It’s been obvious to everyone who’s written about politics since Aristotle that emotion plays an important role in political rhetoric and political judgment. With an increased focus on emotion in cognitive scientists, there has been a flourish of empirical work attempting to elucidate this role. I thought it might be interesting to say a little about this given the recent resurfacing of the Westen et al. study on motivated reasoning in politics. While the recent posts on the study have (unfortunately) described it in terms of confirmation bias, it would be more accurate to describe the study, and motivated reasoning in general, as emotion-driven or emotion-influenced reasoning. What the Westen et al. study showed, for example, is that brain areas associated with emotion were activated when partisans were faced with information that threatened to undermine the conclusion they wanted to reach.

When the study first surfaced (before it had even been submitted for publication! it was that sexy), much of the discussion centered on how irrational the partisans in the study looked, because they relied on emotion instead of (or in addition to) reason. This perceived opposition between emotion and reason is certainly not new, and some empirical results bare it out. For example, people tend to attach emotional reactions to entire groups, and that emotion then influences their judgments about individual members of that group. When people who have a strong dislike for Democrats hear a Democratic politician speak, their negative emotions associated with the Democratic party are likely to lead to negative evaluations of the politician, her speech, and any positions she takes in the speech1. Furthermore, some emotions (e.g., anger) may even make people use less effort, and thus simpler processes, to arrive at a political judgment2. I probably don’t need to tell you how much anger opposition parties can elicit in partisans (if I do need to tell you, then go read a political blog). So emotion can decrease the efficiency with which we make political judgments.

But the relationship between emotion and reasoning, and therefore emotion and political judgment, is much more complex than the common view that emotion is detrimental to reasoning. As neuroscientists like Antonio Damasio have shown, individuals with damage in the emotional centers of their brains have a very difficult time making rational choices3, largely because they don’t have emotions to make sure they make choices that aren’t harmful. In addition, we can sometimes account for our emotional reactions to information while making decisions. For example, our goals influence the role that emotion plays in our decision making. Research on motivated reasoning (upon which the Westen et al. study was based), for example, has repeatedly shown that if we’re motivated to reach a true conclusion, as opposed to being motivated to reach a specific conclusion, we are able to overcome our emotional reactions to particular pieces of information a large extent.

As a further example of this complex relationship in politics specifically, I thought I’d describe a study on emotion and political judgment by Isbell and Wyer4, which serves as an excellent example, and should help rehabilitate the reputation of partisans a bit. Isbell and Wyer note that emotions can affect judgments, even when the emotions are unrelated to the judgments. So, for example, being in a bad mood because of, say, the weather, can lead to negative judgments about things unrelated to the weather, while positive moods can lead to positive judgments about unrelated things. In their study, participants were first told to write a story about a personal experience that either made them happy or sad. Immediately afterwards, they were presented with a newspaper article the described a political candidate and his or her political positions. Half of the participants were told either to evaluate the candidate as they read the article, in order determine whether they would vote for him/her (candidate-evaluation condition). In this condition, participants were thus motivated to evaluate the candidate. The other half of the participants were told to evaluate the article itself to determine whether it was well-written (article-evaluation condition. These participants had no “extrinsic” motivation to evaluate the candidates. Finally, half of the participants rated how favorable they found the candidate, and whether they would vote for him, immediately after finishing the article, and the other half did so after 24 hours.

There are several interesting comparisons in this design. Participants were either happy or sad, either “extrinsically” motivated (candidate-evaluation condition) or unmotivated (article-evaluation condition), and made their evaluation of the candidate either immediately or 24 hours after reading the article. Furthermore, participants were divided into partisans or nonpartisans based on the strength of their identification with a specific political party. Partisans tend to be more interested in politics, and thus should be more “intrinsically” motivated to evaluate the candidates, regardless of whether they were in the candidate or article-evaluation conditions. Based on the motivated reasoning literature, we would expect that since the emotions participants are experiencing (happiness or sadness) are unrelated to the information they read about the candidates, motivated (extrinsically or intrinsically) should attempt to discount the emotions when evaluating the candidates.

Their results were consistent with this prediction. In fact, motivated participants actually overcompensated for their irrelevant emotions. Both partisans and nonpartisans in the candidate-evaluation condition, and partisans in both evaluation conditions, gave the candidates more positive evaluations when they were sad, and less positive evaluations when they were happy. Nonpartisans in the article-evaluation condition, however, were influenced by the irrelevant emotions, and thus rated candidates as more negative when they were sad and more positive when they were happy. There were no differences between the evaluations made immediately and those made 24 hours later, indicating that participants’ initial judgments, whether they were born of the irrelevant emotions or from overcompensating for those emotions, stuck. The data in the figure below is compiled from both the immediate and delayed evaluations (p. 245).

There you have it: motivated participants recognized that their emotions could influence their judgments, and because those emotions were irrelevant, compensated for them. Granted, they overcompensated, but they didn’t let irrelevant emotions get in the way. Partisans in particular were able to avoid being misguided by irrelevant emotions, regardless of whether they completed a task that motivated them to do so. So, in a way, partisans were less irrational than nonpartisans. Does this mean that partisans always make more rational political decisions than nonpartisans? Obviously not, as the Westen et al. data shows, but it does show that the relationships between rationality, partisanship, and emotion are far from straightforward, and we can’t make a judgment about any of them based on one study. From these two studies combined, a better conclusion might be that partisans are too beholding to their biases and prejudices, while nonpartisans are too fickle and open to irrelevant situational influences. Which is better?

¹Gibson, J.L., & Bingham, R.D. (1982). On the conceptualization and measurement of political tolerance. American Political Science Review, 76, 603-620.
²Lerner, J.S., Goldberg, J.H., & Tetlock, P.E. (1998). Sober second thought: The effects of accountability, anger, and authoritarianism on attributions of responsibility. Personality & Social Psychology Bulletin, 24, 563-574.
³Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the
human brain. New York: Grosset/Putnam.
4Isbell, L.M. & Wyer Jr., R.S. (1999). Correcting for Mood Induced Bias in the Evaluation of Political Candidates: The Roles of Intrinsic and Extrinsic Motivation. Personality & Social Psychology Bulletin, 25(2), 237-249.