Yes, Undecided Voters Are Partisans and Have Opinions

by John Sides on October 17, 2012 · 11 comments

in Campaigns and elections

Several of the questions from last night’s debate led commentators to joke about whether the participants were truly “undecided voters,” as they were billed.  Jonathan Chait, for one, says that “Obama enjoyed friendly questions from an audience that obviously leaned left.”

It’s important to remember that undecided voters—at least as best we can identify them—are mostly composed of partisans.  Lynn Vavreck and Larry Bartels laid this out almost three months ago:

For example, despite the marked overrepresentation of independents among undecided voters, most undecided voters are not independents. The accompanying figure shows the distribution of party identification among our 592 undecided voters, as recorded in a C.C.A.P. baseline survey conducted with the same people in late 2011. Only three in ten were “pure” independents (those who denied leaning toward either party), while another 7 percent said they were not sure about their party identification. Four in ten were Democratic identifiers or leaners, while the remaining 23 percent  were Republican identifiers or leaners.

Note that the undecided voters in their sample tended to lean more Democratic than Republican, which may also have been true for last night’s participants.

Go read the rest of their post and you’ll see that undecided voters also have opinions on important political issues—ones that might, say, factor into questions they would ask of candidates, about assault weapons or equal pay for women or Benghazi or whatever.

One of the stupider things that people say about undecided voters is that they’re stupid.  But although they may follow politics less closely than other voters, they’re not somehow devoid of values, beliefs, and attitudes that bear directly on politics.  So we shouldn’t be surprised that last night’s questions were something more than anodyne, and may have reflected honest and meaningful opinions.

{ 11 comments }

RobC October 17, 2012 at 11:38 am

Note also that it was Candy Crowley who chose among the 80 questions submitted by the audience which would be asked. It’s quite possible, especially in light of other respects in which she moderated the debate, that any pro-Obama bias in the questions asked arose from her selection rather than from the universe of questions they submitted.

Nick October 17, 2012 at 2:13 pm

I’m curious, what would be an example of a pro-Romney question?

RobC October 17, 2012 at 3:06 pm

Here are two examples of pro-Romney questions–or if not pro-Romney, then at least questions that appear to have been intended to put President Obama on the spot:

Your energy secretary, Steven Chu, has now been on record three times stating it’s not policy of his department to help lower gas prices. Do you agree with Secretary Chu that this is not the job of the Energy Department?

We were sitting around, talking about Libya, and we were reading and became aware of reports that the State Department refused extra security for our embassy in Benghazi, Libya, prior to the attacks that killed four Americans. Who was it that denied enhanced security and why?

For examples of many pro-Obama questions, see the transcript.

Keith Gaby October 17, 2012 at 1:19 pm

I agree that “undecided” voters are hardly without opinions about the candidates. In fact, because of that, I think the outcome of the election is probably already determined — even if we have to wait for that decision to reveal itself on Election Night. Voters who still consider themselves uncommitted are probably leaning toward one candidate, even if they are not fully satisfied with that choice. Barring an “October Surprise,” if you pushed a swing voter to make a snap decision today, it’s probably the same one they’ll make after another three weeks of consideration. There are very few true-in-their-heart undecideds left – probably not enough to swing the race either way.

If you’re interested, the fuller argument is at: http://www.thebiggerhammermovie.com/blog/2012/10/race-has-been-decided

Ralph October 17, 2012 at 9:58 pm

I believe that this time around that there are a significant number of undecided who want ‘none of the above.’ These range from backers of Ron Paul to people who don’t consider Obama liberal enough to people who have bigoted reasons for not choosing one or the other and so on.

Before people get on a rant about socialist Obama, I remind them to look at his policies and look at Nixon’s. He is slightly to the right of Nixon on a lot of things and Eisenhower on most things. Before you get your bloomers in a knot, there are probably of the ‘undecided’ who can’t vote for Romney because he’s Mormon as there are those who can’t vote for Obama because of his skin color. The Ron Paul voters believe they got a raw deal (and the did) and like Paul, they can’t endorse Romney. But, they can’t vote for Obama either.

I don’t know how many people have told me they’re both the same or they both lie too much or they both …. In other words, “none of the above.”

Brian October 18, 2012 at 4:01 am

Going only on assumption based on my fading political science studies, this post, and the NY Times opinion piece, it appears that the “undecided” lean towards Democrats belies the long known problem Republicans face in changing demographics. Bush and fellow Republicans through much of the early 2000′s successfully made the case to much of the nation that you couldn’t be a Christian and a Democrat. Similarly, they made the term “liberal” a four-letter word. The wedge masked much of the Republicans fading numbers and the scarlet letter may have forced weaker Democrats, especially in red and purple states to take on the “independent” label. One individually or the two combined could be causality for the “undecided” Democrat lean.

Barry October 18, 2012 at 9:29 am

“Who was it that denied enhanced security and why?”

Republicans in the House, as a matter of fact.

Barry October 18, 2012 at 9:31 am

Sorry, that was to Rob C; I blame the iOS.

JPL October 18, 2012 at 7:38 am

Sorry for this ill- considered question, but I haven’t time right now for research. But if the number of undecided voters at this point is so small and the debates don’t have a large effect on the poll numbers (IIRC the prediction on this blog was one or two points after the first debate), how do we account for the seemingly large swing in Nate Silver’s forecast numbers since that first debate? What underlying reality in the electorate do these differences reflect? What are people doing? Does this change reflect people changing their mind in reaction to the debate? This is puzzling to me, since, from what I saw of the debate, in my opinion such a change from Obama to Romney in reaction to it would be unwarranted; but then I was attending to the content of what was said rather than to the spectacle of the performance, and also I don’t think I am, in general, a participant in any processes of mass mimesis.

Keith Gaby October 18, 2012 at 2:22 pm

My sense is that the first debate accelerated the movement of Romney-leaning undecideds towards him — people who would have ended up voting for him anyway. So we got fast-forwarded from Oct 4 to Oct 31 in one night.

NoahB October 18, 2012 at 10:50 pm

Nate had a post about this in the last day or so. He suggests that the swing could be accounted for by the fact that partisans are more likely to answer the poll when they’re feeling stoked. Poll response rates are miniscule anyway, so a small effect in that regard could well cause a swing of a few points, he suggested.

Here’s the post:

http://fivethirtyeight.blogs.nytimes.com/2012/10/17/oct-16-can-polls-exaggerate-bounces/

He wasn’t saying this was absolutely what was happening; just that it seemed like a reasonable or plausible guess.

Comments on this entry are closed.

Previous post:

Next post: