More on "The Logic of Life"

OK, after reading the book through and through, and after some thinking, I find it less compelling.  Though its basic idea (that people respond to incentives in a rational manner) is intriguing, the book makes a mediocre job of justifying it.  The author first tells us to beware of contrived lab experiments (in Chapter 1, commenting on the difference in Tversky's and Lists's conclusions), but later happily relies on such experiments to buttress his claims (most egregiously in Chapter 8, with a classroom experiment at Texas A&M, but also in Chapter 6 with play-pretend hiring decisions made by students with little real-life HR experience).  This inconsistency seems unnecessary, as his arguments would be strong even without using the contrived experiments.

At other times, the book is only too content to leave a subject after testing just one possible explanation.  For example, if the fraction of joint papers by co-located authors increased between the 60s and the 90s, that is enough to support the conclusion that "communications technology also seems to stimulate more local collaboration"?  It couldn't be for any other reason?  And same with employers rejecting (fake) resumes with black-sounding names -- is racism really the only possible conclusion?  Could it be that standout names simply prompt more scrutiny, which then discovers the fakery?  I'm not saying the conclusions are wrong or that the original scientific works didn't properly explore alternate explanations.  But if they did, I wouldn't know it from reading this book.

Finally, the book ranges far afield in search of examples to confirm the rational-humans theory when a few obvious tests are close at hand.  In Chapter 3, it accepts as a given that most people marry within their race and age group.  But what is the rational explanation for this?  The book doesn't examine it.  It also doesn't examine the seemingly irrational strategy that humans routinely exhibit in the ultimatum game (even though chimps don't).  If the theory is any good, why not test it on these obvious examples?

This points to a larger flaw in the basic idea itself: it is not really a theory.  Because it relies on perceived incentives, not objectively real ones, it cannot make accurate predictions without knowing for sure how the average human perceives the incentives surrounding them.  But this is quite often impossible to know with accuracy.  So the predictions made on the basis of this idea will frequently prove false not because the idea itself is wrong but because our knowledge of human perceptions is flawed.  This makes the idea impossible to disprove: any time it mispredicts, it can simply blame it on the complexities of human perception.  By definition, an idea that cannot be disproved is not a theory (at least not a scientific one).

Attractive as it is to think that the invisible hand of reason guides the behavior of human groups (and clever as some of the experiments are), I'm still not sure I can let that thought guide my actions in life.