Bush v . Bin laden : effect of State emotion on Perceived threat Is Mediated by emotion towards the threat agent

The authors conducted an experiment to examine the effect of specific (fear and anger) and global emotional states on perceptions of threat posed by either George W. Bush or Osama Bin Laden. The findings supported a mediator model in which negative emotion towards the threat target mediated the effect of global negative emotion on perceived threat. The authors discuss implications of the findings for theories that postulate an effect of emotion on risk perceptions and for understanding threat perception in the terrorism context. Résumé Les auteurs ont mené une étude expérimentale afin d’examiner les effets d’états émotionnels spécifiques (peur et colère) et globaux sur les perceptions de menaces suscitées soit par George W. Bush, soit par Osama Ben Laden. Les résultats soutiennent un modèle de médiation dans lequel une émotion négative éprouvée envers une cible menaçante médiatise l’effet d’une émotion globale négative sur la menace perçue. Les auteurs discutent des implications des résultats pour les théories qui postulent un effet de l’émotion sur la perception du risque et pour comprendre la perception de menace dans le contexte du terrorisme. Key-words Threat perception; risk perception; emotion; terrorism Mots-clés Perception de menace, perception de risqué, émotion, terrorisme This article has been accepted in the scope of the special issue on social psychological perspectives on terrorism (cf. Arciszewski, Verlhiac, & Kruglanski, 2009). *Thinking, Risk, and Intelligence Group , DRDC Toronto, 1133 Sheppard Avenue West, P.O. Box 2000, Toronto, Ontario M3M 3B9, Canada. E-mail: [email protected]. Phone: 416-635-2000 ext. 3146. E-mail: [email protected]. Bush v. Bin laden: effect of State emotion on Perceived threat Is Mediated by emotion towards the threat agent Bush vs. Ben Laden : l’effet de l’émotion état sur la menace perçue est médiatisées par l’émotion vis-à-vis de l’agent menaçant David R. Mandel* Oshin Vartanian* RIPS_1_2010.indd 5 19/05/10 17:17:53 ép re uv e de c on tr ôl e eMotIon and thReat PeRcePtIon 6 It has long been known that perceptions of risk or threat among the general public are influenced by a multidimensional array of psychosocial factors that include emotions such as dread and outrage (e.g., Fischhoff, Slovic, Lichtenstein, Read, & Combs, 1978; Sandman, 1989; Slovic, 1987). Leaders of terrorist organizations, such as al Qaeda’s Osama Bin Laden (OBL), attempt to use these psychological bases of risk perception to their strategic advantage, seeking to prompt fear and a sense of vulnerability that is disproportionate to the statistical risk actually posed, yet highly representative of the iconic images of terror that acts of terrorism so easily evoke (Slovic, 2004). Some have also suggested that leaders of states threatened by terrorism, such as former U.S. President George W. Bush (GWB), have used those same emotions to bolster political support for hawkish countermeasures by presenting the threats in ways that are, as Mueller (2006) puts it, “overblown.” In this paper, we examined how Canadian participants’ emotions predicted their threat perceptions regarding two key actors in the global war on terrorism – OBL and GWB. Our inquiry was guided by recent theoretical developments in the psychology of emotion, which lend themselves to competing hypotheses regarding the effect of emotion on threat perception, and which we summarize next. Several accounts posit an effect of emotion on risk perception. However, an important distinction between them is whether they are valence-based or emotion-specific. Valence-based accounts propose that how good or bad a person feels at the time they are evaluating risks will be an important determinant of their risk perceptions (e.g., Clore & Huntsinger, 2007; Loewenstein, Weber, Hsee, & Welch, 2001; Schwarz & Clore, 1996; Slovic, Finucane, Peters, & MacGregor, 2002). A classic demonstration of this sort of effect was provided by Johnson and Tversky (1983), who showed that participants who were experimentally induced to feel positive were more optimistic about risks than their counterparts who were induced to feel negative, even when the risks assessed were semantically unrelated to the mood stimuli. In the terrorism domain, Shiloh, Güvenç, and Önkal (2007) found that affect negativity was directly related to perceived costs of terrorism and inversely related to perceived control in Turkish and Israeli samples. As well, negative affect was directly related to perceived vulnerability in the Turkish sample. RIPS_1_2010.indd 6 19/05/10 17:17:54 ép re uv e de c on tr ôl e Revue InteRnatIonale de PSycholoGIe SocIale 2010 n° 1 7 Whereas affective valence theories stress the effect of the goodness or badness of one’s affective state on judgment, emotion-specific theories posit that different emotions that share the same valence may nevertheless lead to different, even opposing, effects on judgment. The basis for this claim is that different emotions are not only the consequence of distinct cognitive (Smith & ellsworth, 1985) and action (Frijda, Kuipers, & ter Schure, 1989) appraisals, but that they also give rise to distinct appraisals that form an important part of the basis for emotion’s influence on judgment (Bodenhausen, Sheppard, & Kramer, 1994; DeSteno, Petty, Wegener, & Rucker, 2000; Keltner, Ellsworth, & edwards, 1993; Lerner & Keltner, 2000; Mandel, 2003; Tiedens & Linton, 2001). According to this view, the appraisal tendencies generated by specific emotions can persist, spilling over to influence judgments even when the target of judgment differs from the emotion-eliciting stimulus (Gasper & Clore, 1998; Goldberg, Lerner, & Tetlock, 1999). In terms of risk perception, two emotions that have received research attention are fear and anger. Although both are negative emotions, fear arises from and gives rise to appraisals of uncertainty and situational control, whereas anger is associated with appraisals of certainty and personal control (Lerner & Keltner, 2001; Smith & ellsworth, 1985). Given that perceived risk is inversely related to perceived certainty and personal control (McDaniels, Axelrod, Cavanagh, & Slovic, 1997; Slovic, 1987), there is reason, as well as mounting evidence, to support the hypothesis that perceived risk might be amplified by feelings of fear and attenuated by feelings of anger (Lerner & Keltner, 2000; 2001; Lerner & Tiedens, 2006). Notably, Lerner, Gonzalez, Small, and Fischhoff (2003) examined the effects of experimentally induced fear and anger on terrorism risk perceptions in a representative U.S. sample shortly after 9/11. Compared to anger-induced participants, fear-induced participants perceived greater risk of terrorism-related threats to the U.S., themselves, and average others. In a sub-sample that was examined a year later, a new induction of fear and anger replicated these results (Fischhoff, Gonzalez, Lerner, & Small, 2005). RIPS_1_2010.indd 7 19/05/10 17:17:54 ép re uv e de c on tr ôl e eMotIon and thReat PeRcePtIon 8 the Present Research We sought to extend the investigation of the effect of emotion on threat-related judgments in the terrorism domain in several important respects. Like Lerner et al. (2003), we experimentally induced fear and anger in different groups of participants. However, as some researchers have recently called for (e.g., Small, Lerner, & Fischhoff, 2006), we added a neutral-emotion baseline condition, which permitted us to gauge whether the effects of fear and anger on perceived threat were symmetric relative to a neutral baseline. This assessment is important in light of recent findings by Bruine de Bruin, Florig, Fischhoff, Downs, and Stone (2006) showing that whereas self-reported fear (controlling for anger and baseline fear) associated with terrorism and natural disaster scenarios was directly related to perceived mortality risk, self-reported anger (controlling for fear and baseline anger) was unrelated to perceived risk. This suggests that earlier findings by Lerner et al. (2003) and Fischhoff et al. (2005) may have been due mainly to the risk-enhancing effect of fear rather than the risk-attenuating effects of anger. Second, by examining participants’ current emotional state across a broad range of emotions, we were able to test whether threat perceptions were predicted by general composite measures of negative and positive emotion. As noted earlier, affective valence accounts posit that perceived risk tends to increase as one moves toward the negative pole of the good-bad continuum. Such proposals suggest a single, bipolar, affective dimension. An alternative valence-based perspective, however, is that there are two, unipolar, “positivity” and “negativity” dimensions that might differentially impact perceived risk. We hypothesized that the negative dimension would be a more influential predictor of perceived threat for four reasons. First, if current emotion is used as a heuristic for judging risk, it seems plausible that negative emotion would be more influential than positive emotion due to its greater representativeness to the target of judgment – namely, events that, by definition, are likely to induce negative emotion. Second, it has long been known that risk perception is influenced by negative emotions, such as feelings of dread (Slovic, 1987). Third, because people tend to be loss averse (Kahneman & Tversky, 1979), they may be more closely attuned to the informaRIPS_1_2010.indd 8 19/05/10 17:17:54 ép re uv e de c on tr ôl e Revue InteRnatIonale de PSycholoGIe SocIale 2010 n° 1 9 tional value of negative emotion. Finally, primary negative emotions, such as fear, are given processing priority in the human brain (Le Doux, 1998), perhaps due to the evolutionary significance of attending to their sources. A third objective of ours was to examine whether any significant effect of emotion (valence-based or emotion-specific) on perceived threat might be mediated by participants’ emotional responses toward the source of threat itself. In the present research, we manipulated whether the evaluated source of threat was OBL or GWB, two iconic figures in the terrorism domain at the time this study was being conducted in 2006. Indeed, around the same time as our study was conducted, an EKOS poll (Harper, 2006) found that Canadians regarded GWB as the third greatest danger to the world after OBL and North Korean leader Kim Jong-il. We showed participants a picture of the relevant threat agent and had them rate their emotional responses toward the target, after which they rated a variety of threats posed by the target. We hypothesized that, to the extent that emotion experienced prior to the target evaluation tasks predicted perceived threat, it would be mediated by emotion specifically evoked by the threat agent. Finally, we examined how our Canadian sample perceived the threats posed by OBL and GWB as a function of their geographic context (namely, Canada or international) and their value domain (namely, threats to national security or individuals rights). We expected that participants would perceive greater threat in the international domain than in the Canadian context, given that Canada has experienced few acts of terrorism. Our examination of value domain was exploratory and motivated by the fact that national security and personal freedoms represent two of the key values underlying debates about the threat of terrorism and counter-terrorism response, and discussions of the appropriateness of counter-terrorism response often focus on the need to balance these values. Therefore, they represent a key value tradeoff in the terrorism domain. RIPS_1_2010.indd 9 19/05/10 17:17:55 ép re uv e de c on tr ôl e eMotIon and thReat PeRcePtIon