Posted by: captainfalcon | December 30, 2009

Just So(min) War

There’s an intramural dispute at the Volokh Conspiracy that highlights a couple infelicitous theoretical tendencies.

Here’s a brief recapitulation of the contretemps. First, Ilya Somin claimed that politicians can up their chances of re-election by promoting ineffectual airport security measures. Orin Kerr responded that, in his experience (and he has some), politicians generally try “to do the right thing” in this area. Somin’s rejoinder is that this cannot be because while politicians would “certainly try to do [the right thing] if it were costless for them. The problem is that successful politicians are unlikely to prioritize ‘doing the right thing’ above staying in power.”

To begin with the bottom line: Somin’s willingness to dismiss another’s first-hand observations on the basis of his pet account of human psychology is another example of his irritating complacency. It is more or less axiomatic for Somin that political behavior is entirely explained by (i) rational ignorance (please stop posting about that, by the way) and (ii) the “fact” that people will never act contrary to their own perceived interests (which consist in money and power) unless it is “costless for them” to do so. What makes Somin’s (rather fatuous) worldview irritatingly complacent is that he does hold it “more or less axiomatically;” that is he declines to modify it in light of new evidence. (The current fracas is case in point: Kerr’s observations contradict Somin’s theory, and for that reason Kerr is mistaken.)

Somin’s criticisms not only reveal a flaw in his character, they are also misplaced. First, they are based on a false account of human psychology (one that, in my experience, has too much suasion among libertarians). Second, the false account of human psychology on which they are based doesn’t even have sufficient predictive power to support them (so even granting Somin his impoverished view of human nature, his conclusions about the type of airport security measures politicians will implement do not follow).

1) False account of human psychology. Do successful politicians care deeply about staying in power? Sure. Do we have reason to think they’ll invariably decline to do anything that hurts to any degree their chances of reelection (I gather this principle lies behind Somin’s view that politicians will decline to do the right thing unless it is costless for them to do so)? Absolutely not. It is well documented that humans in general will act contrary to their own interests (spite and altruism are real phenomena, as is the feeling of moral compulsion), and (prima facie paradoxically) there are good evolutionary accounts why this is so. Absent some reason to think politicians are different, we’ve no reason to expect them to sacrifice (their conception of) the good for their own benefit any more frequently than the rest of us do. (How frequently is that? is a good and, pace Somin, open question.) And absent Somin’s bothering even to wrestle with the kind of literature I’ve alluded to, we’ve no reason (other than soft-headed cynicism) to prefer his account of human psychology to the alternatives.

2) Lack of predictive power. But let’s say Somin is right about the psychology of our politicians. Let’s say politicians will only do what they think is right if they judge doing so is costless to them. Even granting that, Somin’s conclusions do not follow. First, most individual, discrete acts by politicians have zero bearing on their electability. While Somin’s theory should be able to predict patterns of acts, it cannot predict what particular acts a politician will perform. But even the theory’s power to predict act-patterns will be limited; there are many different act-patterns that can enhance a politician’s electability – from demanding outlandish security measures (to help stop the terrorists) to criticizing them for being outlandish (and therefore distracting us from stopping the terrorists). In short, Somin’s theory can at best predict disjunctions of act-patterns, not what politicians will do in response to a particular crisis.

I’ve now surveyed the problems with Somin’s theory: it is wrong and impotent, and he has a lousy attitude towards it. That said, it has heuristic value inasmuch as it highlights two infelicities that afflict many people’s approaches to political behavior. I’ve already alluded to one – the tendency to axiomatically accept a simplistic account of human motivation. The second is the tendency to act as if any theory purporting to explain e.g. political behavior thereby purports to explain every discrete political act. (Often the problem is compounded by an ambiguity in how the theory is presented, as e.g. an attempt “to explain political behavior,” which does suggest it aims to explain all political behavior.)

Both should be avoided.

Update: Three days later, Orin Kerr notices Somin’s regrettable epistemological predilections:

I have an Oakeshottian instinct to root ideas as much as possible in experience and empirical evidence, and I’m skeptical of how much broad theories and first principles can really tell us about the world. In contrast, you tend to reason first from broad theories and first principles, and you seem somewhat more skeptical about what experience and empirical evidence tells us.



  1. The primary difference between a successful politician and a non-politician is that the politician won an election. Presumably, actions that make one more likely to win an election also make one more likely to be a successful politician. So it would be reasonable to expect a higher-than-normal degree of “self-serving” actions (enlightened and/or unintentional though they might be) by successful politicians. Whether that gives any sort of predictive insight, I don’t know.

    It could be worse. We had centuries of economic theory which assumed that people are rational consumers. That’s about as accurate as saying all politicians are entirely self-serving.

    Which brings up an odd thought. If you follow any one consumer, you find all sorts of irrationalities in their decisions, such that a large portion of them are basically unpredictable from a rational-consumer standpoint. But the rational-consumer model actually does a reasonable job of predicting the actions of large groups of consumers, as long as you don’t expect too much out of it. In the same vein, would you predict the actions of a legislature more accurately than of a governor with Somin’s approach?

  2. “The primary difference between a successful politician and a non-politician is that the politician won an election…So it would be reasonable to expect a higher-than-normal degree of “self-serving” actions (enlightened and/or unintentional though they might be) by successful politicians.”

    1. On the other hand, the difference between a successful politician and a SUCCESSFUL non-politician is that one has had electoral success and the other has had success e.g. securing promotion in his chosen field. Why should we think the former requires more self-serving actions than the latter? And if it doesn’t, are we prepared to say that successful people across the board (BOCTAOE) are abnormally self-serving?

    2. Even granting that politicians are on average different from the rest of us (more self serving) as a matter of degree, to get any predictive purchase out of that requires considerable investigation into (a) how self serving we are (and how our tendency to be self-serving vary across circumstances) and (b) how significantly (and in what circumstances) politicians depart from our standard. So, given that human psychology isn’t as Somin says it is, the hypothesis that politicians are on the whole more self-serving by itself tells us nothing about what substantive actions they will perform. Instead you’d need to hypothesize that politicians have psychologies that are different IN KIND from the rest of us (that’s one way Somin could inoculate his theory against the sources I cite, albeit it would open him to a different set of criticisms).

  3. […] of politics (interest groups versus each other). It was also, needless to say, developed before (damn you, Somin!) rational ignorance was a recognized phenomenon. Both of these developments nuke the plausibility […]

  4. […] contempt for Ilya Somin is already on record. This is particularly egregious. Regarding some toothless boilerplate from the Kochs about their […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: