Ultimate attribution error
The ultimate attribution error is a group-level attribution error that offers an explanation for how one person views different causes of negative and positive behavior in ingroup and outgroup members. The ultimate attribution error arises as a way to explain an outgroup's negative behaviour as flaws in their personality, and to explain an outgroup's positive behaviour as a result of chance or circumstance. It is also the belief that positive acts performed by ingroup members are as a result of their personality, whereas, if an ingroup member behaves negatively (which is believed to be rare), it is a result of situational factors.
The ultimate attribution error was first established by Thomas F. Pettigrew in his 1979 publication "The Ultimate Attribution Error: Extending Allport's Cognitive Analysis of Prejudice". As the title suggests, the ultimate attribution error is a theoretical extension of Gordon Allport's work in attribution theory.
The ultimate attribution error is a systematic patterning of intergroup misattributions shaped in part by one's prejudices. Prejudiced individuals are more likely to attribute an outgroup member's negative behaviors to dispositional, internal (possibly genetically determined), causes. These same prejudiced individuals are also more likely to attribute outgroup member's positive behaviours to (a) "exceptional case", (b) fluke or special advantage, (c) high levels of motivation, or (d) situational context causes. Through these explanations, a prejudiced individual may disassociate a positive behavior from an outgroup individual and their group. In comparison, one is more likely to attribute negative ingroup behaviors to external causes and positive ingroup behaviors to dispositional causes.
In general, anyone may commit the ultimate attribution error. However, it is most likely to happen to individuals who possess negative prejudices and stereotypes toward an outgroup. This attribution is considered a root of prejudice, as people who commit this attribution will usually see members of other races, religions, cultures, or even social class as dispositionally inferior or flawed, while people from their own racial, cultural, or religious ingroup, upon committing the same negative behaviors, are seen as good people who are dealing with specific situations the best they can. This reduces the acceptance of outgroup members, as any positive behaviors are downplayed and negative behaviors are highlighted.
Explaining away positive behavior of outgroup members
The attribution of outgroup members' positive behavior is classically categorized into four categories, created at the intersection of perceived degree of controllability of act (low vs high) and perceived locus of control of act (internal vs external).
The "exceptional case" explanation is created at the intersection of low controllability of act and internal locus of control. Using this mode of reasoning, an individual excludes the particular outgroup member from the outgroup. That is, they individuate the outgroup member, disassociating them from the group. This view allows for the maintenance of prejudiced beliefs through categorizing the "good" member as an exceptional case, while the other members of their group are still seen as "bad".
Luck or special advantage
The "luck or special advantage" explanation is created at the intersection of low-perceived controllability of act and external locus of control. This reasoning suggest that the outgroup member's positive behavior is not rooted in their skill, ability, or hard work. Rather, their positive outcome is beyond their immediate control and therefore of little significance. "Special advantage" extends this by suggesting that their group affiliation offers some advantage, and therefore the positive outcome is again of little significance.
The "highly motivated" explanation is created at the intersection of high-perceived controllability of act and internal locus of control. Similar to the exceptional case, the highly motivated explanation individuates the outgroup member and dissociates them from their group. The outgroup member's positive behavior is rooted in their drive to be seen as anti-stereotypic, an external force. Thus, they are not seen as intrinsically exceptional, but externally motivated, and, without this motivation, they would not be able to achieve success. That is, an outgroup member's positive behavior is evidence of their response to external pressures of their interaction with ingroup other. Therefore, without an external source of motivation, the outgroup member is just like any other low-achieving, negative-behavior outgroup member.
Similar to the "exceptional case" explanation, this explanation allows for the maintenance of prejudiced beliefs. That is, the highly motivated outgroup member is seen as hard working, so there must be something wrong with the rest of them.
The "situational" explanation is created at the intersection of high-perceived controllability of act and external control of the act. An outgroup member's positive outcome is not rooted in their effort or ability, but a result of external situational factors that are, at least in some part, influenced by others. Therefore, their positive behavior is not their own, and is of little consequence.
The ultimate attribution error is evidenced in a number of studies. Taylor and Jaggi (1974) found results supporting the ultimate attribution error in the causal attributions between religious ingroup and outgroup members. In a 2x2 between-group design, Hindu or Muslim participants were asked to make casual attributions for undesirable acts performed by Hindus or Muslims. Hindus attributed external causes to undesirable acts committed by fellow Hindus, but an internal cause for undesirable acts committed by Muslims. Conversely, Muslims attributed external causes to undesirable acts committed by fellow Muslims, but an internal cause for undesirable acts committed by Hindus. While Pettigrew and many others to follow would focus on race, this study offered clear evidence that similar mechanisms are at play among religious groups.
Prior to Pettigrew's formalization of the ultimate attribution error, Birt Duncan (1976) found that White participants viewed Black individuals as more violent than White individuals in an "ambiguous shove" situation, where a Black or White person accidentally shoves a White person. In a 2x2 between-group design, White participants viewed a Black or White individual (harm-doer) ambiguously shoving a Black or White individual (victim). In general, when a Black harm-doer shoved another person (whether they were Black or White), their behavior was attributed their high dispositional levels of violence (internal). On the other hand, when a White harm-doer shoved another person (whether they were Black or White), their behavior was generally attributed to external constraints. The results suggested that the White students participating in the experiment possessed a lower threshold for labeling a behavior as violent when the harm-doer is Black (outgroup) than when the harm-doer is White (ingroup).
Morris and Peng (1994) find support for Pettigrew's ultimate attribution error in a comparison of casual attributions made by ingroup and outgroup members about a murder carried out by an ingroup or outgroup assailant. In a 2x2 between-group design, American or Chinese participants learned about a recent murder committed by an American or Chinese individual. They received the media coverage for the murder and were asked to weight the dispositional and situational explanations for the cause of the number. As the ultimate attribution error predicts, American participants were biased toward dispositional explanations for the Chinese murder suspect, and biased toward situational explanations for the American murder suspect. Similarly, the Chinese participants attributed dispositional causes for the American murder suspect and situational causes for the Chinese murder suspect.
Together, these three studies (Taylor and Jaggi's, Duncan's, and Morris and Peng's) establish the foundation of the ultimate attribution error and support its general prediction that negative behaviors by outgroup members are more likely to be attributed to internal causes than negative behaviors of ingroup members. Many other studies have been published using the ultimate attribution error as a theoretical foundation.
Supporting this general statement, a meta-analysis of 19 ultimate attribution error studies was published in 1990. The meta-analysis offers limited support for Pettigrew's ultimate attribution error. Specifically, it finds support for three aspects of the ultimate attribution error:
- more internal attribution for positive acts, and less internal attribution for negative acts, by ingroup than outgroup members;
- more attribution of outgroup members' failures to lack of ability, and more explaining away of outgroup members' successes;
- a preference for ingroup-serving versus outgroup-serving attributions for group differences.
- Pettigrew, T. F. (1979). "The ultimate attribution error: Extending Allport's cognitive analysis of prejudice". Personality and Social Psychology Bulletin 5 (4): 461–476. doi:10.1177/014616727900500407.
- Hewstone, M. (1989). Causal attribution: From cognitive processes to collective beliefs. Basil: Blackwell.
- Taylor, D. M; Jaggi (1974). "Ethnocentrism and Causal attribution in a South Indian Context". Journal of Cross-Cultural Psychology 5: 162–171. doi:10.1177/002202217400500202.
- Duncan, B. L. (1976). "Differential social perception and attribution if intergroup violence: Testing the lower limits of stereotyping of Blacks". Journal of Personality and Social Psychology 34 (4): 75–93. doi:10.1037/0022-3518.104.22.1680.
- Morris, Michael W.; Kaiping (1994). "Culture and cause: American and Chinese attributions for social and physical events". Journal of Personality and Social Psychology 67 (6): 949–971. doi:10.1037/0022-3522.214.171.1249.
- Hewstone, Miles (1990). "The 'ultimate attribution error'? A review of the literature on intergroup causal attribution". European Journal of Social Psychology 20: 311–335. doi:10.1002/ejsp.2420200404.