SuperFreakonomics the book is consistently entertaining, sometimes fascinating, genuinely thought-provoking, excessively glib, fundamentally flawed, and viscerally disturbing. Let me explain.

I’m glad the authors, Levitt & Dubner, address some core aspects of human behavior. Not the detailed analysis of sexual acts, gratuitously provided in great detail, but the ability of humans to engage in cooperative behavior: the prisoner’s dilemma. Because it is so fundamental, let me describe it briefly.

The Prisoner’s Dilemma: Two prisoners have a choice: to cooperate and stand up for each other, or to rat each other out. If both cooperate, both go free. If both rat, both stay in jail. But if one cooperates and the other rats, the cooperator is shot and the rat is freed with a fortune.

Note that the overall benefit is highest when both cooperate¹, yet for any given scenario each player’s best option is to rat. So the question of what causes real life people to cooperate rather than stab each other in the back, to me is fundamental and interesting, and has to some extent guided my own thoughts ever since I read a Scientific American article on it back in 1983. In that experiment, the key was to repeat the game. In a repeated game (using money rather than bullets) in which participants know and can react to the previous outcome, it becomes beneficial for both players to cooperate in any given cycle to maximize their overall benefit. Thus, transparency and accountability are key factors in causing people to behave in a cooperative manner and increasing the net benefit to all concerned.

The authors completely missed this key point, not even mentioning it in passing. Instead, they spent a great deal of time on variations of the ‘Dictator’ game, which shows that given a chance, humans will indeed screw each other over. The “discovery” that people will sometimes steal from each other when given the opportunity, and pretending that finding is somehow winning an argument, is not interesting or compelling. Now if they had really modeled the real world, and allowed some players with large amounts of winnings to change the rules to benefit themselves, to ‘cheat’ on a higher level, and had found a way to stop that from happening, that would be something.

We can disagree on the fundamental principles of human morality, however, and it is a plus to even bring up the subject. They also address other key and interesting points. The explanation of externalities is well done, including the perhaps unexpected logic that simply not preventing theft results in a cost to other property owners.

The stomach-twisting criticism I have, is the authors’ tendency to support one argument in impressive detail, establishing their credentials as the real deal, unbiased economists who are so clever — then step across the flimsiest of bridges and draw an unwarranted conclusion in another area, one whose ramifications might cause real suffering.

The first, most deeply disturbing, example of this leap in logic: jumping from the fact that in the Dictator game, people often cheat each other, to the conclusion that therefore we will never get enough organ donors by volunteers and that we ought to allow individuals from poor countries to be brought to richer countries to donate their organs, paid something, then sent back. And on to the next idea.

Glibly promoting their point of view while implying that opponents are not rationally motivated, wouldn’t matter except for how widely read the book is, and that policy makers might really listen to these guys. If you are a best selling author and you’re going to propose something like paid organ donation, I think it behooves you to really look at the possible suffering and consequences. This is a proposal that could cause intense suffering — people being pressured into selling their organs as a way to survive, to help their family, maybe people being controlled by brokers — possibilities are horrific and should not be passed off as irrelevant. And again, the authors miss a key point of game theory. If an intelligent player can control the rules of the game to benefit them, they will. So putting the rules of the game in the control of the buyer, the rich country recipients of the organs nearly ensures that the game will be stacked in a way that does not protect the rights of the sellers.

And it also sets up a scenario in which a steady stream of impoverished organ-sellers is a necessity to others.

“Lawyers for Attorney General Kamala Harris had argued in court that if forced to release these inmates early, prisons would lose an important labor pool.” - LA Times This is in the court records. How much of a stretch is it from this to arguing that if some changes were made to help people impoverished areas emigrate or find work, that the medical field “would lose an important organ pool”? Perhaps less of a stretch than from a pretend game of “Dictator” with volunteer students.

The second example of an unwarranted and foolish ‘clever’ conclusion, is the criticism of car seats. In excited italics, the authors ask “what if seat belts were designed to fit them (children) in the first place?…or have fold down seat inserts?” and go on to take the EU to task for “requiring most children to stay in booster seats until they are twelve.” Um, booster seats ARE simple seat adapters, no? Booster seats already exist and are the sensible alternative to strappy car seats for older children. So the criticism in this case isn’t really harmful, but just foolish on the face of it, and an attempt to be self important and cleverer than those EU bureaucrats.

Having criticized so pointedly, I’d like to also give credit where it seems to me credit is due.

In the area of climate change, the authors are fairly rational and not kneejerk naysayers, and actually make the argument that given the real danger — they quote economist Martin Weitzman as estimating a five percent chance of a terrible case scenario to the earth — that we should seriously look into the possibilities of geoengineering. Here they go into more detail of what such engineering might look like, address the issue of whether it would be permanent, and at least attempt to address the idea of side effects. I am not deep enough in the field to evaluate whether the ‘Budyko’s Blanket’ idea is feasible or safe, but given the possibility of real danger to the planet I agree wholeheartedly with the authors that it should at least be investigated.

To Levitt and Dubner, I would say this. You have a platform, you understand at some level economics and logic. But most of your models are disproportionate and flawed, partly for a reason that you admit directly: lack of empathy. Empathy, ability to take into account others’ feelings, to place a value on not seeing others suffer, is a valuable trait for any economist who wishes to model the real world. Because a model that only comprehends financial and sexual incentives, thankfully does not successfully predict human behavior. (Though it comes close in the realm of prostitution, the field that features so prominently in your book.)

For what are surely valid evolutionary and meme-evolutionary reasons, humans are often driven by a desire for regard from their peers, and also by a drive to create something meaningful. Without considering these two factors as primary forces, it would be difficult to explain the huge Open Source software movement that powers most of the Internet, and likely runs the server on which this document is being kept. Certainly it should be possible to derive any force in human behavior from the fundamental principles of human and idea reproduction, but as a simple placeholder for biological and ideological fitness, money is insufficient. There is a fundamental reason for this, and it goes back to the prisoner’s dilemma: cooperation happens in long-term, transparent relationships.

Money is fully anonymous and by itself, liquid and entirely short term. To achieve cooperative behavior in the sense of overall benefit to society, it is necessary to have other instruments for connecting humans to each other. Any tool that enhances open, long-term but voluntary, substantial and meaningful relationships over time is likely to have a positive effect. Fortunately, humans have a natural bent for turning tools to this purpose, from telephones to the Internet.

Shared values in a community also have a strong impact on behavior. At the same time I read your book, I also read Enrique’s Journey, an account of the trials of a young Honduran man jumping trains to come north to find his mother. En route, he experiences both extremes of human behavior. In Chiapas, he is set upon by gangs of thieves, beaten and nearly killed. In Veracruz, poor families who live near the tracks throw bags of food onto the trains to help the desperate migrants.

“The towns of Encinar, Fortin de las Flores, Cuichapa, and Presidio are particularly known for their kindness. A young Honduran migrant … points at the ground in Veracruz. “Here”, he says, “the people are good. Here, everyone gives.”….[another migrant] Rodas, his voice cracking with emotion, says “We could not keep going forward without people like this. These people give you things. In Chiapas, they take things away.” ²

When Levitt and Dubner publish a book that can explain this difference, and how to structure a society or maintain a culture like that of Veracruz, and how to change one like Chiapas, that will be a book worth reading.

¹ To make the overall higher value of cooperation explicit in a game scenario, just replace freedom with $3, staying in jail with $1, a fortune with $5, and being shot with $0. This was the actual method used in Douglas Hofstadter’s May 1983 Scientific American Metamagical Themas column that discussed the prisoner’s dilemma at length in relation to game theory, complete with a simple experiment.
² Enrique’s Journey, p 104–105