The power to corrupt

Following on from Sunday's post, Smarter people are more prejudiced against larger numbers of people than anyone else, there is another article of interest.

In my first post, I commented on sociology research indicating that people with higher IQs were more prejudiced against more people than those with lower IQs. Based on the numbers, high IQ are four times more prejudiced than lower IQ people. It's sociology so all the numbers are suspect but an interesting possible insight.

This morning I came across Does Power Really Corrupt by Matthew Sweet. It is an interesting round-up of the state of our current knowledge. The answer is that we don't know whether power corrupts. Early sociology studies strongly suggested that it did but they were poorly designed experiments with strong vectors of confirmation bias. The replicability has been poor. The field is ideologically skewed and biased towards an affirmative answer regardless of the data.

Egloff has been doing research since 1993 and is used to the bloody process of peer review. But he was shocked by the hostility towards his work. “I am not on a crusade,” he says. “I am not rich. My family is not rich. My friends are not rich. We never received any money from any party for doing this research. Personally I would have loved the results of the Berkeley group to be true. That would be nice and would provide a better fit to my personal and political beliefs and my worldview. However, as a scientist…” The experience of going against this particular intellectual grain was so painful that Egloff vows never to study the topic of privilege and ethics again.

Who, then, is right? Are powerful people nicer or nastier than powerless ones? How can we explain the disparate answers yielded by these two sets of data?

It may be that rich people are better at disguising their true nature than poor people. If being generous in public brings rewards, then rich people might be more inclined to help old ladies across roads. Selfish driving is consistent with this idea: the anonymity of the road means that aggressive petrolheads need not worry about damaging their reputations. And Keltner points out that the data come from people’s accounts of their own charitable giving, and not from watching them in the act. “We know from other studies that the wealthy are more likely to lie and exaggerate about ethical matters,” he says. “Survey self-report data in economics and face-to-face data in psychology capture different processes. What I say I do in society versus how I behave with actual people.”

But it is also possible that the problem lies not with the survey data but with the psychological experiments. Over the past year, this possibility has become the subject of bitter debate. In August 2015, the journal Science reported that a group of 270 academics, led by Brian Nosek, a respected professor of psychology at the University of Virginia, had attempted to reproduce the results of 100 psychological studies. Ninety-seven of the original studies had produced statistically significant results. Only 36 of the replications did the same. Those numbers threatened to undermine the entire discipline of experimental psychology, for if a result cannot be replicated then it must be in doubt. In March 2016 a panel of luminaries claimed to have detected serious shortcomings in the methodology of Nosek’s paper. The inquiry was led by Dan Gilbert, a Harvard professor with a history of hostility to the replicators. (“Psychology’s replication police prove to be shameless little bullies,” he tweeted in 2014, defending another researcher whose work was questioned.) When a journalist from Wired magazine asked Gilbert if his defensiveness might have influenced his conclusions, he hung up on them. Psychology’s “Replication Crisis” might not yet be over.

In September 2015, five social psychologists and a sociologist published a paper in the Journal of Behavioral and Brain Sciences that suggested why psychology might show privileged people in a bad light. Left-wing opinion, contended Jonathan Haidt and his co-authors, was over-represented in psychology faculties. This, they suspected, might be distorting experimental findings – as well as making campus life difficult for researchers with socially conservative views. “The field of social psychology is at risk of becoming a cohesive moral community,” they warned. “Might a shared moral-historical narrative in a politically homogeneous field undermine the self-correction processes on which good science depends? We think so.” So does Boris Egloff. “It was a great and timely paper,” he says. “I congratulate them on their courage.” But it came too late for him. “We spoilt the good guys’ party,” he says.

So the answer is that we simply do not know whether power corrupts and we are not likely to know out of the current sociological academy - they are too undisciplined in their science and too ideological in their convictions.

My suspicion is that the answer will be a weak yes, highly contingent on roles, structures, personality traits, industry and circumstance. Some people with some personality traits in some roles, within some structures, in some industries under circumstances of low transparency and low accountability are likely to display some predictable changes in degree of measured corruption.

But we are a long way from knowing that at this point in time.