But since when have the right-wing bloggers and talk-show hosts let facts get in the way of a good story. Right-wing news headlines like "Can't Wait To See What Racist Al Thinks Should Count For Him" and snide commentary like, "As votes appear out of thin air failed radio host and humorless hateful pseudo comedian Al franken appears to be on the verge of stealing the Senate seat in Minnesota" are appearing all over the right-wing blogosphere. Even prominent Republican TV personalities like Joe Scarborough are suggesting that Al Franken is willing to "steal" votes in order to prevail against Senator Norm Coleman. Media Matters reported that MSNBC analyst Pat Buchanan recently said, "Look, you've got a station wagon up in the Iron Range that's hold -- got more of them in there, I'm sure, Joe." Scarborough continued this line of thought and later said, "Buchanan says Al Franken can steal 130 votes easily. He said he can usually hide 130 votes in the back of a station wagon." Not to be out done of course, Minnesota congresswoman Michelle Bachmann has jumped into the fray as well, exclaiming on FOX's popular Hannity & Colmes show that Al Franken, "wants to stuff the ballot box with rejected ballots."
Regardless of what the Media wants the rest of the country to believe, the hard-working volunteers here on the ground in Minnesota continue the tedious recount without complaint. While Minnesota's election process is highly regarded across the country, a number of factors including the use of optical scanners can increase the number of so-called "undervotes" in most races. Some undervotes are purposeful when voter opts not to cast a vote in a particular race or on a ballot question. But in other cases, they result from voters who may have intended to vote in a particular contest, but marked their ballot improperly.
A number of recent studies and media reports have put the total number of undervotes in the Minnesota senate race at close to 34,000 votes. While undervotes are not uncommon for multiple-race ballots some experts believe many of those may be the result of improperly marked ballots. Electronic voting machines of course are set within specific parameters to record votes on ballots that meet these parameters, but machines cannot obviously make decisions about a voter's intent if the voter was not able to cast a clean ballot or did not follow proper instructions. Minnesota election law is very clear that if a voter's intent can be determined that vote must count. Because of this law, recounts are done by hand. Thus in many cases the voter's intent is likely to be clear enough to election officials for those ballots to count and add to the vote totals of either candidate during the recount process. Thus it is not entirely surprising that someone not in the know or intellectually incurious might think that ballots are appearing in thin air for a particular candidate during a recount of this nature.
We did some analysis on the preliminary recount numbers in the counties that have already finished the recount process or are underway. Based on the current trends and from our analysis we believe the following prominent counties will add the following vote totals to each candidate in addition to what the vote spread was at the end of counting on Thursday.
St. Louis = 22 more votes for Franken
Anoka = 9 more votes for Coleman
Ramsey = 73 more votes for Franken
Dakota = 21 more votes for Franken
Stearns = 3 more votes for Franken
Hennepin = 23 more votes for Coleman
Washington = 3 more votes for Coleman
Polk = 5 more votes for Franken
When looking at some of the counties that have finished counting Filmore county may become fascinating to watch because the Coleman camp challenged 27 votes for Franken and the Franken camp only challenged 3 votes for Coleman. And in Watowan county the Coleman camp challenged 6 more votes than the Franken camp. Interestingly the populations in Filmore and Watowan counties were nearly evenly split for Franken and Coleman. Making the assumption that the challenged votes will ultimately followed the overall voting patterns in these two counties it is not beyond the realm of possibility that Franken could pick up another 30 votes just from these challenges. That would leave him down only 17 votes with almost another 1800 challenged votes to be decided upon. Franken would only have to win 50.5% of these challenges to come out on top.
All of these calculations are based on the rather conservative estimate based on current trends that Norm Coleman would actually pick up more votes than Al Franken in Minnesota's most urban county, Hennepin. Franken handily beat Coleman in Hennepin county but so far very surprisingly Franken is not picking up any additional votes on Coleman within this strongly leaning Democratic county. Counting in Hennepin will continue through next week and any changes in our estimates for Hennepin for either candidate could greatly change this already extremely tight recount. Sherburne and Wright counties, which have not started their recounts, are strongly Republican and have moderate populations and could add to Coleman's numbers.
Lastly, a recent study of the "undervotes" in this Minnesota race was assessed by a couple researchers at Darmouth and Harvard. Among their findings they describe the percentages of undervotes throughout the state of Minnesota for the current Senate race. We took their analysis one step further and re-displayed their "undervote" map based on areas that were either Republican voting (red) or Democratic voting (blue) and is displayed in the figure below. In the figure areas displayed in red would be Republican voting precincts that had more than 2% of undervotes for either candidate and the areas in blue would be Democratic leaning precincts with more than 2% undervotes. Without the actual recount data there is no way to determine which candidate the undervote might go to. An interesting phenomenon however that came out of the 2000 Al Gore and George Bush Presidential race was that in the Florida recount Al Gore picked up more votes in Republican areas than he did in Democratic areas. By this measure we think the following counties that have not yet started their recounts could be interesting ones to watch Scott, Blue Earth, Koochiching, Lake of woods, Rock, and Wilkin.
In the end political junkies and pundits will not be able to take their eyes off this race for at least another 3 weeks and the first clear idea of who really won will unfortunately be decided during the counting of the challenged ballots by the State canvassing board in mid-December. After that process the number of possibilities will probably expand again in terms of court challenges.