Strange decisions made by the Minnesota Canvassing Board

Now that it seems we'll have to wait until the new year to find out who will be the next senator from Minnesota, people are starting to review some of the decisions made by the Canvassing Board in how they allocated votes to either Coleman or Franken. Apparently, they indulged themselves in a bit of mind-reading in trying to determine voter intent. And, as John Lott details, they sometimes applied those rules inconsistently and the result has been for Franken to go into the lead. For example, Coleman ovals that were darkened in and then an X was put over the oval were determined to be an indication that the voter had first voted for Coleman and then changed his/her mind. However, when the X was over a Franken oval, a different decision was reached. Ostensibly, the decision driving several of those pro-Franken decisions relied on the voter using the X over a colored-in oval for both the presidential and senatorial races and then supposedly wising up and using just ovals for the rest of the races. Perhaps that is what happened. But we don't really know. In their quest to divine intent, the Canvassing Board is assuming that two X's over filled-in ovals just mean an error that was corrected and not a decision to change a vote but one X doesn't means that the voter had changed his mind. Forget that a voter could ask for and get a new ballot. Such mind-reading is a very iffy thing and we shouldn't be reduced to reading minds to determine who won a Senate seat.

Franken is now up by 46 votes. Lott details how more than that number can be found in contrary decisions that the Canvassing Board made about how to count missing ballots.

There has been much arbitrariness over whether to count newly found votes. Compare these 133 additional votes to the 171 votes that were found in Ramsey County’s Maplewood Precinct 6. In Precinct 6, Franken argued that the original vote total had missed the 171 votes because the voting machine was initially not working and the 171 were not rerun through the new machine that replaced the original defective one. However, Republicans were concerned that the newly “found” ballots might not have been really cast on Election Day.

Since the recount showed that there were 171 more ballots than recorded votes, the decision was to add in the ballots, giving Franken a net gain of 37 votes.

Contrast this to the 133 ballots in Minneapolis. The 133-vote problem arose because more votes were recorded on Election Day than were found when the recount was conducted. The initial explanation was that the ballots must have been accidentally run through the voting machine twice — that no votes were missing, but that 133 had just been accidentally counted twice. The Canvassing Board however decided not to rely on the recount and instead on the original machine totals.

The only commonality in these two decisions was that the outcome benefited Franken. When the recount is in Franken’s favor that is used. When the original machine tally works best that is used.

If the whole idea of the recount is that the recount is more accurate than the original count at the end of a long and tiring election day, why choose the election day number instead of the recount number? The beauty of having a Canvassing Board make these decisions is that they could apply consistent rules to the count. However, it seems that consistency is a hard goal to reach when you're dealing with the myriad ways that people can fill in an oval or count a ballot.