Vernon Smith is a Nobel Prize winner. You quickly realize on reading this book that he got it for economics, not literature. But if you can slog through this book (which took me about 4 months), you will be rewarded with some great insights about development. (But why I am working so hard when Tyler Cowen’s blog is about topless French sun-bathers?)
His big picture is familiar to readers of Hayek: societies develop NOT through the conscious design of some experts (Smith uses the horrible jargon “constructivist” for the design view), but through the “ecological” survival of institutions, norms, rules, firms, and products in a society of freely choosing individuals.
It gets more novel when Smith applies a similar insight to individual rationality. Unlike most of his experimentalist colleagues, Smith is not celebrating findings of human irrationality in the lab as the greatest thing to come along since Sacha Baron Cohen. Instead, Smith wants to look more deeply at “irrationality” and see if in some cases it might be rational after all. The big idea is similar to the above: as humans biologically and culturally evolved, unconscious ways of acting “rationally” passed “ecological” tests of contributing to survival and well-being. The standard “rationality” model that looks at only explicit constraints and logical reasoning (“constructivist”) is way too simplistic for Smith.
An example is the well-known two-player game, the prisoner’s dilemma, when each player gets a higher payoff by cheating if the other doesn’t. The payoff is still very high if neither cheat, and it is the lowest if both cheat. Rational behavior predicts that both players cheat and hence wind up with the lowest payoff. Yet laboratory experiments with real human subjects and real money find that both refrain from cheating surprisingly often.
So players are behaving “irrationally,” yet Smith points out that they have managed to get a better outcome than what “rational” behavior would achieve. He argues that players have unconscious social norms of “fair” behavior (and also they may find ways of “socially” signaling to each other these norms, since one thing we know about humans is that their social skills are highly advanced). Unconscious sociability allows humans to realize gains from social exchange that cannot be captured by the explicit “rational decision” model. He finds more evidence for this idea by subtle variations in the social context of the experiment.
Smith doesn’t address development differences, so the big question is why gains from social exchange are realized more in some societies than others. Maybe he will get to that in his next unreadable insightful book.
12 Comments
We have a bias toward coherent actions?
Somewhere i evolution the certainty of actions must have been slightly modified into the shared certainty of action. Our bias toward action would be enhanced, more than linearly, if nearby members signaled agreement.
We would have a bias toward keeping the factory open a bit longer than its time. We would have a bias toward skipping a step in the production process during good times, and regret its removal during periods of change. We would have a greater than predicted value of money as a collective signaller.
Here’s an interesting, related article by Cosmides & Tooby → http://www.jstor.org/pss/2117853
Does this suggest that the whole model of “rational decision making” should be scrapped? Or is it sufficient to note exceptions (that is, we act “predictably irrational” (Dan Ariely))?
From your summary, I really cannot see anything different from Axelrod’s (1982!) thesis.
Where does religion sit on this. I once saw it described as rational irrationality; people, often well educated, believing in something with a high degree of certainty that cannot be proved.
What about witchcraft? Where I am currently living many people are fearful of “juju”, so much so that they do not report crimes to the police. Recently otherwise ‘rational’ men were ordered to burn a village down by the witch doctor, which they subsequently did. They are now in jail.
Is there irrationality in the book that is not only irrational but also not beneficial?
Personally I think all actions can be given rational explanations after the event, especially if they have a beneficial outcome.
Except perhaps this…
http://www.stabroeknews.com/2009/news/local/07/23/albouystown-man-assaulted-cop-after-being-found-penetrating-dog-%E2%80%93-court-hears/
Any rational explanation? Sorry for sullying your blog Dr. Easterly.
@John:
I think what professor Easterly is referring to particularly are “irrational measures” that allow human beings to “pass “ecological” tests of contributing to survival and well-being”
Clearly, the prevalence of religion in developing countries is often trumped up by misguided social engineering saavy atheists/agnostic as the primary hindrance to development. However, I think while there are issues with religion and development in developing countries, I think there are many constructive ways faith has played a concrete role in development.
One of the most profound ways is the ways that faith communities have helped in poverty alleviation efforts and provided basic social services in the absence of an effective government. The people’s faith might seem irrational but it does something for them–it feeds them and helps them survive. The belief in things like juju also helps to prevent crime or nasty behavior as people realize that there is something that can hold them to account in the absence of accountable community policing.
Clearly, there are cases where this irrationality does not produce the best results-like those you mentioned but it is easy to think all cases are like this when in fact they are not. There are constructive wasy irrationality has contributed to development which I would mention but for limited space and time.
In the areas where there is what I would term “destructive irrationality”, it eventually dies out because it does not contribute to their survival and well being. They are more or less inclined to dump it over time. There are many examples of how this has happened in developing countries that I would mention but for time and space.
In conclusion, I really like this effort by Dr Easterly. Little wonder he is such a great mind in development. However, I would give him unsolicited advice if he intends to dabble into this work of examining “irrational approaches to development”..he should involve local people. Lord knows they have wisdom for development that we too often underestimate.
In the prisoner’s dilemma defection appears to be rational. However, mutual cooperation beats mutual defection, so there is something wrong. The rationality of defection depends upon reasoning by cases. Reasoning by cases seems unassailable, yet there are well know examples where it fails: Simpson’s paradox in statistics and the double slit experiment in quantum mechanics. In all three areas there are interactions that overcome the apparent rationality of reasoning by cases. The cases are not independent.
Yes, Alan, there is no Santa Claus.
If entrepreneurs were rational most of them would not take the risks they do, but if it weren’t for the risks they take, we’d all be worse off.
Also, in an world with imperfect information, it is often better to move forward with limited and wrong information and adjust as lessons are learned, then to wait to gain more information.
In short:
‘rational people’ in many instances achieve suboptimal outcomes — both for themselves and society….
Thanks for this review. I just ordered the book. This looks to me like the future.
Not sure that this is entirely new.
The key difference between the Prisoner’s Dliemma and human interaction is that the latter (usually) involves repeated iterations.
Indeed, in some of the earliest computer competitions of trying to “solve” the PD, involving repeat iterations, the winning strategy was simply to do to the other whatever they had done to you in the previous round.
If they cheated you then cheat back (ie, punish them). If they cooperated then cooperate. Pretty quickly settles down to a “cooperate, cooperate, cooperate” pattern.
Which is pretty much what human societies do really…..
Vernon Smith did not get his Nobel for literature but he took three past and present wives to the festivities in Stockholm:) In your face Bertand Russell, J P Sartre and Paul Samuelson!
Jokes aside, it sounds as though Smith is on the same track as Bowles and Gintis who coined the idea “homo reciprocans” who has evolved tendencies to exchange in a win/win manner. http://bostonreview.net/BR23.6/bowles.html
Of course anti-social actions are not precluded and they suggest that the fabric of polite society becomes strained when a “tipping point” (maybe 25% deviance from reciprocity) is reached.
That evolutionary process may account for the famous “a prioris” of human action that feature in the von Mises account of human action and exchange. He scandalised many people but that theory only needs some tweaking to come into line with the evolutionary epistemology of Popper, Piaget, Campbell et al.
It seems more like, to me, that purposeful action is BY DEFINITION purposeful. Non-purposeful action does not fall under the domain of economics or praxeology. For example, value itself is non-rational, non-purposeful and pre-praxeological. I do not think one requires ‘evolution’ at all to explain Mises’ view of human action or economc calculation, one instead should look towards the philosophical tradition of subjectivism (Bergson, Stirner), especially if it is regarded as an ontological (not EPISTEMIC issue, contra Hayek).