Official: Lost Ballots Could Be Restored In Final Count, Giving Boost To Franken
So what exactly happens if that missing envelope from Minneapolis, which is believed to contain 133 votes that heavily favored Al Franken, is simply never found? Could the recount just be certified without it?
A state official just confirmed to Election Central that the state does have an option: Including the Election Night total instead for this particular precinct.
This could be big for Franken: If the state avails itself of this option to give the candidates their Election Night vote numbers instead of the recounted ones in this precinct, Franken would avoid losing a net 46 votes if they were disregarded -- a scenario that otherwise could potentially give the election to Coleman.
Beth Fraser, the director of governmental affairs for the Secretary of State's office, said that there's actually a recent precedent for such a move, going back to a 2002 recount of a state legislative race that also had ballots missing from the recounted data set. "In that case, the state canvassing board gave each of the candidates the higher number of votes," said Fraser.
But would such a decision, if it made the difference between a Franken win and a Coleman win, lead to a lawsuit? "Probably," said Fraser. "I think that most people would be surprised if there were not a legal challenge in this race."
Late Update: Meanwhile, Sam Stein has confirmed that while city officials were looking for this particular missing envelope, they found another envelope containing votes that weren't counted on Election Night, though the exact number of votes and the margin are not yet known. Things are suddenly looking better for Al.
Late Late Update: City spokesperson Matt Laible just informed Election Central that the new envelope contains about 10 absentee ballots, which are not being tallied for now.
Thank you for staying on top of this, Eric.
I have to say, though, that I'm glad I wasn't aware of TPM in December 2000, because this degree of detail about the recount would have driven me insane or into deep despair, or both.
He's up. He's down. They're missing. They're not. What a roller coaster.
December 5, 2008 12:46 PM | Reply | Permalink
You're not even kidding. And they just found (late update, above) another envelope full of votes. To me, weirdly, this is great fun. But how's this to Minnesotans? Probably pretty excruciating.
December 5, 2008 1:14 PM | Reply | Permalink
So they found a missing sock...it's not "the" missing sock, and doesn't quite match, but in a pinch it will do...
December 5, 2008 2:22 PM | Reply | Permalink
Heh heh (sorry about really late reply), but this sock's much smaller. Kind of like a baby sock, I guess.
December 7, 2008 10:48 AM | Reply | Permalink
Who knew MN, of all places, was such a den of intrigue?
December 5, 2008 12:48 PM | Reply | Permalink
C'mon...you know Minneapolis was the site of the Senator in the mens room tapping feet in the next stall. Minneapolis has lots of intrigue.
December 5, 2008 12:54 PM | Reply | Permalink
As much as gay sex seems to pervade your thoughts maybe you should consider coming out of the closet...
I'm just sayin'
December 5, 2008 1:02 PM | Reply | Permalink
My thoughts exactly as of a few minutes ago, after reading the latest of his endless ravings on the thread about Mary Beth Maxwell!
December 5, 2008 1:13 PM | Reply | Permalink
I really wanted to join that thread (having a dual interest in it being gay and a government contractor) but I was in turns depressed that we still have to have those same sorts of conversations and elated that Ion covered all the bases so eruditely...
December 5, 2008 1:43 PM | Reply | Permalink
I bet Coleman is scrambling to find a way to lose ballots now that he heard this.
December 5, 2008 12:53 PM | Reply | Permalink
So, I remember more than a few folks saying when Bush v Gore was handed down that the opinion's claim that it should not be regarded as precendent was nonsense on stilts. How could a Supreme Court decision not be regarded as precedent (or at least as persuasive authority so weighty as to be, in effect, dispositive) by future judges in analogous circumstances? I comforted myself, however, by thinking "how likely could it be that an analogous circumstance might ever arise again?". Well, it sure looks like we are about to have an answer to that (once rhetorical) question.
December 5, 2008 12:55 PM | Reply | Permalink
You missed the fine print: Bush v. Gore should only be cited as precedent in cases in which it's holdings will support a Republican's claim.
December 5, 2008 2:19 PM | Reply | Permalink
Jeezis, this race has turned out to be the most bizarre story of the year!
December 5, 2008 1:10 PM | Reply | Permalink
At the rate things are going it might end up being a contender for the most bizarre story of next year, too.
December 5, 2008 1:17 PM | Reply | Permalink
My polling place in suburban Chicago has three precincts. Overall we had 1613 registered voters in those precincts. Over 80% turned out to vote for the 11/4 election (1294). Obviously a bigger number than usual with the local guy at the top of the ticket, Durbin up for re-election to the senate, a competitive House race and other tight local races on the ballot. I can't imagine what would happen if our judges "lost" 133 of the ballots.
That's some funky chain of custody they have going on in that ward up there.
December 5, 2008 1:15 PM | Reply | Permalink
Wow, I have about zero faith that my vote is ever counted. For an activity so central to the very core of democracy, we really suck at it.
December 5, 2008 1:20 PM | Reply | Permalink
This is a completely bizarre twist, but there are only really three possible outcomes:
1) Coleman wins by more than the margin of error
2) Franken wins by more than the margin of error
3) The race is declared void and re-run.
If the missing votes make the difference the only viable options are to use the election night totals or declare the race void and re-run it. Coleman cannot win on lost ballots.
Since neither will be seated unless and until the legal challenges are over, both candidates might well agree to a re-do or the Senate itself might order a re-do rather than have the issue in litigation for years.
Coleman would expect to win a re-run, but I suspect that Franken would come out on top because the independent party is not going to be a factor in a second go-around and it appears that he cost Franken more votes than Coleman.
Also, Coleman's histrionics may have cost him support.
December 5, 2008 1:24 PM | Reply | Permalink
What do you mean by "winning beyond the margin of error"? There is only a margin of error in a poll. There is no margin of error in a vote! By definition the margin of error is the "fudge factor" of sampling and so on - when you're estimating the likely behavior on election day.
We can all speculate till the cows come home, but we need to be clear that we're speculating. Sit tight and watch the show!
December 5, 2008 1:29 PM | Reply | Permalink
I believe hallam is referring to the standards which vary according to state laws. In Georgia, they had to have a runoff Senate race despite Chambliss winning by a bunch of votes, because the state law requires a runoff if no candidate breaks 50%. There was a 3rd party candidate in the race, and neither Chambliss nor Martin broke 50%.
However, if I understand correctly, there is no comparable standard in Minnesota, so either guy could technically win by just one vote.
December 5, 2008 1:44 PM | Reply | Permalink
Ooops - I may be wrong in multiple ways. Perhaps there was no 3rd party candidate in MN, so the 50% thing would be a moot point anyway. Plus when I reread hallam's comment I think he may mean the number of missing ballots when he says "margin of error".
December 5, 2008 1:49 PM | Reply | Permalink
Of course, that is also true in the states that require and absolute majority.
December 8, 2008 12:06 PM | Reply | Permalink
I think that, by "margin of error" he means that either of them wins by more than the missing 133 votes. Otherwise, it makes no sense.
Why do I think it possible that some GOP operative, thinking that he or she would get away with it, removed Envelope #1 (with the 133 ballots) and disposed of it -- knowing, given the precinct, that it probably favored Franken) -- I hope it turns up, but it stinks, and it sounds awfully suspicious. Sounds like the Senate may have to decide who to seat.
December 5, 2008 3:01 PM | Reply | Permalink
It's quite a bitter pill to swallow, I know, and it seems counter-intuitive, but yes, there is often a margin of error in ballot counting, even in a hand recount. I don't know how many times this has actually mattered in an election, in a hand recount it is so small that it almost never comes into play. I would guess that in the modern era you could count those times on one hand if indeed there are any at all.
(I don't know your educational or professional background, I hope that the following does not seem condescending as that is not how I mean it.) Margin or error is not a "fudge factor" and does not only relate to polling. Rather, it is a measure of the error in a statistic due to error in your sample of the population (in the broad sense of the word; here population means all the ballots). It is half of the width of the confidence interval for that statistic.
There would indeed be no margin of error if instead of a sample we simply accurately analyzed the entire population. One would hope that in a hand recount we could do that but the sad fact is that while as a percentage of the population the sample is huge, it does not encompass the whole thing. As you know, ballots may be rejected or discarded because voter intent was not clear even though it is certain that most of the time the voter did indeed have an intent, or they may be rejected due to errors on the part of someone besides the voter e.g. a voter registration worker. There are plenty of other reasons the intent of an eligible voter who submitted a ballot on which they meant to express their intention would not be counted. Heck, a bunch of ballots might be lost, perhaps around 133, for instance.
These inaccuracies make the ballots that are counted a sample of the population, albeit one VERY close to the whole thing. The fact that the sample isn't random because, hypothetically, a bunch of ballots were lost from a particular precinct, makes matters even worse for your confidence interval.
Even in light of the above, the margin of error in a hand recount of an election is so minuscule that the very idea of it tipping an election to the candidate who in the "true" population had less votes is generally laughable. But a margin of 4 votes in our ballot sample? FOUR VOTES?! When 33 times as many are missing? When so many ballots are facing legal challenges that could easily remove them from the sample that were in actuality a legitimate part of the true population?
I would bet money that at this point, statistically speaking, there is no more than a 50% chance that the candidate which won in the true population will prevail in the recount. And what's sad is that it's no one's fault and there's nothing anyone can do about it.
December 5, 2008 5:21 PM | Reply | Permalink
It is simply impossible to count 2.5 million of anything by hand perfectly. We found this out in Washington State, where I live, in 2004. (Washington uses optical scan ballots, supposedly the gold standard for recounting, as in Minnesota.) The initial (machine) count in the governors race showed challenger Dino Rossi (R) beating Atty. Gen. Christine Gregoire (D) by 261 votes, out of 2.7 million. Then there was a state-mandated machine recount, which had Rossi up by 42 votes. The hand recount (paid for by the Democrats) was completed on Dec 23, 2004, and gave the election to Gregoire by 129 votes. Sound familiar?
Naturally, there were a series of lawsuits, that in the end left the result standing, but much lingering bitterness.
Postscript: Gregoire and Rossi had a grudge rematch this year, which led to a Gregoire victory by about 6%.
See http://www.historylink.org/index.cfm?DisplayPage=output.cfm&File_Id=7336
for a complete history.
But the bottom line is that there is a margin of error when dealing with big numbers, no matter what.
December 8, 2008 11:46 PM | Reply | Permalink