Psychology

All power tends to corrupt

But power without status corrupts absolutely

DURING the second world war a new term of abuse entered the English language. To call someone "a little Hitler" meant he was a menial functionary who employed what power he had in order to annoy and frustrate others for his own gratification. From nightclub bouncers to the squaddies at Abu Ghraib prison who tormented their prisoners for fun, little Hitlers plague the world. The phenomenon has not, though, hitherto been subject to scientific investigation.

Nathanael Fast of the University of Southern California has changed that. He observed that lots of psychological experiments have been done on the effects of status and lots on the effects of power. But few, if any, have been done on both combined. He and his colleagues Nir Halevy of Stanford University and Adam Galinsky of Northwestern University, in Chicago, set out to correct this. In particular they wanted to see if it is circumstances that create little Hitlers or, rather, whether people of that type simply gravitate into jobs which allow them to behave badly. Their results have just been published in the Journal of Experimental Social Psychology.

Dr Fast's experiment randomly assigned each of 213 participants to one of four situations that manipulated their status and power. All participants were informed that they were taking part in a study on virtual organisations and would be interacting with, but not meeting, a fellow student who worked in the same fictional consulting firm. Participants were then assigned either the role of "idea producer", a job that entailed generating and working with important ideas, or of "worker", a job that involved menial tasks like checking for typos. A post-experiment questionnaire demonstrated that participants did, as might be expected, look upon the role of idea producer with respect and admiration. Equally unsurprisingly, they looked down on the role of worker.

To manipulate their power, participants were told there would be a draw for a $50 bonus prize at the end of the study and that, regardless of their role, each participant would be able to dictate which activities his partner must engage in to qualify to enter the draw. Participants that Dr Fast wanted to imbue with a sense of power were informed that one other element of their role involved dictating which "hoops" their partners would have to jump through in order to qualify for the draw, and that they controlled the amount of effort the partner had to exert in order to win the $50. They were also told that the partner did not have any such control over them. In contrast, low-power participants were informed that while they had the ability to determine the hoops their partner had to jump through, that partner ultimately had more control because he could remove the low- power participant's name from the raffle if he did not like the hoops selected.

Participants were then presented with a list of ten hoops and told to select as many as they liked (but a minimum of one) for their partner to jump through. Unknown to the participants, Dr Halevy and Dr Galinsky had conducted an independent test, using 58 people not involved in the main study, to rate how demeaning, humiliating, degrading, embarrassing and uncomfortable each of the ten possible activities actually was. Five of the ten were rated as deeply demeaning. These included things like: "say ‘I am filthy' five times" and "bark like a dog three times". The other five were not considered particularly demeaning. They included: "tell the experimenter a funny joke" and "clap your hands 50 times".

Participants who had both status and power did not greatly demean their partners. They chose an average of 0.67 demeaning activities for those partners to perform. Low- power/low-status and low-power/high-status participants behaved similarly. They chose, on average, 0.67 and 0.85 demeaning activities. However, participants who were low in status but high in power-the classic "little Hitler" combination-chose an average of 1.12 deeply demeaning tasks for their partners to engage in. That was a highly statistically significant distinction.

Of course, not everybody in the high-power/low-status quadrant of the experiment behaved badly. Underlying personality may still have a role. But as with previous experiments in which random members of the public have been asked to play prison guard or interrogator, Dr Fast's result suggests that many quite ordinary people will succumb to bad behaviour if the circumstances are right.