Mixing Memory

Humans are strange animals. We have such a deep need for social order that, when that order is threatened, we’ll do irrelevant things in an effort to preserve it. For example, when people are told that the conviction rate for a particular crime is low, they’ll assign harsher punishments to individuals who commit that crime1. Because the low conviction rate is perceived as a threat to social order, they take it out on the next person associated with that threat, even though increasing one person’s punishment does nothing to actually affect the perceived threat. Since terrorism is clearly one of the most visible threats to social order in Europe and the United States, an interesting and important question is, how does this threat to social order affect our behavior? This question was addressed in a series of studies presented in a paper by Fischer et al., in press at the Journal of Experimental Social Psychology.

Their first study was conducted one day and four weeks after the London subway bombings in July of 2005. One hundred forty-four German participants were told that the researchers were doing a study for the German government, which was considering changing the financial punishments for various crimes. For one of those crimes, car theft, the participants read a short description of the crime, and were asked how much the perpetrator should be fined once convicted. They were also asked how likely (on a scale of 1-10) they believed a terrorist attack Germany was. Unsurprisingly, with the terrorist attacks in London fresh in their minds, the participants tested one day after the attacks thought the probability of a terrorist attack in Germany was higher than did participants tested four weeks after the attacks. Participants tested the day after the attacks also suggested a higher fine (8503 Euros) than participants tested four weeks later (5996 Euros). This result suggests that the salience of the terrorist attacks one day after they occurred, and the resulting perception of a threat to the social order, led participants to suggest a higher punishment for a crime totally unrelated to terrorism.

Studies two and three sought to replicate the findings from the first study, with a little more control over “terror salience.” This time, Fischer et al. had participants read a short paragraph in which they learned that German intelligence services believed the likelihood of a terrorist attack in Germany to be either high or low, and then suggested fines for car theft. As in the first study, participants for whom terror salience was high (after reading that the German intelligence services thought the likelihood of an attack was high) suggested higher fines (4770 Euros) than did low terror salience participants (3442 Euros). In study three, terror salience was manipulated by presenting participants with either photos of terrorist attacks or neutral photos, and once again, participants in the high terror salience condition suggested higher fines for car theft (10158 Euros) than low terror salience participants (4608 Euros).

In order to determine whether terror salience was to blame for more severe suggested punishments, rather than simply negative affect (terrorism is depressing), Fischer et al. again showed participants different types of photos to manipulate terror salience in the fourth study, but this time used three types: photos of terrorist attacks (high terror salience condition), neutral photographs (low terror salience, low negative affect condition), and photos of natural disasters (low terror salience, high negative affect condition). They were then presented with the story of the court trial of a man accused of rape, presented as a newspaper article (participants were told that they were participating in studies about the presentation of information in the media). They were asked several questions about their perception of the crime, including how many months the convicted rapist should spend in prison. Participants in the high terror salience condition suggested longer punishments (149 months) than participants who saw photos of natural disasters (105 months), and participants who saw the natural disaster photos suggested longer jail times than participants who saw neutral photos. Regression analysis indicated that the effect of terror salience was largely due to participants desire to prevent future crimes (one of the questions they were asked), indicating that terror salience caused participants to act to preserve the social order.

So, in four studies, participants for whom the possibility of terrorist attacks was very salient suggested harsher punishments for crimes unrelated to terrorism. It’s almost as though participants, frustrated by their inability to prevent attacks, or punish the perpetrators (in the London bombings, the perpetrators were all suicide bombers, and thus dead), lashed out at the next person who did anything that could be perceived as a threat to social order, even if that person’s actions were completely unrelated to terrorism. I suspect that terror salience, and the resulting perceived threat to the social order, can lead to all sorts of reactive behaviors. Hopefully, future research will explore this, because one can easily imagine situations in which reactionary behaviors caused by terror salience become threats to social order themselves.

1Rucker, D. D., Polifroni, M., Tetlock, P. E., & Scott, A. L. (2004). On the assignment of punishment: the impact of general-societal threat and the moderating role of severity. Personality and Social Psychology Bulletin, 30, 673-684.
2Fischer, P., Greitemeyer, T., Kastenmüller, A., Frey, D., & Oßwald, S. (In Press)> Terror salience and punishment: Does terror salience induce threat to social order? Journal of Experimental Social Psychology.


  1. #1 Michael Anes
    December 13, 2006

    I’m noticing an interesting pattern to the results of the 4 experiments – In experiment 1 and 2 (verbal messages conveying salience) the ratio of high to low punishment is nearly identical (1.42 in Exp 1 and 1.39 in Exp 2). In the pictorial condition of experiment 3, the ratio is 2.20. Can you say what the result was for the low punishment in exp 4 (yes, I know it’s months rather than euros -still I’m interested). If it was 65-70 the ratio would be the same as in exp 3.
    Did the researchers give a range of options or constrain punishment options in some way (multiple choice?) that might result in the similarity of ratios high to low?

  2. #2 Alex Leibowitz
    December 14, 2006

    1. “Human beings are strange animals.” As opposed to what?
    2. What is your working concept of “the social order”?

  3. #3 Chris
    December 14, 2006

    As opposed to our intuitions, for one.

    For a working concept of “social order,” start with Durkheim.

  4. #4 Alex Leibowitz
    December 14, 2006

    I can’t exactly put my finger on it, but research like this bothers me. Maybe because it tries to support certain philosophical claims with empirical data?

    Or maybe I’m just not enough of a naturalist.

  5. #5 Chris
    December 14, 2006

    Alex, I understand your concern, and if we were just using this experiment, “Social order” would be a stretch, but there is a pretty large literature on this stuff. If you want, I can email you a copy of the study, where you can get some of the references. Just drop me an email.

  6. #6 Neil
    December 15, 2006

    when people are told that the conviction rate for a particular crime is low, they’ll assign harsher punishments to individuals who commit that crime1. Because the low conviction rate is perceived as a threat to social order, they take it out on the next person associated with that threat, even though increasing one person’s punishment does nothing to actually affect the perceived threat.

    This last claim might be false. If we cannot raise the conviction rate, we might nevertheless increase deterrence by raising the associated punishment. If people are utility maximizers, their actions should be sensitive to the cost x probability of detection.

  7. #7 Neil
    December 16, 2006

    And a question: to your knowledge, has anyone tested the effect of threats to the social order on compassion? On, for instance, attitudes to social welfare programs?

  8. #8 Chris
    December 16, 2006

    Neil, the claim might be false, and it might even be one way that people justify the behavior, but I would be surprised if it’s actually part of the decision process. That’s an interesting empirical question, though.

    And I don’t know of any that look at compassion, but I’m going through the literature now, so I’ll keep an eye out.

  9. #9 Smillie George
    January 3, 2008

    This research just confirmed the quite well-known fact that when someone is in a bad mood because he either believes that his country will get attacked or because the weather is bad or because he had an argument with his girlfriend or whatever,it’s highly likely that he will be more harsh toward everything and everyone that annoys or might annoy him,for example someone who steals cars(most people consider car theft to be an antisocial act,an act that is really annoying).