"Power corrupts" and the science behind it

I'm sure most of us have heard the old adage "Power corrupts, and absolute power corrupts absolutely". (If you hadn't before, now you have.)

Well, there's been a recent study on the subject. that concludes there's a kernel of truth there. The study basically concludes that we tend to internally categorize people based on their social status, and that our empathy is greatest for people in the same categories as us. By extension, if you put someone in a position of power - increasing their social status relative to others - they tend to lose empathy for those other people, since they no longer share a common category.

It seems to me that this effect would scale with the number of people involved. The larger the group you hold power over, the greater the power is. Following this theory, that leads to more disconnect and less and less empathy. (It would explain a lot about the behavior that big businesses and government officials get away with.)

Of course, this is the first study that reached that conclusion, as it's apparently something that hasn't been investigated very much. (I suspect such a study might have trouble getting funding, for obvious reasons.) So I have wonder if this is true. Are we built psychologically so that managing too large a government/business/organization is inherently destructive?

Quite honestly, I don't know what to think about this. Abuse of power is certainly something that happens disturbingly often, but I would not argue it as some kind of inevitable, inescapable condition that takes hold on everyone with a meaningful position of leadership.

My take? People are less likely to abuse power if they have most of the same responsibilities as the people they have authority over. My first job was a dishwasher in a restaurant, and the manager of the restaurant never even slightly abused what privileges his position gave him—most notably, he was always the last to leave at the end of the day, and he was pretty much on his feet the entire time every day he worked. (he was also the one who personally made me lunch on my workdays.) What is worth noting is that he did all the same kitchen work as the rest of us did, so I could say he related to us more than say, the store owner. The store owner on the other hand was different. She spent half her time at work texting at the bar, always left early and talked down to the kitchen workers quite a bit.

That's really all I have to say about it off the top of my head.

Well, I'm not sure what that shows, given that owning or running a restaurant is a far cry removed from running a city or a country. Differences in individual personality probably show through more when the difference is 10 people versus 20 people (which is a rough guess of the parallel at this restaurant), as opposed to 100,000 people versus 200,000 people.

I think the OP missed an entire section of the article:

I think that says it all: power corrupts, but those who are not corrupted by it gets to keep it.

Also, a reminder that anecdotal evidence is all but one data point: it might be an exception rather than the norm when compared with other data points. Which means that I'll have to actually have a look at the publications involved (as the article says nothing about the details of the study: how many subjects were involved and where did the researchers get them? How is the mirroring measured? What were the questions asked? etc.

edited 28th Oct '13 4:56:31 PM by IraTheSquire

I found the study.

Incidentally, I didn't miss that. I was just focusing on the implications for larger organizations like governments, big businesses, and so on. A lot of the mid-level business managers rarely have to deal with that many people, or have their decisions affect that many people.

Just a look at the figure in the results already makes me doubtful: the standard deviations of the three power conditions overlap with each other (which means that there is a 95% chance for that result with any two of them can be equal) and there's no R square value for the line of best fit in the other graph, where the data points look very wide spread.

That's not how standard deviations work, since the farther away you move from the center of a curve, the lower the chance of occurrence becomes. The actual chances of overlap, based on the graph, are somewhere south of 25% for both the high-medium and medium-low cases. (I'd have to run my own calculations with the raw data, which obviously we don't have, to get the actual values.)

That said, it is a little problematic (and odd) that the r-squared value is not mentioned in the study. That does make it harder to verify the accuracy of the line of best fit. Hmmmm.

My take is that it's not so much that 'power corrupts' but that 'corruption begets power'. That is to say, those who make it to the top are those who are already corrupt, those who were willing to be totally unscrupulous in their efforts to get there. This is a more hopeful view because it means corruption is not inevitable, but a result of a culture that rewards corruption and punishes integrity and honesty. So if we really wanted to, we could change that culture.

That aside, I certainly see every reason to believe that people who have a great deal of power/wealth become detached from the difficulties of the have-nots, and invent justifications for holding onto (and expanding) their power/wealth even by immoral means, and tend to see other rich/powerful people as closer to them than those in the lower echelons. Again, though, this is a problem that comes from culture, or perhaps human psychology (related to the monkeysphere, for instance); not an inevitability.

Maybe not, but the significance of the data is still far too small to say that the result is not due to random (significance should be at least two before you can say that the result is not due to random with a 95% confidence). And 25% is still quite huge in terms of probability and scientific analysis (considering p-value is normally has to be less than 0.05 and confidence needs to be at least 95%).

edited 28th Oct '13 9:59:09 PM by IraTheSquire

I'd agree, Ira. The paper is interesting insofar as it suggests a better constructed, larger study with more intense data analysis is probably worthwhile having a go with.

It's a possible fund raiser, rather than anything else. <shrugs> It goes on the "interesting, but flawed" pile.

edited 29th Oct '13 2:27:57 AM by Euodiachloris

@ Meklar: Then how would you go about changing base human psychology; you must be aware it won't be as easy as you think it is?

well that assumes that the idea of power corrupting IS a part of base human psychology, and not a base part of our civilization since human were still trying to thrive against nature in Africa and being the dirtiest meanest person was possibly smartest way for a single person to live.

And... that so misses the point of what makes hominids hominids. <_< The group.

Individuals can be right nasty or even downright nice, but for some true terror, you get a band together and cause havoc to others to survive. We're pretty good at it.

Being individual horrors beyond hope? Usually leaves you a groupless wonder, doomed to wind up as cat food.

edited 29th Oct '13 11:18:33 AM by Euodiachloris