Mary Washington Poll is a Red Klaxon

There’s always one poll that changes the narrative in every campaign. If it’s an outlier, the change doesn’t last; if not, it becomes the where-were-you-when-you-saw-that poll of the year for those of us who are proud citizens of Election Nerd Disneyland.

In 2009, it was Rasmussen – of all people – giving Creigh Deeds a 6-point lead (yes, Virginia, that really happened). Twelve years prior, it was the WaPo – yes, that one – that gave Jim Gilmore his first poll lead (I was getting my fingers ink-stained from the paper’s front page on my way to a graduate school class).

This year, it’s the University of Mary Washington, declaring that Terry McAuliffe’s lead among registered voters (+5) translates into a deficit among likely voters (-5).

Some context is in order here. Virginia has voted Democratic at the presidential level for each of the last four elections. Republicans haven’t won a statewide election here since 2009. This leads to two perfectly logical, yet firmly contradictory conclusions. The first is that Virginia, as a “blue” state, will be electing Democrats for the next generation. The second is that the law of averages must eventually catch up someday, and vigilance (within the majority party) is critical to avoid an upset.

Personally, as much distaste as I have for the RPV (and for the GOP in general), I’m a second assumption guy. Of the 19 states who have also gone Democratic for the last four presidential elections, ten of them have elected a Republican governor later than Virginia has. Additionally, the last time either major party won four Virginia gubernatorial elections in a row, my parents hadn’t even met (I turn 49 next month).

So, for Democrats, this is a far more uphill race than conventional wisdom held.

That doesn’t mean the UMW poll is a harbinger of November (see Rasmussen, 2009). Indeed, there may be issues with the poll itself. 40% of the survey was conducted on-line rather than by phone – and we don’t know how many of them made it to the Likely Voter sample. More to the point, said likely voter sample had a 54-46 gender split, in favor of men. Neither 2017 nor 2013 saw that.

Still, Democrats can’t just dismiss the poll for that reason. Rather they (ahem, we) should take note that the more Democratic gender is clearly the less motivated one at this point in time. Likewise, the 2020 vote preference among the likely voter sample shows Trump at 45% and Biden at 46%. Given that Biden won the state by 10 points, the enthusiasm gap is real, even if and no matter how it’s exaggerated.

So, after they get over the shock and use the aforementioned crosstabs to take the edge off the news, T-Mac et al need to redouble their efforts on getting Biden voters to the polls.

For the “base” voters, a slew of reminders that a Republican governor means the Big Lie hawkers have control of all election infrastructure should be a start. For center-right Biden voters demoralized by the Afghanistan debacle, greater publicity for fellow Afghan-withdraw critic Bill Kristol should help. Both should be more incentivized to vote by Youngkin encouraging folks not to take the COVID vaccine.

Just rolling your helmets on the field? That never really worked, and it certainly won’t now.