About That Recount...

(AP Photo/Getty Images)
The following e-mail dropped into the "C&C" mailbag this morning. It's a detailed analysis of what might happen with a Virginia Senate race recount, from someone in a position to know. -- Ed.

I ran both the 1989 statewide recount for Doug Wilder and the 2005 recount for Creigh Deeds. It's nice to be on the sidelines for the first time in state-wide recount history. The Wilder and Deeds recounts were about as many as I wanted to run in a lifetime!

There is a big lesson for us this year from those recounts:

The Certified Winner will not lose a recount unless the certified results differ by 50 votes or less. Whichever candidate wants to change the election night result MUST do it in the next few days during the canvas process.

Here is why:

1. The Wilder recount changed about 20 votes net from the CERTIFIED results. The Deeds recount changed about the same number from the CERTIFIED results. Since the certified results are based on the results of the canvases that will be conducted over the next few days, history tells us that there is virtually no chance the RECOUNT will make up a 100 vote difference in the certified results, much less nearly 8,000.

2. Attorney General McDonnell drastically decreased what little chance the CERTIFIED losing candidate might otherwise have of winning a recount. Deeds argued to the 3 judge court that in his recount the counting calculation in the voting computers should be re-run as a part of the recount and scannable paper voting sheets be re-run through the scanners . McDonnell argued that the counting calculation should not be run again, but rather the results showing on the machine should simply be rechecked to insure accuracy in the transcription of it by the voting officials and scannable ballots should not be rescanned absent affirmative evidence of inaccurate counting. McDonnell won both arguments, virtually assuring that the results from the computer and scan card voting machines would not change. It is inconceivable that McDonnell would argue or the 3 judge court would grant any different process this year. Since nearly everyone voted this year on the computer, the recount cannot make any real change. The few changes that were made in '89 and '05 came almost exclusively from paper ballots. There won't be enough paper ballots in this recount to make a couple of hundred vote difference.

3. The election night loser's ONLY chance to win lies in the canvas. If you check the differences between the vote totals reported in the "next day" press in each of the last five or so statewide cycles, you will see anywhere from a 400 to 5,000 net vote shift from the "next day" results to the Certified Results. For example, in Allen's race against Mary Sue Terry, the certified totals for both candidates resulted in an over 4,000 net vote loss for Allen between the results reported the morning after the election and the certified results announced a couple of weeks later. I did a chart of several of these results during the Deeds/McDonnell process after Rosenthal and a couple of other ex-AGs held a press conference to suggest that it was suspicious that Deeds was picking up votes as the canvas results were reported to the Board of Elections. In point of fact, Deeds "gain" in the canvas process was pretty typical. In almost every state-wide race, one candidate or the other gains significant numbers of net votes - FAR, FAR more than were gained in either of the recount procedures.

In short, Allen might as well forget about winning a recount and putting VA and the nation through weeks of waiting (remember, under the statutory timetable, a recount cannot realistically be completed until the week before Christmas). However, Allen and Webb better have a lot of lawyers on the ground over the next three days to observe the canvases. The canvases are the ONLY place this election can change.

Larry Framme