The polls themselves. Not the ones you should go to on November 4, but the ones you see on almost a daily basis (or, if you're as much of a political junkie as I am, that you inflict on yourself via Web-surfing on an almost hourly basis). Party polls. Media polls. Pollster polls. And, as I've followed them over the past several months, I've noticed a discrepancy between the numbers in all of these polls, and the narrative for this election cycle that has been pushed in the media, especially by "traditional" media outlets (print and broadcast).
That narrative, of course, is that this is a year of destiny for the Republicans, especially in the Senate races. You've heard the basics by now many times. Most of the seats up this year are in red states. Obama's popularity is in the tank. The voters registered with the party controlling the White House don't show up for midterms and, on top of that, there's a "six-year-itch" in two-term Presidencies that drives unaffiliated voters to vote for the "out" party in that circumstance. Historically, all true. And, if history was inclined to repeat itself, then surely the poll numbers would reflect that.
Except that they aren't.
A little bit of context. By this time four years ago, the polls were producing across-the-board numbers that pointed to a major GOP wave. You have only to look back at those numbers to see that. And, thanks to the data available on the Real Clear Politics Web site, we can do just that. By this time four years ago, the RCP average across polls for party preference in Congressional elections favored Republicans by 6.5%. Today, however, they only have a 3.0 lead. A statistical tie, for all practical purposes.
And even that is inflated by off-the-chart Republican leads from polls produced by three "old media" outlets: CBS, ABC News/Washington Post, and Associated Press/GfK. Take those out, and the difference between the parties is zero. And, frankly, there's every reason in the world to take them out. In the digital age, old media have lost almost all but their oldest viewers, the ones that more typically vote Republican. The other pollsters, while in some cases biased by the source (e.g., Fox, Rasmussen), without exception show no more than a one or two percent lead for either party. There's little doubt that they are trying to achieve a true demographic balance in building their samples.
If that's true, it shatters one more myth about midterm elections: that they are dominated by older voters. The absence of a true Republican "wave" in this year's polls suggest that 2014 might not run true to that form. There's yet another reason to think that. Namely, that the pollsters themselves are fearful that they might be undercounting younger, Democratic-leaning voters. Both The Huffington Post Pollster and The New York Times have recently raised this as a possibility. Again, this is a problem related to the demographics of communications; younger voters are harder to reach by telephone than older voters, so pollsters always run the danger of oversampling the latter. The problem, however, is that, at present, there's no statistically valid way to "correct" for that.
What does all this ultimately mean for November 4th? Jump ball, in my opinion, with each party's ground game being the key to success. This election certainly will test whether Obama and his organizers really have found a way to revolutionize getting progressive voters to the polls.
You, of course, can do your part to help prove that they have done it. Vote. Not because the polls tell you to do it, but because your values make you do it.
No comments:
Post a Comment