Where are the missing Senate votes?

There’s an unusual math problem I’m having as I look at MPR’s election results.

Number of votes cast in the race for president:

Number of votes cast in the race for Senate:



Twenty-five thousand Minnesotans voted for the top of the ticket and left the second race blank? It’s possible. But why?

(More calculations to follow)

(2:39 p.m.) The AP has a slightly different Senate total: 2,880,764. I believe the AP’s vote tallying is independent of the Secretary of State.

(2:52 p.m) The AP has 2,898,691 votes cast for president. That’s a difference of 17,927 votes. Still that’s quite different from what the Secretary of State has, leading one to believe the AP might be closer to accurate, except that the Secretary of State has 100% of the votes counted, and the AP still had precincts outstanding.

It’s also quite possible the AP numbers don’t include absentee ballots.

(3:06 p.m.) Not surprisingly, the biggest drop in votes is in Hennepin County (I’m using the AP figures at the moment) where the difference in votes cast in the two races was over 4,000.

Percentage-wise, the biggest drop was in Lake County (4%), followed by Houston (2%), Lake of the Woods (2%), Traverse (2%), Wilkin (2%) and Winona (2%). Sen. Norm Coleman won four of those six counties.

(3:27 p.m. ) In comments, people smarter than me say this is a normal thing. I downloaded the spreadsheet for the 2000 race — the last time we had a presidential contest and a Senate contest) and I get virtually no difference in the vote. In fact, I get that there was just 1 more vote cast in one race than the other. I may be making a mistake there so double-check the math.

  • I am actually surprised it isn’t higher.

    I’ve heard from numerous people working phones and door-knocking in CD5 a few months ago that many people were saying they were going to “vote for Obama” (or McCain) but didn’t care about anything else on the ballot. Granted, organizers did great work to get people to fill more of the ballot out (the joint Obama+Franken lit piece helped a bunch) but even still I’m not surprised that number isn’t larger.

  • Jeff

    i too am not surprised that number isn’t higher.

    if the Dems could have just fielded an even remotely decent and much less polarizing candidate for Senator this year, Obama’s support would have handed that Dem the victory

  • I can’t find same-day registration, but I would think it would be higher than that number – meaning a lot of people showed up only to vote for Obama. Its actually quite common.

    In addition, those may have been people who hated Coleman and Franken, but did not know who Barkley was. “A pox on both houses” is another common reason for drop-off.

  • MR

    Maybe Barkley should have changed his last name to Anderson.

  • GregS

    Gosh, what happens when an Anderson runs against an Anderson?

    Does the “Sven” beat the “John”

    Inquiring minds want to know.

  • Bob Collins

    Do you think more people would’ve voted in the Senate race if Obama had campaigned in Minnesota and stuck Franken on the podium?

  • MR


    I think that it’s not terribly likely. It seemed that Franken was only helped minimally by the “Obama bounce,” such as it was. Enough people had predetermined opinions about Franken, especially Democrats, that I think Obama wouldn’t have helped that much. But it might have gotten Franken another 750 votes…

  • I found the missing Senate Votes!

    Exit polls suggest that 45% (National Election Pool) to 75% (CNN) of Barkley’s voters would not have voted at all on the Senate line had Barkley not been there (Bob- you would have to be evil to disenfranchise these people).

    Your missing voters….are well…..missing. They chose not to vote on the line. It seems the $50 million dollars of mud finally destroyed part of the electorage. Thank god for Dean….or the voter participation rate in Minnesota could actually go down (btw- without the IP, you can kiss that highest in the nation goodbye).

    BTW_ the IP base numbers a way up, from 6% to just shy of 11%. Next election is the first in 12 years without any statewide federal race and all that ugly money…thank god!

    Peter T (again)

  • Bob Collins

    Unrelated to the original post but on the subject of the IP. The story of the election for the IP is that an IP candidate didn’t lose support in the last weeks of the campaign. He didn’t really gain it either, but he didn’t lose it.

    That to me is the most significant change in an election that had very little of it to offer.

  • Eric (@MSPDude)

    Something appears to be odd with the 2000 election spreadsheet you linked to on the SOS site. If you add each party’s votes for the presidential candidates, you get more then the “Ballots” column. It may just be that I don’t understand the layout.

    Instead, I looked at the SOS’s overall 2000 results for the President and Senate races.

    2000 President: 2,438,685

    2000 Senate: 2,419,520

    2000 difference: 19,165 (0.79% of presidential ballots)

    Doing the same for the 2008 results from the SOS page:

    2008 President: 2,910,312

    2008 Senate: 2,885,357

    2008 difference: 24,955 (0.86% of presidential ballots)

    So, if I did my cutting/pasting/functions correctly, down ballot drop-off is just slightly more than in 2000. It doesn’t seem to be enough higher to raise flags for me.


  • Bob Collins

    Great job, Eric. I was talking to Mary Lucia tonight and she said she saw a few people while she was voting come in, put the circle in the “president” race and leave.

    That’s ironic to me because part of Barack Obama’s call has been for people to care more and do more and volunteer more and it strikes me that that would include giving enough of a rip that — since you’re already voting — spent another 30 seconds making a difference.

    That 20,000 or so people didn’t do that tells me the totality of the Obama message that brought them to the polls, didn’t fully sink in. Of course, not all of the one-race votees may have been Obama supporters.

  • ML

    Yep, I voted for the future President, skipped the Senate and then filled out the rest of ballot. I thought about this a lot. I didn’t want to vote Coleman back in, but I couldn’t bring myself to vote for Franken, even as “the lesser of two evils.” I thought for awhile I could vote Independent. But after the final debate, I couldn’t vote for him either. It’s the first time in my voting life I’ve left a race blank.

  • Bob Collins

    Funny you should mention your dilemma, I was talking to the boss tonight and told him that I find that the older I get, the less willing I am to hold my nose when voting. I wonder if this is shared by anyone else?

  • Bob is going to want to kill me but I am going to bust another myth (see posting on Anderson btw–everything is cool Bob).

    You are correct that the IP candidates did not dade this year. The observation that they usually do, however, is not correct.

    The theory that 3rd Party candidates fade in the stretch comes from “Third Parties in America” which does some pretty poor research on only seven presidential races dating from the 1920s. I remember North Dakota Unionist candidate Lemke was mixed in with Anderson, Perot, Wallace…x…x…

    That book has some good stuff in it, but the fade conclusion is pretty weak. That was brough up at a Toronto conference last year.

    I did some analysis of IP candidates since the early 1990’s using Rosenstone’s methodology that led him to conclude the fade (I don’t like the methodology because it does not account for MOE…but hey….he used it). The results are mixed, some campaigns go up, some go down, and some remain the same…pretty random actually.

    If you ever get into social science research you will quickly discover that the greates bias is not ideological but intead result driven…there must always be results…you can’t answer I don’t know.

    There was a conference at the Kennedy School a couple years past which talked about 3rd parties. It was noted that the Third Parties in America was not very robust.

    Want to learn some really cool stuff about third party performance, fades, and especially strategic voting….google Michael Alvarez (Cal Tech), Fred Boehmke (Iowa) …good stuff.

    For the record I have asked the U of M to research strategic voting. I happen to know that Jeane Massey from Fairvote asked pollsters this year to look at and test the issue.

    They don’t want to.


  • bsimon

    “I was talking to the boss tonight and told him that I find that the older I get, the less willing I am to hold my nose when voting. I wonder if this is shared by anyone else?”

    Generally I try not to hold my nose, but made an exception in this case, for Barkley. I wrote-in Tammy Lee for MN-5, rather than leave it blank.

  • Bob Collins: “Do you think more people would’ve voted in the Senate race if Obama had campaigned in Minnesota and stuck Franken on the podium?”

    Yep. If Barack Obama dropped in and did an event at the sports center in Blaine (for example) with Tinklenberg, Madia and Franken, it would have been game over.

  • Here’s the latest on the drop-off from President to Senate to the sum of all the US House races:

    –Race –Total Drop-off

    President 2910109

    Senate 2885125 24984

    House 2800918 84207

    Hope that formats.