What the Virginia Governor Race May Portend for 2018
People often complain about how frequently the U.S. holds elections. That’s understandable. Only six months after a national election in which 60 percent of eligible adults cast their ballots, voters in Kansas, Georgia and California have had to vote to fill congressional seats. And in the coming months, South Carolina, Virginia, New Jersey, Georgia and Montana will fill major positions -- either by sending new members to Congress or new governors to state capitals.
But there are benefits to the never-ending stream of races. Frequent elections (along with polls) help politicians figure out if the American people approve of what they’re doing. If a new president overreaches early on (as often happens) then voters can push back in off-year and midterm contests. The primary fights for these offices can also reveal important splits within a party before they materialize on the national level.
And in 2017, the most important electoral gut check will probably come from Virginia. The Old Dominion is the only swing state that holds a gubernatorial election in an off year, and there are contested primaries on both sides. Once the dust settles from string of special House elections from now till late June, many journalists, activists and political donors will likely start to focus on the race there.
So it’s worth asking -- how should we interpret what’s happening in Virginia?
There’s more than one way to tackle this question, and the governor’s race is multifaceted. So I’ll keep it simple and run through three narratives (some of which have already been floated elsewhere in the press), assess how accurate they are and thus get a handle on how to interpret the eventual results.
So Far, the Democratic Primary Isn’t Round 2 of Clinton vs. Sanders
At first glance, the Democratic gubernatorial primary looks a bit like the Democratic presidential primary of 2016. Ralph Northam, Virginia’s lieutenant governor, declared his candidacy early and gained the support of much of the state establishment. But Tom Periello, a highly progressive former congressman, unexpectedly decided to run as an outsider and is trying to squeeze Northam from the left. And while there’s a real insider-outsider dynamic in this race, the analogy is far from perfect.
First, Ralph Northam isn’t Hillary Clinton.
Clinton worked in politics or policy for the majority of adult her life and has always been a Democrat. Her ideology changed over time, but her stances mostly reflected those of the typical Democrat. She is also, obviously, a woman. Northam’s biography doesn’t match up. He spent most of his life serving in the military or practicing medicine (sometimes simultaneously). He voted for George W. Bush twice before being elected as a Democrat to the state Senate in 2007. And in 2009, Republicans attempted to get him to switch parties in an effort to tip the Senate majority in their favor. Northam rejected those entreaties, was re-elected in 2011 and has been serving as lieutenant governor since 2013.
This contrast isn’t meant as a dig at either candidate. Both are clearly qualified for the offices they’ve sought. But the differences in their personal and ideological history are obvious.
Second, Tom Perriello (pictured) is not Bernie Sanders.
If you want a great, in-depth treatment of this topic, I would recommend this piece by Clare Malone at FiveThirtyEight. Here’s my version of the argument: Sanders has been a socialist and an Independent running on a relatively consistent liberal platform for most of his career. Perriello represented a conservative congressional district for one term -- during which he supported the Stupak-Pitts amendment (a failed effort to limit Obamacare abortion coverage) and got an “A” rating from the NRA. His DW-NOMINATE score (which measures economic liberalism using voting records and complicated math) during his congressional term put him to the right of the vast majority of his fellow House Democrats. Perriello has since changed some of his policy positions, and he’s bolstered his progressive credentials by working for John Kerry’s State Department, the Center for American Progress and the Obama administration. He even earned Sanders’s endorsement earlier this year. But his liberalism hasn’t been as reliable as Sanders’s, and he’s overall more of an establishment candidate (see endorsements by David Plouffe and John Podesta) than Sanders was.
Most importantly, if this primary was a rerun of the 2016 Democratic primary, Northam would be winning handily. Clinton beat Sanders in the Virginia Primary by nearly 30 points, in part by winning support from blacks across the state and suburbanites near Washington.
But most non-partisan polls show a close race with a high number of undecideds (a recent PPP survey did show Northam significantly ahead with a coalition somewhat resembling Clinton’s, but the poll was sponsored by a group that endorsed him). Obviously, the polls can (and will) change, and maybe the battle lines will end up looking something like a Sanders-Clinton rematch. But we’re not there yet.
Trump Will Matter -- But Sometimes Gubernatorial Races Are Weird
Analysts will likely also use the results of Virginia’s gubernatorial contest to try to measure how voters feel about the current administration. I think there’s some merit to that (see the final point). But it’s worth noting that gubernatorial races aren’t perfect referenda on the president for a number of reasons.
Gubernatorial races don’t revolve around the same issues as presidential or senatorial elections. Governors generally aren’t involved in foreign policy decisions, and they often don’t get the opportunity to directly vote on the president’s agenda in the same way that a senator or congressman does.
Voters understand that, and that’s why the relationship between presidential results and senatorial results is stronger than the relationship between presidential and gubernatorial results. (I measured this by mapping presidential lean onto Senate and gubernatorial results while controlling for incumbency and then comparing fits and related statistics.)
While ticket-splitting is certainly declining, gubernatorial candidates are less tied to the national party than congressional candidates. That’s why Republican Phil Scott is the governor of deep-blue Vermont and Democrat Jim Justice is the governor of West Virginia, one of Donald Trump’s most-supportive states. The list of governors whose parties don’t quite fit their state is reasonably long: Republican Charlie Baker in Massachusetts, Democrat Steve Bullock in Montana, Republican Larry Hogan in Maryland and more.
If Trump’s approval rating stays low, he might still cost the eventual GOP nominee (probably Ed Gillespie) the governorship. But it’s important to keep an eye on local conditions and how each candidate performs, because that can shift the race.
Virginia Is an Important Test of What Happens When Trump Is Off the Ballot
Virginia could provide valuable information about what happens to both party coalitions when Donald Trump is off the ballot. Specifically, it gives us a first glimpse at whether college-educated whites will view the GOP as the party of Trump, and if non-college-educated whites will stick with the Republicans in conditions that are less favorable than 2016.
Hillary Clinton (depending on your data source) either won or came close to winning college-educated-white voters on the national level.
That’s not normal.
In presidential contests, white voters with a college degree typically lean Republican, and the GOP candidate almost always wins them by a healthy margin. And in Virginia, that trend doesn’t stop at the top of the ticket. In 2014, Gillespie barely lost a Senate race to incumbent Democrat Mark Warner while winning college-educated whites by 10 points. In 2009, Republican Bob McDonnell won the group by 22 points, according to exit polls. Exit polls (which have their issues) showed Trump winning the group by only four points in Virginia. And, when compared to Mitt Romney, Trump tended to lose support in the counties and cities where white levels of education were the highest.
(Note: Each bubble on the chart represents an administrative subdivision within Virginia. The vertical position represents the shift between 2012 and 2016, the horizontal position represents the rate of college education among whites and the size shows the population.)
If Democrats can hold or expand on these gains when Trump is off the ballot, that would be a good sign for them. It would indicate that the GOP is the Party of Trump up and down the ballot, which would rightly worry House Republicans in well-educated, white districts.
If Republicans manage to do well with these well-educated whites, it might be an initial sign that they can hold on to them while driving their numbers up among rural, non-college-educated white voters. And if Trump wants to expand his coalition in 2020, that might be the easiest place to start.
Finally, it’s important to watch the working-class and rural white vote in 2017. This group has slowly but surely become more Republican over the last couple of decades, and Trump’s candidacy represented another step in that direction. If Democrats can roll back those gains just a bit, it’ll put them one step closer to making a national comeback.
Right now, the Republican Party is heavily leveraged on non-college-educated white voters. Democrats currently win commanding majorities of every major nonwhite group, and Republicans fought them to a draw with college-educated whites in 2016. In other words, Trump’s version of the GOP is heavily reliant on working-class white voters. And whenever a party is too reliant on one subset of voters, its coalition can become unbalanced and vulnerable.
Consider the Obama coalition. Many analysts thought that President Obama’s version of the Democratic Party would be able to win presidential elections for the foreseeable future by keeping high turnout and margins among minority voters. But when Obama was off the ballot in 2016, black turnout dropped in key states. That was one factor that led to Hillary Clinton’s defeat.
So Republicans should keep an eye on working-class white margins in Virginia this November. Losing even some of those voters while failing to gain others would represent a significant problem for them.