The Frontal Cortex

Motivated Reasoning

In response to my post yesterday which argued that Democrats and Republicans are both vulnerable to what’s politely referred to as “motivated reasoning” – in other words, we’re all partisan hacks – some commenters objected. They pointed out that the actual study I was discussing found that conservatives, perhaps due to their rigid beliefs, were especially vulnerable to such cognitive flaws. Here’s a sample:

Imagine any other post like this generically. “A new study found something that supports my worldview! Haha. Except part of the conclusions undermine my worldview, so anecdotally I disagree with that and quote my personal experiences to prove I’m write”. That’s incredibly un-scientific.

Anyway, gasp, I’m a liberal, whatever. I acknowledge both liberals and conservatives can be irrational, but when there’s studies showing that one group is MORE irrational, then maybe we can cut the “pox on both your houses” malarkey for just a second.

That’s a perfectly valid point, but I’m sticking with my bipartisan view of partisanship. Consider this study, by Drew Westen. (It’s worth pointing out that Westen is a devout Democrat and even wrote a book that gives Democratic politicians advice on how to reach “the emotional brain.”) During the run-up to the 2004 election, Westen showed a group of 15 Democrats and 15 Republicans multiple statements from John Kerry and George Bush that were clearly contradictory. For example, the experimental subjects would read a quote from Bush praising the service of soldiers in the Iraq war and pledging “to provide the best care for all veterans.” Then, the subjects would learn that on the same day Bush made this speech, his Administration cut medical benefits for 164,000 veterans. Or they would a statement from Bush praising Ken Lay (the CEO of Enron) and then a statement explaining that Bush now avoids any mention of Lay. Kerry, meanwhile, was quoted making contradictory statements about his vote to authorize war in Iraq and his pledge to reform Social Security.

After being exposed to the political inconsistencies of both candidates, the subjects were asked to rate the level of contradiction on a scale of one to four, with four signaling a strong level of contradiction. Not surprisingly, the reactions of voters were largely determined by their partisan allegiances. Democrats were troubled by Bush’s inconsistent statements (they typically rated them a four), but found Kerry’s contradictions much less worrisome. Republicans responded in the opposite manner. They excused Bush’s gaffes but almost always found Kerry’s statements to be flagrantly incoherent.

The study was done in an fMRI machine, but there’s no need to get into the cortical substrate of motivated reasoning. Let’s just say that brushing aside a contradiction seems to activate the rewards mechanisms of the brain. (Self-delusion, in other words, feels really good.) My point here is that Westen observed no difference between liberals and conservatives. Here is a relevant quote from the paper:

Because the focus of this report is on partisans’ responses to threatening information about their candidate (rather than on differences in neural processing between Democrats and Republicans), and because Democrats’ neural and behavioral responses to Kerry contradictions resembled Republicans’ responses to Bush contradictions, we aggregated the data across parties.


  1. #1 Anon
    September 17, 2008

    The study’s methodology introduces an artifact that undermines your conclusion. To be included in this study, subjects had to pass a particular criterion of partisanship (30 points or more on the “feelings thermometer”). They do not present any data for what percentage of initial applicants passed this test. The results thus (artificially) present extremists in both camps; you should not say that Westen “observed no difference between liberals and conservatives”, but rather “observed no difference between liberal and conservative extremists.

    It may not be relevant, but we cannot know. I, personally, would be very interested in knowing what percentage of applicants met the criteria, and whether there were meaningful differences between extreme partisans and mere leaners.

  2. #2 Coturnix
    September 17, 2008
  3. #3 Lucas
    September 17, 2008

    It’s also difficult to be confident of a result from a study with 30 subjects. One person’s opinion can change the results (for that party) by almost 7%, so it’s entirely possible that democrats might be less susceptible to contradictory thinking, but the sample was skewed by one person. I haven’t read the study, but the example texts you summarize include one example of gross hypocrisy (the veterans example) on Bush’s part which seems a bit more severe than the other examples. If I were the study’s designers, I probably would have picked a more minor example to match the others.

    However, I do find the results intuitively plausible, even on a personal level. I find myself much less influenced when I disagree with Obama than with McCain. I’m not exactly a partisan Democrat either. If foreign policy were removed from the debate, I would probably go with McCain. I think I made my decision based almost entirely on foreign policy, so perhaps I’m subconsciously reinforcing that decision when I hear them speak on other issues.

  4. #4 Nicholas
    September 17, 2008

    Your last two posts make it seem as though the only difference between the left and the right is the content of their ideology. However, there is good reason to believe that conservatives are more cogitively rigid than liberals. Jost et al.’s meta-anaylsis found that conservatism correlated positively with intolerance of ambiguity, openness to experience, and needs for order, structure, and closer, and correlated negatively with uncertainty tolerance and integrative complexity.

    Although the study was criticized for it’s title, “Politcal Conservatism as Motivated Reasoning”, the authors we by no means suggesting that liberalism is NOT motivated reasoning.

    Further, Block and Block examined the temperaments of children in nursery school and found that temperament predicted their ideology 20 years later (!!!). This goes a long way in demonstrating that the differences in cognitive style are real, and that, if anything, ideology falls in line behind our motivations for perceiving the world a particular way.

  5. #5 Pierce R. Butler
    September 17, 2008

    Regarding, say, the war on Iraq: Bush’s statements and actions were significantly consequential; Kerry’s weren’t.

    Might not that have something to do with why reality-based persons might assess one differently than the other?

  6. #6 John Daly
    September 17, 2008

    Actually, Westen did report a difference between Democrats and Republicans in his study. The Democrats were more negative about the Bush contradictions that were the Republicans about the Kerry contradictions. The Republicans were less negative about the Bush contradictions than the Democrats were about the Kerry contradictions. The Democrats were also more negative than the Republicans about the neutral contractions. Thus while the evidence supports the contention of “motivated reasoning”, the small sample of Democrats appeared to judge the small sample of inconsistent pronouncements more negatively than did the Republicans.

    Westen provides only two examples of inconsistencies. I was left with the question as to whether the inconsistencies of Kerry and Bush were really equivalent, or whether one was guilty of more serious inconsistencies. I don’t understand how one would construct an index of the seriousness of such an inconsistency. If in fact the inconsistencies attributed to one of the candidates were objectively worse than those of the other candidate, then drawing a conclusion from the study that the Democrats and Republicans were equally guilty of emotional thinking might well be wrong.

  7. #7 Ren Galskap
    September 17, 2008

    In statistics, 30 is considered a large sample, provided the sample is unbiased. You don’t get significantly better results from a sample size of 60. However, if your sample is biased, the size of your sample doesn’t matter. A biased sample of 1,000,000 is just as likely to give biased results as a biased sample of 30. In this case, as Anon points out in the first comment, the sample may be biased.

    Nevertheless, I agree that the idea that liberals are less prone to the backfire effect is dubious. The increased backfire effect among conservatives may be due to the issues selected: liberals have been portrayed as spenders for so long that the stereotype may override the facts. However, stem cell research is relatively new, and liberals don’t relentlessly promote the message that conservatives are opposed to stem cell reasearch.

  8. #8 Cerulticle
    April 24, 2009

    Acomplia Drug
    It is recommended to keep taking the drug in order not to regain weight, though weight lost in the initial stages will not be regained if the drug is continually taken.

  9. #9 Kelly
    September 14, 2009

    Weston’s differences may not be attributable directly to Democrats vs Republicans, but more attributable to personality differences and the way those personality differences self-sort into either political party.

    In addition to Nicholas’ citations above, Professor Bob Altemeyer’s work on authoritarian follower personalities ( clearly demonstrates seven significant cognitive characteristics of this personality type:
    1. Illogical Thinking,
    2. Highly Compartmentalized Minds;
    3. Double Standards;
    4. Hypocrisy;
    5. Blindness To Themselves;
    6. A Profound Ethnocentrism;
    7. Dogmatism: The Authoritarian’s Last Ditch Defense
    (more extensive descriptions are found in Altemeyer’s chapter 3).

    Although Altemeyer calls this personality type “right-winged authoritatians” he makes clear that this refers to the type rather than the political position and that these types can also be found amongst Stalin’s followers. Nevertheless, these followers appear to be more attracted to our political right and tend to concentrate on our political right. The left might attract these folks with a greater use of different influencing tactics.

    Also, although all of us are somewhat prone to these same cognitive biases, these authoritarian followers are much more prone to these biases than the rest of us. According to Altemeyer’s research, about a third of the population appear to be authoritarian followers. These folks tend to be more easily influenced by informational tactics more common (but not exclusively found) on the right.

  10. #10 Jenny
    February 28, 2010

    Yes, yes. Those *other* people have a bigger problem being blind to themselves than *we* do.

    And I was considering examining male/female differences in my research. It’s so good to be reminded of the devastation that a little bit of one-sided research can wreak.

  11. #11 John
    August 30, 2011

    From the few examplse of Comments I’ve read on this post, It seems to reinforce Motivative reasoning as a force. Most of the comments seem to be trying to make the statement that “my side is more reasonable then yours” What better motivator can their be then to protray oneself as more reasonable then others.

    I know I have been guilty of motivative reasoning, the challange is to try and recognize it, accept it for what it is and try to move beyound it.

  12. #12 John
    August 30, 2011

    Sorry about the spelling errors, I should have previewed the post before I posted it.

New comments have been disabled.