Roy went on a behavioral economics kick some time last year, and at that time I at least made it through Nudge and Sway, leaving Predictably Irrational and a more academic book on the subject on the pile and collecting dust. Well, I finished Predictably Irrational today. The book does have a lot of interesting experiments providing insights that are sometimes amusing, often disturbing, and certainly worth examining for their ramifications. I was struck at times, though, by the somewhat naive solutions the author proposed as a way to deal with the world as it really is. I would tend to agree with his oft-repeated assertion that the model of the rational decision maker that is so central to traditional economic theory is inadequate to say the least. However, I am less convinced that he has a more realistic view on aggregate.
One example proposal of his illustrates my point here. He argues that, rather than dividing up the bill amongst all parties when a group dines together, the group should adopt the “one person pays and this rotates” strategy. While I can buy this for a very small group that often dines together and knows each other well, I think it breaks down pretty rapidly. Even his example of four payers to me seems too large to work effectively. While he acknowledges the risks of people moving away or the same people not coming or other problems with the idea, he still thinks it deserves more serious consideration. Perhaps he just isn’t the one who would normally feel compelled to pick up the check if no one made a move for it, even if I had paid the last time around.
I do also somewhat question how broadly one can reason from the conclusions of his experiments. While they are certainly well designed experiments, by their nature there is little opportunity for group behavior to emerge or for the effects of time to be taken into account. His group behavior is really more the sum of the individual behavior rather than there being much opportunity (with one exception) for individuals to influence the behavior of others in the group. To me, both of these characteristics play an important role in reasoning about behavior in economic systems. However, I think he makes several points about our tendency towards trust, fair play, dishonesty, procrastination and our unrealistic assessment of costs and benefits that more traditional economists should at least attempt to incorporate in their models.
Of course, one experiment that I thought had its original source in this book wasn’t there. Now I have to try and find it for someone somewhere else. Sigh.