Do organizations have an inherent tendency to promote immorality? This is the question posed by a recent experiment (pdf) by Armin Falk and Nora Szech.

They offered subjects the choice: to gas a mouse to death for 10€, or to let it live and get nothing. 45.9% of subjects chose to kill the mouse.

Then, they put other subjects into groups of eight, and gave each individual a choice of A or B. Option A was to let the mouse live and get nothing.Option B was to choose to kill the mouse and get 10€. The subjects were told that if one or more of their group chose option B, eight mice would die. Faced with this choice, 58.6% of people chose to kill the mice. That's a statistically significant 28% more.

The difference, they believe is because of what they call "diffused pivotality." In the second experiment, subjects can convince themselves that they aren't directly responsible for the deaths of the mice, as they can tell themselves: "if I don't choose to kill the mice, someone else will, so I might as well pocket the euros."

In other words, if people can pass responsibility onto others, they are more likely to act immorally.

This is not a new finding. It corroborates Stanley Milgram's famous experiment which showed that people are prepared to give others potentially fatal electric shocks if they believe they are sanctioned to do so by an authority figure. And it has been traditional for organizers of firing squads to put a blank bullet into a gun, so that members of the squad can convince themselves that they fired the blank and so were not responsible for the killing.

These results, say Falk and Szech, "demonstrate the power of organizations to promote immoral outcomes."

But what type of organization? It could be any, for example:-

- In hierarchies, junior members can absolve themselves of responsibility by thinking they are doing what their boss wants.

- In democracies, we might be willing to vote for selfish policies (such as immigration controls?) because we feel our vote won't be pivotal.

- As Steve points out, regulation can actually engender bad behaviour by giving people a "means of neutralization"; responsibility can be passed onto the regulators.

- In markets, people can claim that they are compelled to act badly by the forces of competition. People who sell arms to dodgy regimes commonly claim that if they didn't do so, others would.

Because of self-serving biases, I suspect that people will be quick to pass responsibility to others. However, my hunch is that hierarchical organizations are more conducive than other forms to convincing people that "diffused pivotality" exists and that they are not responsible for outcomes. If this is the case - and several million deaths in the 20th century are consistent with the hypothesis - then hierarchy is, literally, murderously wrong.