Virginia Polls Should Alarm Democrats

O ne of the hazards of projecting national trends from a small number of off-year elections is that they may take place in states and localities that are unrepresentative of the country. Democrats have been doing poorly so far this year in special elections, yet they did very well, despite a summer-long scare in the polls, in the California recall. The campaign to recall Gavin Newsom led in the polls all August, trailed by 14.5 points in the final polls by mid September, and lost by 27 in the actual voting. That outcome should also remind us of the hazards of getting overexcited about early polling.

The big-ticket races this November are all in blue areas, such as the governor and House of Delegates (the lower house of the state legislature) in Virginia, the gubernatorial race in New Jersey, and the mayor’s race in New York City. But that also means that there can be warning signs even if Democrats pull out victories across the board.

Opinion polling in Virginia in recent years has not always been the most reliable; in 2014, for example, some pollsters rather famously suppressed polls showing a late surge for Ed Gillespie against Mark Warner, thinking the polls were embarrassingly out of step with the consensus. Had they all been published, there would have been more warning that Warner would win by only 0.81 percent of the vote. As the state has turned bluer, misses of that many Republicans have become rarer; the RealClearPolitics average nailed the presidential vote in 2016 in Virginia, and underestimated Ralph Northam in 2017 when he beat Gillespie by 8.9 points after leading by 3.3 points in the final poll average. In 2020, however, the poll average overestimated Democrats again; Joe Biden was ahead by 11.5 points and won by 9.4, and Tim Kaine’s reelection was ahead by 17.8 and won by eleven.

With those caveats in mind, let’s take a look under the hood of current polling in Virginia — both the gubernatorial campaign between Terry McAuliffe and Glenn Youngkin and the presidential approval ratings for Joe Biden.

First, let’s look at a comparison of how Democrats have fared in the exit polls with various subgroups in Virginia over the past five years with how McAuliffe is polling with them. I have not included here all the recent polls, because some of them did not have easily available crosstabs showing polling among the relevant groups.

McAuliffe is in a closer race than we have seen in the last five years of Virginia elections, so we would expect to see him losing ground in some places. It remains noteworthy where that is happening. Independent voters, largely divided in 2016–17, broke hard for Democrats in 2018–20. Nearly all the recent polling in Virginia shows them swinging back in significant numbers to Republicans. That is a common trend against the party in power in off-year elections, and it bodes poorly for Democrats.

Hispanic voters have been a real concern lately for Democrats. Only a few of the recent polls break out McAuliffe’s showing with Hispanics, but the two that do show erosion. Even if we consider the Emerson poll (55 percent of Virginia Hispanics backing Youngkin) to be a bit of an outlier, the trend here is alarming. So is Youngkin’s nearly doubling the Republicans’ small share of the black vote in Virginia in two of the four polls. Trends with non-college and female voters are likewise troubling for Democrats.

Virginia being as blue as it is, McAuliffe may yet survive the national trends. Like Newsom, he is trying to do so with a campaign that mostly ignores his record, his party in Washington, his opponent, and the issues facing Virginia, and focuses laser-like on the one thing holding the Democratic coalition together: Donald Trump. Even the most Trump-obsessed precincts have noticed this.

Amanda Carpenter in the Bulwark:

The Democrats bet they could turn Youngkin into a Trumpkin, but Youngkin managed to keep the Trump stink off his name and make the race about something else entirely: education. That’s left McAuliffe and the Democratic Party of Virginia, who apparently think it’s a brilliant idea to troll Youngkin supporters with a giant chicken balloon that looks like Trump, fumbling for a strategy. . . . Team McAuliffe’s response? “Glenn Youngkin’s entire campaign has been based on Donald Trump’s divisive conspiracy theories, and tonight we saw more of the same — angry Trumpian conspiracy theories and constant threats against public school funding.”

Whether Youngkin’s attack is fair — whether it is right to blame McAuliffe for the school issues that people are angry about — is not the point. The charge, which comes with a bevy of policy proposals, warrants something more substantial than another chant about Trump. . . . The notion of tying Youngkin to the unpopular former president — Trump lost to Biden by ten points in Virginia — isn’t a dumb one per se. But it seems to be a generic template straight out of national party headquarters.

Jonathan Martin of the New York Times, in an article noting the apathy of Virginia Democrats towards McAuliffe, says:

In interviews outside Fairfax’s early-voting site, every McAuliffe voter cited Mr. Trump as a reason for supporting the Democrat. Transportation, education and taxes — longtime core issues of Virginia governor’s races — were scarcely mentioned.

James Hohmann writes in the Washington Post that:

Democratic strategists involved in the race argue that [Youngkin’s] focus on CRT will backfire by stoking their base as well, especially African Americans. Their research shows many voters associate CRT with Trump, who lost Virginia last November by 10 points. Democrats hope that the more Youngkin mirrors the former president’s messaging, the more their own people will show up. . . . McAuliffe responded [to criticisms of his own debate gaffe] with a commercial claiming that “Youngkin would bring Donald Trump and Betsy DeVos’s education policies to Virginia.”

If Trump was still in office, this would likely be a viable midterm strategy for Democrats nationwide, as it was in 2018. But consider that there is a new president, Joe Biden, and this is how his polling looks in Virginia:

On this chart, for space and simplicity, I’ve averaged out the previous-year exit polls from the first chart. If the news in McAuliffe’s polls is concerning for Democrats, the news in Biden’s polls is awful. Across the board, independent voters in Virginia have turned heavily against Biden, after being supportive of Democrats in recent years. Non-college-educated voters tell the same story. Young voters are markedly less Democratic. Hispanics range from modestly less supportive to outright split. Black voters are dissenting from Biden at twice the rate before. The Democrats’ decisive advantage with women in Virginia is gone.

This, mind you, is not only voters in a blue state, but in the state where anti-Trumpism is most firmly grounded, from the state’s large population of federal employees to the memories of Charlottesville. Project all of those shifts into any electorate less naturally favorable to Democrats, and the outcome is a bloodbath. And that is before we consider the likelihood that shifts in opinion also drive shifts in the relative turnout of different groups. No wonder we are already seeing headlines like this one, in The Hill: “Democrats face grim political reality in midterms.”

I’ll return another day to the broader national-polling picture. But even if Terry McAuliffe survives this environment on the strength of Virginia’s partisan tendencies and his own well-known name and reputation in the state, if the results on Election Day 2021 look anything like the polls, the prognosis for Democrats a year down the road will be grim.