Wednesday, October 03, 2012

Wednesday Wonderings: Confirmation bias and inability to prove a negative

Scott Lemieux has an article in the American Prospect about "Dems and Reproductive Rights: BFFs", and he makes a very useful point - one that many pundits fail to make, since the narrative is far more interesting than the facts, figures, and statistics:
The conventional wisdom would suggest that supporting reproductive rights is, on net, a national net negative for the Democrats. I see little basis for this belief. The consistent two-for-one public support for Roe v. Wade certainly puts the burden of proof on those who argue that abortion is a net loser for the Democrats in presidential politics. Moreover, contrary to the assumption of many pundits, affluent, educated voters (who are relatively more socially liberal) generally place a higher priority on social issues than working-class voters do. To give another example, claims that same-sex marriage swung the 2004 election for George W. Bush turned out to be empirically wrong. Again, too many pundits focused on the opponents of same-sex marriage that referenda might have mobilized, while ignoring the affluent suburban voters who were alienated by Republican demagoguery on the issue.
To me this is another example - of many - that seem to show that people get all caught up with narrative, and the (conveniently?) forget that there are actual measurables out there that would break their confirmation bias to smithereens. However, so long as they don't look at them, or so long as the narrative continues, there is no need to actually face the chance that their biases are wrong.

What makes this comment different from the right-wingers and GOP-watchers who are currently saying that the polls are all biased against their candidate, Mitt Romney? Well, I'd like to think that their position is consistent with the hypothesis of confirmation bias - they believe that their candidate is better than the other candidate and any attempt to show an alternate reality is actually a overt campaign of attacking their reasonable and wonderful candidate. Therefore, when polls were showing a widening gap against Mr. Romney, it couldn't be that their wonderful candidate was falling behind in the race to a socialist/communist/muslim/atheist/Kenyan not-president who has to read everything off a teleprompter and had everything handed to him due to his race. No: it has to be due to their candidate getting short shrift from "the lame stream, super liberal media". I mean, it had to be true, since all the media outlets kept talking about it, and all the major media outlets had long been labelled as "liberal". The process has gone so far that many on the right have taken to the notion that all the polls have a consistent liberal bias; every single one of them (except for Rasmussen), and there's even been a website created that has "unSkewed" the polls (by skewing the results in favor of Romney by mis-understanding and mis-using the methodology of the Rasmussen polls. (Unfortunately, even "Unskewed Polls" is now showing Obama leading Romney by 4%... so does this mean that this wonderful site is actually ... skewed?)

In comparison, I'd like to think that I can change my opinion when presented with facts of how things are actually happening or how they actually turned out. To that end, I think that this "political wisdom" that Democrats shouldn't endorse so many topics of the bygone culture wars - gay marriage, abortion rights, religious tolerance, teaching science in public schools, gun control, etc. - is just wrong, and built on a false narrative of a sweeping campaign that swept Reagan into two presidential terms. It may have been something that people stood for 30 years ago (which I doubt; people can vote for a President for reasons other than wanting to ban gay marriage and abortion), but times, they are a-changin'.

Oh: and make sure to watch the debates tonight!

No comments: