May 29, 2008

The New Yorker Writes About Power

With authority.


I've been a subscriber for almost thirty years, and I'm always hopeful it won't again contain another Updike story or Hendrick Hertzberg screed, but inevitably disappointed am I.  I guess it's hard to find different writers for a magazine that purports to be for the literate when it is really a middlebrow primer for an unsophisticated political bias.   If that bias catered to the liberal or leftward leaning affluent it would be at least worth the price of subscription, but it promotes a more general bias: power is bad.  Anything the powerful does is wrong.

To illustrate this, consider the latest issue containing an article, "Power Hour."  It describes an impromptu experiment on the world's powerful  (i.e. Paul Wolfowitz) that unfortunately failed to show the intended result, namely, that he's a jerk.

The fact that the experiment failed didn't stop The New Yorker from trying to portray Wolfowitz as a jerk anyway, even though his performance theoretically showed that he was not a jerk, by the outcome measure of the study.  They even commissioned a caricature of Wolfowitz to accompany the article, because, well, he's a jerk, of course.  No matter what science says.

My own unscientific study suggests The New Yorker's readership is especially those who believe that the little guy always gets trampled on by the powerful, but that they, the readers, are neither little nor powerful, they're in the middle.  They are a select group that lacks power, but possesses wisdom.  They understand what's really going on.  Coincidentally, this describes another overlapping group: academics.


This coincidence makes the second sentence of the article "Power Hour" so perfect:

To academics, one of the best [inverse] indicators of a person's place in a hierarchy is his tendency toward "perspective taking"--"stepping outside of one's own experience and imagining the emotions, perceptions, and motivations of another individual..." An assistant is perspective taking when he gets the boss's coffee before he asks for it; when the boss forgets to pay the assistant back the $3.75, he is not.

The distinction may seem like the main point, but there's a subtext:  the boss is a jerk in both examples.  I'm willing to bet real money that if you had to guess who reads The New Yorker, it would be the assistant.


All of this is besides the point.  The point is how an amateur researcher redid an experiment published in 2006 in Psychological Science about the effects of being powerful, this time using attendees at Time's 100 Most Influential People banquet.  I'll describe the experiment in detail, but let me preface this with the observation that it seems not to have occurred to the researcher, nor anyone at The New Yorker, that "Most Influential" might be a different group than "Most Powerful."

Here's the experiment: take powerful and not powerful people, and have them draw an E on their foreheads.  Powerful people, as it suggests above, have no interest in "perspective taking" and so will write it facing themselves, because they are self-oriented; not-powerful people will write it facing out (legible to observers) because they will be better able to assume another's perspective.

If this sounds like crap to you, you're right.

But it gets worse.

The New Yorker writes:

The hypothesis of the Psychological Science study was that the more power a person has, the less capacity he has to take another person's perspective.

Which is exactly not what the hypothesis of the article in Psychological Science was.  The actual hypothesis was:

we hypothesized that power would decrease perspective taking.

That gerund is extremely important.  The actual hypothesis doesn't claim less capacity, or less desire-- simply that it happens less.  It could happen out of lack of capacity, but also maybe because of lack of time, need to distance oneself from the cacophony of a myriad of individual perspectives, a blaring iPod, solar flares, alcoholism, etc.  How a powerful person writes an E tells you absolutely nothing about why he did it that way, if indeed it is even true that there is an association between being powerful and drawing an E.  But The New Yorker wants it, wants it very badly, to be about a defect in ability.

In fairness to The New Yorker, the authors of the study themselves make this same mistake.


Start from the middle-- study design.  The subjects were not actually powerful: they were college students, age about 20.  They were primed for power by being asked to write an essay about

  • either an experience where they had power over someone (high power group), or
  • an experience where someone else had power over them (low power group)

You can see that simply the ability to recollect such an experience is a huge confounding variable.   Imagine an actually powerful person trying to recall being powerless-- is this person now primed for powerlessness, or-- rage?  Which is entirely different than the priming a powerless person asked to recollect being powerless might experience (e.g. shame.)

It seems to me that a better priming method-- if you're going to do it at all-- would simply be to have them pretend to be powerful or not;  a king, or a slave. 

It's also worth noting that 70% of the college student subjects were women.  Discuss.


Next comes the drawing of the E. You should probably sit down for this:

We used a procedure created by Hass (1984) in which participants are asked to draw an E on their foreheads. One way to complete the task is to draw an E as though one is reading it oneself, which leads to a backward and illegible E from the perspective of another person. The other way to approach the task is to draw the E as though another person is reading it, which leads to production of an E that is backward to oneself. We predicted that participants in the high-power condition, compared with those in the low-power condition, would be more likely to draw the E in the self-oriented direction, indicating a lesser tendency to spontaneously adopt another person's perspective.

drawing the E.JPG
Jean-Luc Picard on the left would be, therefore, in the high power condition, and Hendrick Hertzberg on the right would be low power.

Which is fine, except: why does the direction of an E have anything to do with being self-oriented?   You'll observe Picard is left handed, and Hertzberg is right handed.  Think about that.  Are there other reasons to think this E reflects self orientation and not, for example, the ability to write backwards ("power-primed are more dexterous?" "Power primed people draw horizontal lines from left to right?" )  or any of a billion other possibilities?  In other words, just because the group is called "high power," doesn't mean it has anything at all to do with power.

So while the results of the study were that the "high power group" was three times more likely to draw the E backwards, is there any conceivable way we can infer that actually powerful people do it, or that this has to do with power at all, or self-orientation?  The New Yorker cleverly adds a distraction:

(Or: B = -1.51, SE = 0.76, prep = .88.)

Unexplained, just sitting there, as if to say, "oh, look, we don't understand all that nerdy stuff, just accept that science says it's true."

But is there any validity to it at all?  No.  and not just because it's ludicrous.

The authors cite Hass as the originator of the method, and as evidence for the validity of the method.  Hass believed that people who were self-focused looked at themselves by occupying an external perspective, looking back.  Read that again.  In other words, a legible to the external observer E meant you were more self focused, not less.  It doesn't measure the ability to adopt someone else's perspective: it measures your ability to see yourself.

So let's review: Hass has a theory that is entirely untestable, it is then used backwards in  Psychological Science to draw conclusions about something different, and then applied wrongly for The New Yorker.  Awesome.  Science is relative anyway, and the real point is  Wolfowitz is still a jerk, right?


But let's look not at the protocol, not even at the premises, but at the meaning of the words.  Since the power level is hardly quantifiable, could the study be about something else-- that would make more sense of the results?

...recent views that power can lead to objectification, the tendency to view other people only in terms of qualities that serve one's personal goals and interests, while failing to consider those features of others that define their humanity... [making] it easier for a power holder to use them as tools in the service of his or her goals...

I don't know about power; but it is definitionally narcissism.  Replace the word "power" with narcissist and it becomes clear:

the high-power individual's self-concept remains more rigid; individuals with more power in their marriages resist the identities imposed on them by their less powerful spouses, and when relationship partners become more emotionally similar, it is the lower-power partner who has done the majority of the adapting...

The powerful, on the one hand, are less accurate than the powerless in estimating the interests and positions of other people and are more likely to make self-serving attributions. People with more power form less complex interpersonal impressions than people without power, basing their impressions of others on expectancies and stereotypes.

The problem is the conflation of power with narcissism.  Not all-- not even the majority-- of narcissists are powerful.  And not all powerful people are narcissists.

You might think I'm splitting hairs, but the authors themselves define power differently than they actually use it:

Power is often defined as the capacity to influence other people; it emerges from control over valuable resources and the ability to administer rewards and punishments.
Oh well.


It should be obvious that Hass's version is just as empty and meaningless as the Psychological Science version.   Hass's study is an experiment testing a hypothesis based on articles of faith, a slightly distorted version of the famous logical fallacy:

witches float
Lisa is a witch

therefore, Lisa should float.

Never mind that Lisa is really Arthur and he knows how to swim.  The authors of the Psychological Science study believe it, so let's grant them their premises are indeed facts.  But they then go on to devise an experiment that actually tests the opposite of the (preposterous) hypothesis.

Published science this bad rarely happens by accident, it requires planning and execution and is thus inevitably associated with having an agenda.  A quick look at the authors' bios reveals the agenda: power corrupts absolutely, and absolute power is worse.


But back to Wolfowitz.  Let's predict the outcome.  He's probably self-focused? So according to Hass, he should draw the E legible to observers.  But according to Psychological Science, powerful Wolfowitz should write the E backwards to outside observers, because he doesn't take other's perspectives.  (Either way, it proves he's a jerk.)

Before I tell you what he did, I'll give you The New Yorker's explanation of the results, which includes the words "bar," "alone," "deposed," "shaky," and "hesitant"-- all in the same sentence; followed by a completely made up explanation that allows you to be right even when you're wrong:

He apparently was not demonstrating the "carryover effect," by which a person who has been stripped of influence continues to behave as if he were still in power.
Thank God for science.

The pretense of The New Yorker is that they, by virtue of not being powerful, must be smarter than Wolfowitz, who is a priori an idiot.

Well, is it possible that Wolfowitz sees a set up?  Maybe he's more clever than you credit him?

Wolfowitz drew his E lowercase.


It seemed odd that, in a room full of powerful people, no one was acting the way powerful people are supposed to.
Oh, I don't know, given that there are so many things wrong with that sentence alone, it's not so odd.

[One of the authors of the PS paper] attributed the anomalously unpowerful results of the Time 100 trial to the use of blue Post-its [stuck to their foreheads, on which they drew the E]. "Blue creates a negative set of emotions, and when people are experiencing these emotions they think more deliberately," he said. "If you had used red, it would have gotten at their more spontaneous inclinations."

Got that?  But Wolfowitz is still a jerk, right?