There is only one outstanding Senatorial recount from the election- that of Franken vs. Coleman in Minnesota. That's right, Al Franken, the humorist turned satirist turned anti-Limbaugh, turned "politician." I always sort of expected that his candidacy would in the end turn out to be the highest form of Satire, like a more-plausible Borat sneaking into the Congress and giving an address on the virtues of wifebeating. (I kidded my virulently anti-Democrat grandfather that he could do stand-up routines for Filibusters) I don't think that's the case though. Either he always was a serious contender, or he confused himself along the way. In the end, it doesn't matter, as he doesn't get to decide if he is elected. It turns out that maybe no one will.
To digress briefly, we need to discuss statistics a little bit. In any set of data there will be uncertainty about the values contained therein, from natural noise, human error, or some systematic bias that alters the data due to an unseen cause. The data are "known" to a point, but comparisons between individual or groups of data cannot be relied upon unless the difference is substantially greater than the overall error in the system. That is, you may see (or think you see) a difference, but you can't prove that it is real unless it exceeds the natural inherent error. Remember those "margins of error" listed in tiny letters below all poll findings?
This is the greatest difficulty with our electoral system, as many elections are not decided by a statistically significant margin. That is, the results are not dependable. Whoever comes out "ahead" was pronounced the winner by error, not citizens. Whether it is random error (favoring no one), or a bias error (favoring one group or candidate) then makes a big difference, and a lot of the coverage of recounts focuses on bias error (e.g., high disqualification rates in poor areas). Most people will believe that close elections have been decided, mostly because they don't know what they are talking about. It turns out that in Minnesota, they have the most transparent appeal for just this situation.
According to an excellent op-ed in the NYT (that agrees with this analysis and inspired some of it), if an election in Minnesota results in a statistical tie, the election can then be decided by drawing lots. That is a coin toss. A true, unbiased, totally inhuman means of determining the winner. Currently, around 50 votes out of 3,000,000 cast separates the Minnesota senatorial candidates. This is far, far less than the 1% human error expected in a recount, a hopeless situation from a statistical perspective.
I find it laudable that the Minnesota constitution provides a means to choose the winner of a statistical tie by blind luck rather than dumb error. However, I do find it ironic that a satirist who may have entered into a race with a wink and a shrug, then ground his fingertips bloody in the catfight to win, will in the end have the race decided with a coin toss and a shrug. Al Franken, I'm sorry: that's just not funny, but it is ironic.
Here's a link to the op-ed. There was a similar one during the 2006 election too.
Posts
No comments:
Post a Comment