What Went Down During The 2021 Elections

I’ve been in Las Vegas for the past 32 nights, working on my book project and playing in my share of poker tournaments. A month in Vegas is a long time. But for better or worse, it tends to give you some critical remove from the daily stories that dominate the political news.

And with that perspective, here’s what I’d say: From a predictive standpoint, the outcome in Virginia isn’t likely to tell us very much about what will happen in the 2022 midterms.

That qualification — “from a predictive standpoint” — is important, because it means that Virginia probably doesn’t provide much marginal information given what we know already about 2022. And most of that news is already pretty bad for Democrats.

One thing we know about 2022, for instance, is that the president’s party — in this case, Democrats — tends to perform poorly in the midterms. Something else we know is that Biden has become rather unpopular, and while there’s still time for him to turn that around, an election held today would probably be pretty rough for Democrats.

There are some more optimistic indicators for Democrats: Namely, the generic Congressional ballot still shows them narrowly ahead. (Although most of those polls are conducted among registered voters, not likely voters, which could potentially reflect an “enthusiasm gap” between Democrats and Republicans.)

In that sense, it’s not surprising that Democrats are underperforming a bit in Virginia. Relative to Biden’s 10-point margin of victory in 2020, either, say, a 2-point win for McAuliffe or a 2-point loss would count as an underperformance.

Beyond that, I’m not sure we’ll glean very much insight. Virginia has had a decent predictive track record, but it’s a small sample size and it’s just one data point. Moreover, it’s a gubernatorial race, which — even as statewide elections become more nationalized — can turn more on local issues (say, education) than a national election might. A voter who agreed with Youngkin on some issues — say, lower taxes — but didn’t like Senate Majority Leader Mitch McConnell or the rest of the national GOP agenda wouldn’t be taking the same risk by voting for Youngkin tonight as they would in voting for a Republican candidate for Senate. Youngkin would be governor of Virginia and Virginia only, and won’t affect control of Congress.

But even if Virginia won’t tell us much about the odds in 2022, it could have a bigger effect upon how the 2022 election is contested. A lot of journalists and political elites live in Virginia (or in neighboring Maryland or Washington, D.C.) and are likely to overinterpret any signals coming from the state. (The same is true for places such as New York, of course, where FiveThirtyEight is headquartered.)

The polls are now closed in Virginia, as well as a handful of other races we’re watching, such as the Democratic primary in Florida’s 20th Congressional District. Now we sit back and wait …

To help understand the Virginia governor’s race as the votes are tallied, we created benchmark vote margins for each county and city using their 2020 presidential vote share. The way the benchmarks work is simple: If a place is voting more Republican than its 2020 benchmark, that’s a good sign for Youngkin; if it’s going more Democratic, that’s a positive development for McAuliffe.

Here are the benchmarks, based on the two-party vote in the 2020 presidential election (so just the Democratic and Republican votes) and ordered by each locality’s share of the statewide vote:

A few things to keep in mind. First, the numbers are based on the two-party vote, so it excludes third-party and write-in votes, which will probably constitute, at most, around 2 percent of the total vote. We did this because the Democratic and Republican votes are all that matter in a race without a significant third-party presence. (Not to say the lone third-party candidate couldn’t have an impact, as progressive candidate Princess Blanding could win a few left-leaning votes).

Second, these benchmarks are calculated based on Biden’s 10-point win in Virginia last November, but adjusted to represent a 50-50 race. Yet they are also based on the share of the vote each locality had last November, and some places might make up a larger or smaller share of the vote than they did in 2020. If bluer or redder places turn out at a very disproportionate rate, that’ll affect these benchmarks.

Lastly, we know that election officials will mostly report absentee votes first, which will be Democratic-leaning, so the types of votes each locality has reported will be an important caveat to keep in mind. As such, we plan to carefully contextualize these benchmarks tonight to avoid misleading you about how the race is going.