Study: Undervotes should favor Franken in MN Sen recount
A new study by researchers at Dartmouth bodes well for Al Franken in the Minnesota Senate recount. In a nutshell, the researchers found that Democratic districts are more likely to experience significant undervotes.
There are two types of undervotes: Intentional and accidental.
Intentional undervotes don't matter to the recount. These are the folks who didn't even try to vote for senate.
Victory in Minnesota will probably hinge on accidental undervotes discovered during the recount. These occur when a voter tries to cast a vote but that vote doesn't get properly recorded.
As a teaching assistant, I would see accidental "undervotes" when we scored multiple choice exams with optical scanners. Sometimes, for whatever reason, the machine would fail to read a mark that a human being could clearly see.
The upshot of the Dartmouth study is that both accidental and intentional undervotes are more likely to occur in Democratic districts. There are probably a number of reasons why Democrats undervote more. For one thing, Democratic voters are more likely to be poor, less educated, and/or new to voting. Perhaps more importantly, poor and minority neighborhoods are more likely to be stuck with old, clunky voting machinery that doesn't work very well.
So, if we assume that undervoters vote like their neighbors, we can expect more of these ballots to break for Franken.
While there is a significant Democratic voting bloc that tends to be more uneducated as compared to the entire population, most college-educated people tend to vote for Democrats.
Posted by: JD | November 15, 2008 at 04:02 PM
Al Franken needs to pay lawyers.
If you want to help with the expenses, then donate here.
Posted by: Eric Jaffa | November 15, 2008 at 04:04 PM
JD, you're wrong. Obama's performance among voters worsened as their education background got better, up to and including college graduates. He only did well among people with graduate education.
Lindsay, Minnesota uses new optical scan machines everywhere. The problem of different voting machines was so prominent in 2000 that subsequently states started solving it. In 2004 it was nonexistent.
Posted by: Alon Levy | November 15, 2008 at 07:20 PM
Alon, do you have a source on the numbers of college voters? Not saying you're wrong, but your claim contradicts just about everything I've read on the matter.
Posted by: Tyler DiPietro | November 15, 2008 at 10:09 PM
According the Exit Poll at CNN's website, Obama got at least 50% of every educational group:
No High School 63%
H.S. Graduate 52%
Some College 51%
College Graduate 50%
Postgraduate 58%
============================================
Posted by: Eric Jaffa | November 15, 2008 at 10:20 PM
Tyler, the data set I use is the one Eric quotes. Obama actually did better with high school dropouts than with people with postgraduate degrees, unlike Kerry and Gore.
Posted by: Alon Levy | November 16, 2008 at 02:42 AM
Obama did better with high school dropouts and worse with increasingly schooled voters. Jesus. One has to ask: "is our children learning?"
Posted by: cfrost | November 16, 2008 at 06:08 AM
That's not a new pattern Obama established. When looking at voting behavior by education level, it usually forms a "U" shape for Democratic candidates. The least-educated vote strongly Democratic, gradually declining as you move across the page, and then jumping up again when you reach post-grads.
Whether or not the percentage among "least-educated" is higher than the "most-educated," in any given year, the basic "U" pattern is consistent from election to election.
Posted by: Another Chris | November 16, 2008 at 09:31 PM
That's right Frosty - even republicans voted for Obama.
*****
You can always leave it to republicans to want to stop counting votes when they think they're ahead.
Posted by: mudkitty | November 17, 2008 at 07:11 AM
or like in hillsborough county (where I live among the real criminals of the world of all stripes and even know/love some of them) they 'accidentally' 'neglected' to mail page 2 of 2 for folks to vote with.
Solution: Take yourself and twenty minutes, wait in line with witnesses. Get it handed directly TO yourself. Vote then. Better yet, go elsewhere. Concentrate, access the internet for the obscure stuff that REALLY MATTERS (sand and water conservation, etc.... these folks weasel up to even more heinous duties if the wrong ones get a toenail hold)
THEN when you are sure you've done the best you can amidst the lies the scandals and the subterfuge, personally hand-deliver your vote again. Or have a trusted person take all of the ones for your family.
Just sayin ...
Having recently had a large, necessary deposit stolen from the mail --- if they want it out of the mail they'll take it right out ---- I don't mind that less than an hour it took to be sure caribou barbie and the mav. end up on dancing with the hasbeens. A new show I hope to produce.
Go ahead steal the idea. I won't mind.
Posted by: voxy | November 17, 2008 at 05:26 PM