It wasn’t just Virginia. The contest in North Carolina, when Biden’s lead is a more modest 19 points, was also called quickly. In both states, as with so many of the contests this year, Sanders earned a lower percentage of the vote than he had in the Democratic nominating contest four years ago. What’s remarkable about the shift over those four years, though, is how geographically pervasive it was. Sanders’s support in Virginia and North Carolina contracted, with Biden winning a number of precincts that Sanders had secured four years ago.
Consider North Carolina. Below are three maps, showing the 2016 presidential vote by precinct, the per-precinct vote in the 2016 primary and the vote on Tuesday night. In the 2016 primary, there is a lot of green, representing Sanders, in the western part of the state, as well as some closer to the eastern coast. On the 2020 map? A lot of that turned blue. In 2016, Sanders won 35 percent of precincts, but in 2020 that number shrank to 7 percent.
The 2016 map is included because the red-blue spectrum from that election cycle provides a lot of demographic and geographic information. You can easily pick out cities (blue, ringed with white swing precincts) and large precincts, which denote rural areas.
In Virginia, the pattern is similar. There was less Sanders green in 2016, and most of it evaporated on Tuesday night. Sanders won half as many precincts on Tuesday as he had four years ago.
You can see where Sanders maintained his support in Virginia: in the far western part of the state and southwest of Roanoke.
In both maps, the areas where Sanders lost support appear at a glance to be areas that voted more heavily Republican in 2016. In Virginia, where precinct-level data from the 2016 general election were available, that wasn’t actually the case. The places where Sanders’s percentage of the vote dropped the most since 2016 were generally the places where Hillary Clinton did the best against Donald Trump.
It’s hard to know what this means without other information. Is it a function of Biden performing more poorly in places that lean Republican? Of Sanders supporters in more heavily Clinton areas staying home? Of those supporters having been more likely to back Sanders in 2016 out of opposition to Clinton?
What’s particularly interesting about Virginia is that Clinton won the state by about the same margin as Biden did on Tuesday. Despite that, Sanders did about 12 points worse in each Virginia precinct on average than he had in 2016, while Biden fared only about 6 points worse on average than Clinton had. One result? A much more geographically extensive victory.