This post has been updated.
Rep. Eric Cantor's unexpected Republican primary loss Tuesday kicked off rapid speculation on how the House majority leader lost to a poorly-funded opponent in a Virginia district whose primary Cantor won with 79 percent support in 2012. One possibility is that district Democrats crossed over and voted for Cantor's opponent, David Brat.
Virginia's lack of party registration makes it difficult to pin down whether Democrats crossed over in large numbers, but local level turnout provides some indirect clues on whether this phenomenon was widespread. On two counts, the data cast doubt on whether Democratic cross-over voting caused Cantor's loss.
While Republican primary turnout spiked by 28 percent over 2012, according to the State Board of Elections, Cantor received nearly 8,500 fewer votes this year than he did in the 2012 Republican primary, a drop that was larger than Brat's 7,200-vote margin of victory. Regardless of how many Democrats turned out to oppose Cantor, he still would have prevailed had he maintained the same level of support as in his 2012 landslide.
If Democrats showed up in large numbers to vote against Cantor, turnout should have spiked highest from 2012 in Democratic-leaning areas, with Cantor seeing an especially large drop-off in support. In fact, turnout rose slightly more in counties that voted more heavily for Mitt Romney in the 2012 presidential election.
Likewise, Cantor saw the biggest drop-off in support in Republican strongholds of Hanover (-44 percentage points) and New Kent (-44 points), counties where Obama drew just over 30 percent of the vote in 2012. In Henrico County where Obama won 55 percent support, Cantor's drop-off was a smaller 32 points. The overall correlation between Obama's county support and Cantor's drop-off was clearly negative at -.60, indicating that the higher Obama's 2012 support, the lower Republican primary turnout rose this year. This is consistent with the idea that Republicans largely drove the rise in turnout.
Update: Turnout expert Michael McDonald of the United States Elections Project found similar results analyzing precinct-level data Tuesday night, reporting GOP primary turnout was lowest in the most Democratic-leaning areas of the state.
Weak evidence Ds swung VA-7. Highest D precs have lowest total votes (not all localities in VA-7, but highest D incl) pic.twitter.com/iqxWgwFKkj
— Michael McDonald (@ElectProject) June 11, 2014
The evidence of Democratic cross-over voting is weak even after narrowing to the precinct level to 10 locations in Henrico County, which all voted 60 percent or more for Cantor's general election opponent in the 2012 general election. If Democrats banded together to embarrass Cantor in the GOP primary, this is where it would be most evident. Indeed turnout did rise an average of 66 percent above the 2012 primary level in these precincts. But vote shares don't bear the rest of this out -- on average Cantor won 24 percentage points less support in these precincts, smaller than his 35-point drop in support statewide.
Some Democrats surely selected a Republican ballot and voted for David Brat, but Cantor's loss seems to be much more the result of weak support among Republican voters, some of whom showed up for a race they typically ignore to vote for the tea party conservative who was besieged with attack ads.