We are reading it in my Behavioral Economics class and so far we have finished the first 5 chapters which make up Part I of the book “Anticipating Future Preferences.” In Ran Spiegler’s typical style, perfectly crafted simple models are used to illustrate deep ideas that lie at the heart of existing frontier research and, no doubt, future research this book is bound to inspire.
A nod also has to go to Kfir Eliaz who is Rani’s longtime collaborator on many of the papers that preceded this book. Indeed, in a better world they would form a band. It would be a early ’90s geek-rock band like They Might Be Giants or whichever band it was that did The Sweater Song. I hereby name their band Hasty Belgium. (Names of other bands here.)
Many of the examples in the book are referred to as “close variations of” or “free variations of” papers in the literature. And Rani has even written a paper that he calls “a cover version of” a paper by Heidhues and Koszegi. So to continue the metaphor, I offer here some liner notes for the book.
In chapter 5 there is a fantastic distillation of a model due to Michael Grubb that explains Netflix pricing. Conventional models of price discrimination cannot explain three-part tariffs: a membership fee, a low initial per-unit price, and then a high per-unit price that kicks in above some threshold quantity. (Netflix is the extreme case where the initial price per movie is zero, and above some number the price is infinite.) Rani constructs the simplest and clearest possible model to show how such a pricing system is the optimal way to take advantage of consumers who are over-confident in their beliefs about their future demand.
A conventional approach to pricing would be to set price equal to marginal cost, thereby incentivizing the consumer to demand the efficient quantity, and then adding on a membership fee that extracts all of his surplus. You can think of this as the Blockbuster model. The Netflix model by contrast reduces the per-unit price to zero (up to some monthly allotment) but raises the membership fee.
Here’s how that increases profits. Many of us mistakenly think we will watch lots of movies. Netflix re-arranges the pricing structure so that the total amount we expect to pay when we watch all of those movies is the same as in the Blockbuster model. Just now we are paying it all in the form of a membership fee. If it turns out that we watch as many movies as we anticipated, we are no better or worse off and neither is Netflix.
But in fact most of us discover that we are always too busy to watch movies. In the Blockbuster system when that happens we don’t watch movies and so we don’t pay per-unit prices and we Blockbuster doesn’t make much money. In the Netflix system it doesn’t matter how many movies we watch, because we already paid.
My only complaint about the book is the title. (Not for those reasons, no.) The term “Bounded Rationality” has fallen out of favor and for good reason. It’s pejorative and it doesn’t really mean anything. A more contemporary title would have been Behavioral Industrial Organization. Now I agree that “Behavioral” is at least as meaningless as “Bounded Rationality.” Indeed it has even less meaning. But that’s a virtue because we don’t have any good word for whatever “Bounded Rationality” and “Behavioral” are supposed to mean. So I prefer a word that has no meaning at all than “Bounded Rationality” which suggests a meaning that is misplaced.