On November 5, 2013, Virginia held statewide elections for Governor, Lieutenant Governor, and Attorney General. In the Attorney General race, then-Virginia Senator Mark Herring (D-33rd) faced then-Virginia Senator Mark Obenshain (R-26th). The winner is determined by simple plurality.
The result of the race certified by the Virginia State Board of Election had Herring winning by a margin of only 165 votes—about 0.0075% of the total. Off on a Tangent is unable to verify the result or determine the winner beyond a reasonable doubt.
At 12:03 a.m. on November 6, during my live coverage of the election, Obenshain was ahead by enough of a margin that it was mathematically impossible for Herring to close the gap. I called the race accordingly. The next morning, published results from state election officials had changed drastically (with no explanation) and were showing Herring in the lead. I reluctantly rescinded my call at 9:10 a.m. By noon, the published results were again showing Obenshain in the lead. That evening I again calculated that it was mathematically impossible for Herring to pull ahead, and called the race for him again at 10:40 p.m.
There were rumors that “thousands of votes” in Fairfax County had not yet been counted, which is part of why I did not call the race again earlier in the afternoon. I made my (second) call only after Fairfax County election officials categorically denied that any uncounted ballots remained. There were not enough provisional and uncounted ballots left in the state for Herring to make significant gains.
On November 8—the Friday after the election—Fairfax County officials announced that “nearly 2,000” votes had not been counted due to a “technical error” with a voting machine in the county’s 8th District (a Democratic Party stronghold). The reported “nearly 2,000” votes ended up being 3,033 votes, and two-thirds of them went to Herring.
Officials claimed the “found” ballots were stored in a secured facility for all those days . . . but their chain of custody cannot be verified, and the circumstances around their “discovery” are extremely suspicious.
The number of ballots affected by these irregularities is more than eighteen times Herring’s alleged margin of victory, and similar (but smaller) irregularities like this also occurred in the City of Richmond. My precinct-level analysis suggests (but does not prove) that Herring gained at least 450 votes through manipulation or fraud in Fairfax County, Richmond City, and elsewhere, which, if true, was more than enough to have turned the election in his favor.