The Ultimate Attribution Error in the Killing of George Floyd

Derek Chauvin’s behavior reflects his own stereotypes, not George Floyd's acts.

Posted Jun 09, 2020

In 1979, social psychologist Thomas Pettigrew coined the term “ultimate attribution error” to describe a combination of demonstrably false assumptions that observers from one socially defined group (for our purposes, the ingroup) often make regarding the behavior of certain people from a different socially defined group (for our purposes, the outgroup).

Pettigrew was especially interested in the assumptions that ingroup members make regarding the positive behavior of outgroup members, which ingroup members are likely to regard as atypical. In turn, I'll focus on one particular assumption (i.e., “manipulable situational control”) that occurs when the ingroup member perceives the outgroup member’s positive behavior as (1) highly controllable, yet (2) controlled by forces outside (rather than within) the outgroup member.

I believe that that this particular aspect of the ultimate attribution error was on full display when Derek Chauvin, a police officer of European descent, killed George Floyd, an African-descent detainee, in Minneapolis, Minnesota on May 25, 2020.

For nearly nine minutes, Chauvin pressed his knee into Floyd’s neck, as Floyd lay handcuffed, face-down, and completely prone on the street. Objectively speaking, anyone who witnessed that fateful encounter in person or via social media would conclude that Floyd was compliant, not resisting arrest — and, most important, not saying or doing anything that represented a threat to Chauvin’s physical or psychological existence.

Nevertheless, Chauvin behaved as if Floyd posed an imminent threat to somebody. Paraphrasing Pettigrew’s 1979 description of the “manipulable situational control” scenario that I mentioned, Chauvin’s oddly dispassionate yet sustained behavior seemed to suggest, “So what if he isn’t fighting back? My three fellow officers and I have him surrounded; of course he’s complying with us! I had better extinguish this threat before he extinguishes me [or someone else], now or in the [foreseeable] future.”

Consistent with Pettigrew’s perspective, I conclude that Chauvin was responding to a stereotype regarding African-descent people in general, being predisposed toward violence even when completely subdued. In contrast, Chauvin was not responding to the actual behavior of George Floyd in particular. As Pettigrew would have anticipated, the “ultimate attribution error” is magnified when the observer (in this instance, Chauvin) is a racist.

Although I have made such arguments throughout my career as a social psychologist (e.g., Gaines & Reed, 1994, 1995; Reed & Gaines, 1997), I feel compelled to repeat these arguments as I grapple personally with the amply documented, cold-blooded brutality with which Chauvin killed Floyd. It’s the least that I can do, within academia, to complement the waves of protests and demonstrations that currently are surging within and outside the “ivory tower.”

Some critics (most notably Hewstone, 1990) have questioned whether the “ultimate attribution error” is as far-reaching, or as inherently mistaken, as Pettigrew contended. However, from the moment that I saw the initial image of Chauvin’s face seemingly masking emotion — yet clearly conveying prejudice — as Chauvin killed Floyd, I was convinced that I had witnessed the “ultimate attribution error” all over again, albeit in an exaggerated (and particularly grotesque) form.  Especially when one considers that American society taken as a whole (not just individual Americans taken separately) readily accepts European-descent persons and police officers as ingroup members, even as American society readily rejects African-descent people and detainees as outgroup members, I believe that Pettigrew got it right.

If social psychologists are going to play a role in improving intergroup relations after (if not during) the conflict-driven Age of Trump, then I think that we, myself included, will need to reacquaint ourselves with Pettigrew’s still-timely message.         


Gaines, S. O., Jr. & Reed, E. S. (1994).  Two social psychologies of prejudice:  Gordon W. Allport, W. E. B. Du Bois, and the legacy of Booker T. Washington.  Journal of Black Psychology, 20, 8-28.

Gaines, S. O., Jr. & Reed, E. S. (1995).  Prejudice:  From Allport to Du Bois.  American Psychologist, 50, 96-103.

Hewstone, M. (1990).  The “ultimate attribution error”?  A review of the literature on intergroup causal attribution.  European Journal of Social Psychology, 20, 311–335.

Pettigrew, T. F. (1979).  The ultimate attribution error:  Extending Allport's cognitive analysis of prejudice.  Personality and Social Psychology Bulletin, 5, 461–476.

Reed, E. S., & Gaines, S. O., Jr. (1997).  Not everyone is "different-from-me":  Toward an historico-cultural account of prejudice.  Journal of Black Psychology, 23, 245-274.