National Geographic

Seeing an American flag can shift voters towards Republicanism

[WARNING: A THOROUGH REPLICATION ATTEMPT BY MANY LABS FAILED TO REPRODUCE THE RESULTS IN THIS STUDY. THE EFFECT DESCRIBED THIS POST IS HIGHLY UNLIKELY TO BE REAL.]

As a visitor to the USA, one sometimes gets the feeling that it’s hard to move or look around without seeing a flag. They are seemingly everywhere, an omnipresent reminder of national identity. But the star-spangled banner is more than a symbol; it can also influence minds in unexpected ways. Travis Carter from the University of Chicago has found that when people think about voting decisions, the mere sight of the American flag can subtly shift their political views… towards Republicanism.  It’s an effect that holds in both Democrats and Republicans, it affects actual votes, and it lasts for at least 8 months.

In the run-up to the 2008 US presidential election, Carter recruited a group of around 200 volunteers and asked them about their political views. A month or so later, he split them into two groups that were comparable in terms of their political beliefs, voting intentions and other variables. Both groups rated how likely they were to vote for either the Democrat Barack Obama or the Republican John McCain on an online questionnaire. The questionnaires were identical except for one small detail – in the top left corner of the screen, one group saw a small American flag and the other saw nothing.

That tiny difference was enough to swing their voting preferences. Carter found that the volunteers who saw the tiny flag became more likely to vote for McCain than Obama (relative to their answers at the start of the experiment). They claimed to feel more positive towards the Republicans and even when Carter tested their unconscious atittudes, a small Republican bias still came through.

After the election, Carter contacted the volunteers again and asked them who they actually voted for. He found that those who saw the flag were less likely to have voted for Obama than those who didn’t (73% versus 84%). They were also more likely to think that the media were unduly harsh in their treatment of McCain. Remember that there were no differences in the political leanings of the two groups before one of them saw the flag-bearing questionnaire.

Finally, in July 2009, Carter caught up with his volunteers one last time. Even though eight months had passed since half of them saw the tiny flags on-screen, these recruits still showed some Republican bias. They were less happy about Obama’s job performance than their peers, less warm about other liberal leaders, and even held slightly more conservative views. (Bear in mind that in this final round, only a third of the original sample answered Carter’s call; however, both the flag and no-flag groups were equally represented).

The effect of Carter’s simple questionnaire is stark in both its size and duration He writes, “A single exposure to an unobtrusive American flag shifted participants’ voting intentions, voting behaviour, attitudes, and beliefs toward the Republican end of the ideological spectrum.“ This was true whether the volunteers identified as liberal or conservative – people from both ends of the spectrum shifted towards Republicanism.

This isn’t the first time that a national flag has provoked such a striking effect in a psychological study. In 2007, Ran Hassin (who led Carter’s study) found that the sight of an Israeli flag could affect the attitudes of people involved in the Israeli-Palestine conflict. The flag appeared too briefly to be consciously seen, but still it drove the volunteers towards a more moderate stance in the political centre. And the brief flash of flag even shifted the volunteers’ votes.

But there is one important difference between the two studies: the Israeli flag pushed people towards the political centre, but the US one shifted people to the right. Why?

Perhaps the volunteers moved towards the dominant party at the time? Carter thinks not. In the spring of 2010, with Obama a year in power, Carter recruited 70 people and asked them to look at four photographs. Half the people saw buildings with flags in front of them; the others saw photos where the flags had been digitally removed. Even though the two groups had the same spectrum of political beliefs beforehand, the flag group shifted towards a Republican worldview after seeing the photos. It doesn’t seem to matter who is sitting in the White House at the time.

Instead, Carter suggests three alternative explanations. First, it’s possible that the flag does shift people to a more moderate position. Carter’s recruits tended to be more liberal than conservative, so if they all moved towards the political centre, that would come across as a shift to the Republican end. The fact that conservative volunteers shifted further to the right argues against this, but it would be simple enough to test by repeating the study with a group of predominantly Republican volunteers.

Second, people might associate the American flag with Republicans more than Democrats. Carter demonstrated as much in a small pilot study of 50 people – they associated brandishing the flag with Republicans more than Democrats. And indeed, previous studies have found that conservative Americans are more like to own or display a flag than liberals are. Carter writes, “The American flag conjures up Republican beliefs and attitudes, and these primes collectively push people in the Republican direction.”

Third, people might simply believe that the average American is more conservative than they are. Carter argues that people associate national flags with the archetypal citizen, and if they see a flag, they might shift their attitudes towards that imaginary every-American.

All three possibilities can be tested in future studies. For now, one thing is clear: these results come as a shock to most people. Indeed, Carter found that 90% of people believe that the presence of a flag wouldn’t affect their voting behaviour.

We like to think that their political beliefs and choices are the result of thoughtful consideration and objective analysis. In truth, several studies have now shown that voting simply isn’t that rational. Our choices are affected by unconscious preferences, our reflexes, and even local sports results. We are so predictable that people can guess the victors of elections with a surprising degree of accuracy based only on fleeting glances. In this context, the idea that a powerful national symbol like a flag could affect political preferences is not unreasonable.

It does, however, seem unbelievable that one exposure to an innocuous flag could have such broad effects, especially since the recruits will have seen hundreds of flags in their daily lives. Carter acknowledges this incredulity. “Considering how often Americans are exposed to their flag, why would this one exposure have any impact at all?” he writes.

He thinks that the answer lies in the context of the experiment. During his study, people saw the flag while explicitly declaring their voting intentions. That’s a very powerful act, and not one that people do very regularly. Carter says, “For some participants, explicitly declaring voting intentions may have been a rare event that further crystallized their stated intentions and attitudes, incorporating any bias introduced by the presence of the flag at that critical moment.”

Reference: Carter, Ferguson & Hassin. 2011. A Single Exposure to the American Flag Shifts Support Toward Republicanism up to 8 Months Later. Psychological Science http://dx.doi.org/10.1177/0956797611414726

More on politics:

There are 18 Comments. Add Yours.

  1. Coturnix
    July 10, 2011

    Fascinating. Someone should try to repeat this with an image of a cross wrapped in the American flag ;-)

  2. Ed Yong
    July 11, 2011

    I know you’re semi-joking, but it *is* interesting that the Israeli flag has a massive religious symbol smack in the middle of it.

  3. Aaron
    July 11, 2011

    I thought americans were a bit wrapped up in our flag til I visited denmark. My GOD those danes love their flag, it’s EVERYWHERE, to an extent that puts republicans on the 4th of july to shame.

  4. Taylor Burns
    July 11, 2011

    Ah, good ol’ cultural violence. The 21st century is a world of images – if only we could move beyond this mid-19th century nation-state hangover…

    Apropos: At a psychology conference in Istanbul right now, and a Canadian social psychologist just presented results from an experiment where he alternatively primed participants with an image of the earth from outer space and a national flag. In subsequent moral dilemmas, those exposed to the planetary image significantly increased in measures of empathy, egalitarianism, etc, while the moral compass of those exposed to a national flag became much more punitive. Was considering penning a blog post about the study – this seems all the more appropriate in light this similar and contemporaneous American flag study. Would make for a good comparison, once the data is actually released to the public.

  5. David Hobby
    July 11, 2011

    Ed, thank you for a great blog!

    My first thought on reading this was “it’s probably a statistical artifact”. For all I know, the author has dozens of unpublished studies with no results. His sample sizes are small, so he’ll have a lot of random variation mixed in with the effects he’s looking at.

    As for the main study, I bet the matching of the two groups wasn’t perfect. That would certainly explain why the “effect” lasted so amazingly long–the flag group actually WAS more conservative when it was chosen.

  6. Noah
    July 11, 2011

    Weird, since the flag of the Republican Party is the Stars N’ Bars…

  7. Nathan
    July 11, 2011

    Seems like there might be a fourth explanation for the results. The appearance of the flag on the questionnaire could (consciously or unconsciously) cause respondents to believe that the questionnaire itself was produced by conservatives. If the official ethos of the questionnaire seems to take on conservative overtones, I wouldn’t be surprised if people adjusted their answers in accordance with the audience they believed they were addressing.

  8. Ed Yong
    July 11, 2011

    @Nathan – Ooh I like that. The flag only appeared on that one questionnaire, but if the volunteers believed that the experimenters were more conservative than they were, that might indeed have influenced their behaviour in later stages of the experiment.

    @David – Regarding your second point, as far as I understand, the results were all adjusted for the political views expressed in the very first stage of the experiment. So they represent *shifts* in political viewpoints, rather than absolute position on the spectrum.

  9. Miller
    July 11, 2011

    @Noah Funny you should say that, because it was the Democrats who seceded from the Union and supported slavery. Every single one of those guys who ordered fire hoses turned on Civil Rights protesters were Democrats. George Wallace, the guy who stood in the school house door to prevent integration was a Democrat.
    Learn a little bit of the history of the party you support. Historically it’s been Republicans who have supported civil rights and it has been the Democrats who have opposed it.

  10. Sam McNerney
    July 11, 2011

    I am not buying the explanations so far. Here is what I think is going on. The flag is pushing, or “nudging” (as Thaler and Sunstein would say) borderline Democrats that may be doubting their party. Its like when you are having a difficult time deciding between two options, and just the slightest thing can nudge you towards one of the options. And although you say that you want one option, the truth is that you are undecided. Maybe its deciding between two sports teams to win a game…. you say you want the Heat, but really (unconsciously) you want the Mavs, and because of that, your brain unconsciously looks for cues or reasons to switch you preference.

    I doubt that the flag actually “controls” peoples decisions. From what we know about confirmation bias, true democrats would never be persuaded by an American Flag to switch votes. I think the effect is only true for borderline democrats who are unconsciously looking for reasons to switch parties. Unfortunately, I don’t think these types of democrats would show up on a political survey, which they took before the experiment….. so it would be tough to test my idea out.

  11. Owen
    July 11, 2011

    @Miller: What you’re saying was true up until the sixties. Then the Democrats decided to back the Civil Rights Act along with many Republicans. Most of the opposition came from southerners of both parties. However, since President Johnson, a Democrat, was the driving force behind the act’s passage, southern racists deserted the Democratic party in droves. The Republican party actively sought their votes in later years.

  12. Emmy
    July 12, 2011

    I was thinking along the same lines as Sam. Granted, I have not read the full study, nor any other studies related to such topics. But from what I’ve seen, this study and the one conducted by Haidt are disappointing. Because they use such vague categories. Haidt’s study did not seem to be based on registered Democrats or Republicans, but rather it used the flimsy categories of “Conservative” and “Liberal” (was there even a category for moderates?)

    This study would have been great had there been tighter categories: Democrats who have voted in their party for 5 years. Democrats who have voted in their party for 10 years. Republicans who voted in their party for 15 years. Democrats who voted for the Green party twice. Republicans who voted for Democrats 4 times. Etc. Instead we now have no idea who in the study was among the large number of undecided American voters, and which ones were hardcore left or right. Having precise numbers in studies may not give us much more to work with so far as interpretation goes, but it makes for a much clearer picture.

    I think the study is interesting though and worth doing again; although when people jot down how they plan to vote, versus actually casting a ballot, I wonder if we can take that seriously as well.

  13. djanes1
    July 13, 2011

    Does any of this have to do with McCain and Obama specifically? McCain is a decorated war hero, while Obama’s dad wasn’t American. Right or wrong, these personal subtexts could greatly influence the way the survey subjects contextualize the flag image. Would the same effect hold true, for example, a race between (say) John Kerry and Arnold Schwarzenegger?

  14. Sam McNerney
    July 14, 2011

    Ha, did anyone see that this study was mentioned on the Colbert Report last night?

  15. the_Butcher
    July 16, 2011

    The flag is to remind you where you came from. Political parties come and go, what’s that go to do with the national flag?

  16. David Hobby
    July 17, 2011

    @Ed–

    You wrote: “@David – Regarding your second point, as far as I understand, the results were all adjusted for the political views expressed in the very first stage of the experiment. So they represent *shifts* in political viewpoints, rather than absolute position on the spectrum.”

    Sorry, it’s not that simple. An ATTEMPT was made to adjust for political views expressed at the start. In this kind of research, such an attempt will necessarily not be perfect. (There are a LOT of random factors.) The result can be groups that are imperfectly matched.

    One of the conclusions of the paper is that a single flag exposure can have LONG-TERM affects on political outlook. To me, this is VERY implausible. So one looks for alternative explanations. Mine was that the researchers were unlucky, and their groups were actually different in their political views. Nathan@7 had another alternative explanation.

  17. Rusty Reardon
    July 22, 2011

    There is a term used in the field of “The Psychology of Influence”. “Social Proof” is the term used when people are influenced by repeated exposure to a consistent visual or verbal message. Marketing companies know this as well as politicians. When the bad little boy is caught red handed with his hand in the cookie jar, and he repeatedly denies your own observation, you will begin to doubt what you observed.

  18. Karen Hauber
    October 20, 2011

    I just came across this article (a few months after it was published). As owner of a 40+ year old family business featuring American flags, http://www.flagandsignplace.com, we see first hand that a large percentage of our US flag customers are veterans and families of veterans, and those of a more conservative political leaning. We have a significant drop in American flag sales in Presidential election years, which would seem the opposite of what would be expected. Dissatisfaction with government is the focus in the media and in everyday conversations for months on end in these years, and it plays out in the flying, or the not flying, of the US flag! When things are “good” in our country, approval ratings up for a current administration, the economy better, world events favoring the US, etc, we see an increase in American flag sales. Of course we also see increases in times of crisis, such as Sept 11th, 2001 or during Desert Storm and so on.

Add Your Comments

All fields required.

Related Posts