I am among the strongest advocates of polling, but, I'm dismayed by shallow headlines that lead what might actually be pretty interesting work. So, in line with the "restoring sanity" theme for the weekend, I thought I'd try to bring some perspective on how polls can offer much more "reality" than "headlines" if you simply look at the data.
The latest AP-KN Poll headlined with, "Obama Primary Challenge? Nearly Half Of Dems Want 2010 Fight." Yahoo.com ran the headline, "Poll: Democrats divided over Obama as nominee in 2012." I won't go too much into the self-selection issues related to the sample; these data are not a representative sample of 2010 Democrats, they are members of a recruited panel from 2008 that have agreed to participate in repeated surveys via the internet. So, these are not "of Dems" they are "of Dems in the panel," and, the 6.5% margin of error for Democrats in the study is probably much larger than reported. It's important to keep this in mind.
But, I digress.
The poll's lead-in came from numbers showing 47% of the self-reported Democrats would like to see Obama face a serious challenge in the primaries (51% would not). However, all of these Democrats didn't vote in 2008. Among those who did, about 5% points fewer (42%) want to see a serious challenge (58% would not). Even more interesting, among the Democrats who specifically "voted for Obama" in 2008, 11% points fewer (36%) want to see a serious challenge to him (64% would not). This final number showing one in three of Obama's past Democratic supporters want to see him challenged, is actually more telling than the headlined 47%...even if significantly more, 64% in total, of these supporters do not want to see any challenge (and don't forget about that margin of error issue).
Of course, the AP-KN poll story was focusing on the potential for a Democratic primary challenge, but it's virtually political suicide for a party to push for a serious challenger against an incumbent president because it signals low party unity to voters. While anything is possible, this is an highly unlikely scenario; especially given Obama's ability to raise money, stump on the campaign trail, and his skill speaking to tough issues. Regardless of whether one agrees with the President's policy positions, it's arguable that no other contemporary political figure in the country has the breadth, depth, and intellect to speak to, and explain the issues on America's front burner (and please don't channel Bill Clinton, this is not the 1990s).
So, let's focus on the big picture, and assume the President's base is his coalition of supporters from 2008. It's important to remember that Obama won Independents and made strong inroads among Republicans in 2008. IF he's facing a serious uphill battle over the next two years, THEN we should see a solid erosion of this coalition of supporters.
What do these Democrat, Republican, and Independent (plus others) Obama voters who are members of the AP-KN panel think? Actually, they overwhelmingly still believe in the President and appear to be widely engaged:
After less than two years of highly contentious policy battles on health care, wall street reform, recovery spending, and military strategy, Gulf oil spills, Pirate attacks, criticisms from former Vice President Chaney, squabbling within his own party (see Ben Nelson), and a series of slip ups regarding police acting stupidly, and raising questions about the "wisdom" of a Mulsim community center near ground zero, these numbers actually signal a teflon president, as much, if not more than than they do battered one. But, I guess, some just see the glass half empty.
I'm not saying the headlines presented by AP-KN are wrong, I simply want to point out that there's much more to polling data than normative headlines, and I would encourage the public to click on the links that say "full results" so they too can become more sophisticated poll readers. Don't disregard, ignore, or fear the data; embrace that polls are here to stay and become critical, rather than passive, readers.
I enthusiastically applaud AP-KN for providing in-depth topline information detailing their numbers and their methods. This makes it much easier to find the interesting data nuggets and test whether or not many of the narratives we see in the media are true.
But, I think the AP-KN analysts have misread some of the tea leaves in their data. In less than two years the President has lost some Democratic support, but not anywhere near a majority of it. And, the data are hardly "glum" as the story empahsizes. In fact, the numbers are pretty consistent with what we normally see during mid-terms, of 1st term presidents; a point that the AP-KN press release eventually acknowledges.
Understandably, all survey organizations want people to "please, look at my poll" but we should all work a little harder on our headlines.
Let's try to restore some sanity, even though times are tough.
Follow David C. Wilson on Twitter: www.twitter.com/dcwilsonphd