Blogs

Science Of Small Talk

The science of social behavior, one interaction at a time

by Sam Sommers

Sam Sommers, Ph.D., is a social psychologist at Tufts University. See full bio

October 14, 2008, Politics

Comments on "Who's Going to Vote for "That One"?"

Who's Going to Vote for "That One"?

By Sam Sommers on October 14, 2008 in Science Of Small Talk

If you thought we've been drowning in political opinion polls for a while now, just wait to see what the next three weeks brings as Election Day approaches. And the $64,000 question, as always, remains how accurate are these polls? This is a particularly perplexing query this year given the potentially biasing impact of the notorious Bradley Effect... Read More

Huh?

Submitted by michaelkane on October 14, 2008 - 9:09pm.

Hello,

I can't refrain from pointing out a fairly obvious hole. You state:

"A White voter has little reason to think, I have to tell this guy I plan to vote for Obama or else he'll think I'm a bigot. Especially since the next questions in the survey will probably allow the respondent to explain her candidate preference."

If you replace the subject 'white voter' with 'white, staunch Democrat,' the flaw becomes more apparent. How would one go about explaining that? "I've voted Democrat in every single presidential election since 1980, but I'm making an exception this time because, uhhh... because...ummm..."

Regardless of whether the 'effect' can be empirically proven to have influenced the outcomes of elections, the premise at least jibes with common sense: people have a painfully obvious reason to lie which does not exist in other situations. As you know, the fact that we lie about [or are unaware of] how racist or otherwise bigoted we actually are has been empirically proven several times over; I don't see what makes a poll different, or why [Democrats] would need a study to be legitimately concerned, particularly in view of the last two elections.
Will significant numbers of people 'freeze up at the moment of truth and pull the lever for McCain'? Of course not, but significant numbers of ambivalent Democrats might not actually show up, along with significant numbers of just plain lazy Democrats who support Obama every step of the way up to, but not including, actually getting off their arses and voting. And Yes, I’m self-righteous on this point: there is no possibility of me merely talking about voting in this election because I have already mailed in my absentee ballot.
I hope your conclusion is correct because I am both a Democrat and an Obama supporter [and white, incidentally]. The premises that got you to that conclusion seem a bit shaky, however. Personally, I'd prefer to see frightened Democrats with more motivation to actually cast ballots and bring people with them, and I'm deeply suspicious of any reassuring information that might contribute to the consistently abysmal Democrtatic turnout.

re: Huh?

Submitted by Sam Sommers on October 15, 2008 - 5:14am.

But your example presupposes that the white voter in question thinks that the interviewer is aware of his or her identity as someone who's voted Democratic in every election since 1980. Why would they think that? The process you describe is not one where the poll questions themselves should trigger concerns about appearing prejudice, but rather a self-focused process by which a "staunch Democrat" feels bad about not supporting Obama and therefore has to lie about it to appease themselves. So it would be that self-focused deception you'd be concerned about with the Bradley Effect, and I just don't see it happening too often here. Where was it in the primaries? Where is it in the small, but vocal numbers of "staunch Democrats" who keep telling pollsters they won't vote for Obama? Your other concerns I share-- people need to follow through on their commitment to a candidate by actually voting. And certainly the motivation for this entry was not to assure voters of any ilk that they don't need to show up on November 4. I find it hard to believe that it was the desire to disprove the "Bradley Effect" that was going to drive Democrats to the polls this fall. The objective of the post was simply to address the accuracy question regarding the current polls.

1) Do you have a party affiliation? 2) Who are you voting for?

Submitted by michaelkane on October 15, 2008 - 8:34pm.

Why would they think 'that'? You took 'that' to literally mean 'staunchness' and past voting habits, as opposed to rhetorical devices for emphasis, but my 'that' just meant party affiliation. And they would think 'that' if they had indeed been asked. If a pollster asks a Democrat's party affiliation before asking whom he or she is voting for, if he or she doesn't plan to vote for Obama and is embarrassed by either the real reasons or the prospect of being misunderstood, viola: a reason to lie, rational or not, which stems from the questions and the situation. A phone conversation with a human is not the same as a voting booth.

And why should we presuppose that someone being polled would assume they would get a chance to thoroughly explain that they have reasons which have nothing to do with race, and more to the point, would care to go into such an explanation?

The hypothetical 'staunchness,' while beside my point, does indeed do more to describe issues of self-image. However, I would maintain that such issues would also be significant motivators here. Would people lie to themselves? Sure; and if they did and then coincidentally never confronted themselves by actually voting, it could easily contribute to the 'effect'. Also, any staunch Democrat who is 'vocal' about not voting for Obama definitely has some point to make, which is a whole 'nuther can of worms. Otherwise, why advertise?

You asked where [the 'effect'] was in the primaries-- I'd say somewhere between Iowa and New Hampshire, apparently:

http://www.thenation.com/blogs/campaignmatters?pid=268328

Note the different settings [caucus vs. private voting booths].

reply

Submitted by Sam Sommers on October 16, 2008 - 4:12am.

I just used the hypothetical you used in your post. Again, throughout this election, from an anecdotal perspective I haven't noticed self-described Democrats having any trouble telling pollsters they plan to vote for someone else (or abstain altogether). Only recently do the numbers suggest they've coalesced around Obama. The Iowa and NH example is a good one-- it was in my original draft of the post but I cut it for space. At that point, in January, it seemed very possible that we were seeing a Bradley Effect in the private voting booth but not in the public caucuses. But that trend didn't hold with other contests. Your citation is to a blog from January; analyses of the entire primary season, including the Nate Silver link I include in the post, suggest no reliable Bradley Effect for the primaries taken as a whole. That's the danger of basing conclusions on cherry-picked or salient examples instead of controlled analysis.

Supporting Obama, but Voting for McCain

Submitted by Dan Rust on October 16, 2008 - 4:39pm.

See the YouTube video below:

http://www.youtube.com/watch?v=PpsdKUQ1nKA

for one person’s perspective on how you can make a decision to support Obama AND vote for John McCain. Feel free to forward this to any “undecided” voters you know of. This will be a VERY close election and we need every single vote we can get.

Demonizing Obama will get us a few votes (and also lose us at least a few votes) but I hope that a rational, positive thought process will persuade a few people who are still on the fence.

Regards,

Dan

not on topic

Submitted by Sam Sommers on October 16, 2008 - 6:14pm.

Dan, this isn't a blog for advocacy for particular candidates or causes-- it's for psychological analysis of current events and daily interactions. So please confine your efforts to promote your video to more appropriate venues. But since you opened the door for this discussion, it's difficult to maintain the status you strive for in the video as an impartial undecided voter just sharing his humble, objective opinion when your post on this site uses phrases such as "*we* need every single vote *we* can get" and "demonizing Obama will get *us* a few votes." It undermines your effort to present your video as anything other than a partisan message cloaked in the disguise of impartial advice. Your video focuses on the importance of "winning" the war for the sake of reputation and morale. This treatment of war as comparable to a sporting event that is easily defined as "won" or "lost" is a recent development that I find troubling. It's quite the oversimplification, not to mention a potential trivialization of the human cost in question. But even if one wanted to go down that road, one could argue that the U.S. "won" the military objective of the war a long time ago, and could leave at any point without fear of "losing." It's the reconstruction that has been botched.

I'd say you actually set up

Submitted by michaelkane on October 17, 2008 - 12:10am.

I'd say you actually set up a straw man on top of a misunderstanding of the example, but I've already passed the point of quibbling, so enough of that.
You essentially said you can’t imagine a reason for lying, but I can. Because I can, the theory is plausible to me [as one, not the only, factor]. I don’t give two hoots about proving or disproving it and wouldn’t attempt to do either. The polls now predict a lead, not a landslide, and that concerns me.
The candid people you reference are completely beside the point; a supposed Bradley effect would by definition exclude Democrats who had already said they wouldn't vote for Obama, as you yourself pointed out. Their existence does nothing to disprove the existence of their counterparts.
I’m not quite sure if you’re suggesting it constitutes ‘controlled analysis’ as opposed to just broader analysis, but in looking at that link you provided, I see nothing in the overall primary numbers which would dismiss the idea that this ‘effect’ could have existed and been a factor in some states with disparities between polls and results. What I do find of particular interest is which states the disparity did appear in: nearly all on the East Coast—precisely where common sense would predict an appearance of the issue of white Democrats saying one thing and doing another as opposed, to say the South, where common sense [prejudice, if you like] would suggest white Democrats would feel less conflicted. And what about states with no variance, or the opposite? The people polled and the people who voted are definitely all the same people? An ‘effect’ might have existed but been outweighed by other factors. Bluntly put, the BLACK man vs. a white WOMAN dynamic in the primaries also suggests the plausibility of a double [Bradley] effect counterbalancing itself [in a way that could not be paralleled by a vice-presidential candidacy].
This is why you’ll never find a political ‘science’ class in a science department. No one can prove or disprove much of anything in scientific terms, which arguably makes common sense as credible a tool as any other.

Bradley-Wilder effect

Submitted by votetheday.com on October 21, 2008 - 11:00am.

Is it possible, that Obama's lead could evaporate on election day because of Bradley-Wilder effect? Or nowadays Americans are significantly less reluctant to vote for an African-American? Vote here - http://www.votetheday.com/america/secret-racism-will-subvert-obamas-adva...