections, and there’s an erosion of confidence in polling,” added Perriello.
It certainly doesn’t help that the seat flipped in both the 2014 and 2010 elections, and neither party has had a firm grip on it for decades.
In 1970, Republican Linwood Holton—who is U.S. Sen. Tim Kaine’s father-in-law, incidentally—won the Virginia governor seat, breaking a nearly century-long streak of Democratic control in the statehouse. Since then, the position has become increasingly prone to flipping. Two more Republicans succeeded Holton, followed by three Democrats, then two Republicans, then two more Democrats, then another Republican, which brings us to Democrat Terry McAuliffe, the current incumbent.
But while this odd, off-year race in a swing state might feel pivotal to next year’s mid-term elections given the rocky political climate lately, experts say that it’s actually not.
Last month, FiveThirtyEight analyzed historical election data that proved Virginia’s gubernatorial races have not been predictive of the outcome of mid-terms since at least 1993.
Additionally, a single race is highly unlikely to indicate any kind of broad pattern, and party affiliation is a much weaker factor in the outcome of gubernatorial races than on offices farther down the ballot. In fact, gubernatorial elections have never been much of a bellwether for federal races, and they are terrible predictors for presidential elections.
While it may not tell us what to expect in 2018, Virginia’s gubernatorial race will still have a national impact, first and foremost because the winner will preside over the congressional redistricting that will take place in 2021 after the U.S. Census is completed.
Because each state conducts its redistricting independently, the states’ majority parties have on many occasions redrawn the lines in their own favor, a process known as gerrymandering.
For instance, Texas has been in and out of court since 2011 over it’s post-2010 census redistricting plans. After the first plan was rejected, a second set of maps were drafted in 2013 and a federal court in San Antonio subsequently declared that several of the districts marginalized minority voters.
The case made it’s way up the Supreme Court, where justices decided 5-4 that the redistricting plans could be implemented while the court considers the state’s appeal of a lower court’s order to redraw the lines.
According to David Wasserman at The Cook Political Report, another strong influence of Virginia’s governor race will be its ability to drive voter turnout in the state legislative races, and those seats do matter when it comes to 2018 predictions. Wasserman says that if Democrats pick up at least 5 to 10 seats in the Virginia House of Delegates, it could signal a corresponding shift in the U.S. House next year, as the 2009 state legislative races did for the 2010 mid-term elections.
With plenty at stake, it’s no wonder that the yo-yoing polls have caused anxiety for Democrats, but that’s just how polls work. Statistician Nate Silver says that when poll results resemble each other too closely, it is usually an indication of herding, which is “the tendency of polling firms to produce results that closely match one another, especially toward the end of a campaign.”
According to Silver, when drawing broad conclusions about the general population based on a sample group, there is an inherent and unavoidable amount of error—specifically, he says, at least 32 percent. “If you’ve collected enough polls and don’t find that at least 32 percent of them deviate from the polling average by 3.5 percentage points, it means something funny — like herding — is going on.”
So, the disparity in Virginia’s polling results means that it is less likely that herding is occurring, and that the overall polling averages might actually be good indicators of the state of the race—unlike in races where herding does occur.
We’ve gathered polling data over the past three months, and polling averages put Northam at 46 percent and Gillespie at 41 percent. Accounting for the ± 4.1 percent average margin of error, that gives Northam a slim, one-point lead over his Republican opponent:
Virginia voters will head to the polls on Tuesday.