Here’s?another hilarious xtranormal send-up – this time lampooning the Becker/Murphy theory of rational addiction:
(HT: Peter Siegelman)
Addicted to Rationality
Naked Self-Promotion
Keep up with the latest Freakonomics news and chatter at Big Buzz. And here's more:
- What to feed your brain and ease your commute? Business Insider says Freakonomics Radio!
- Stephen Dubner talks with Dan Schawbel at Forbes about the Freakonomics phenomenon.
- Freakonomics: The Movie is now on Showtime! It is also available on DVD
and Blu-ray
. You can also watch it on iTunes or Netflix.
- Freakonomics tops Forbes' list of books that rock the "New Author Platform".
- Daily Beast Q&A with Dubner & Levitt on SuperFreakonomics paperback edition.
- NY Mag on Levitt’s new mind-for-hire gig.
- Econ profs rank Levitt as their fourth favorite economist under age 60.
- TIME has named the Freakonomics blog one of "The 25 Best Financial Blogs."
- Freakonomics Radio has hit the air! Find out when our second season of five hour-long specials will air on your local public radio station. You can download/subscribe at iTunes, get the RSS feed, or grab our widget.
Read Thomas Schelling on addiction.
The amount of ‘tail wagging the dog’ in prediction models is disturbingly rampant in economic and ultra-numerical philosophical argumentation. If the result of a model is contradicted by centuries of experience across billions of people, that is asymptotically equivalent to a proof by contradiction, showing the model to be wrong. In that case, the model needs to be thrown out or revised substantially.
If a weather model fails to predict the weather, we revise it. We do not sit back and marvel at an amazing model that predicts an Earth perfectly that, while complex, is 10^12 times less complex than the one we live on.
Here, the connotation for the lay reader of ‘rational’ is wildly divergent from the anemic definition used in economics which involves some equation involving an integral and the value of fixed preferences, wherein the pleasure of all experiences in life are reduced to comparable numeric values.
Models of rationality fail to model the greater interests of group dynamics. Any decent model of morality would show that when a rational actor makes certain types of decisions the only moral choice is to coerce non-rational behavior from that individual. (Does it feel good, even best, to kill? Too bad.) When we talk of rational acts we implicitly imagine group dynamics whether we realize it or not. As well as other sentient beings, the welfare of the Earth (at least measured locally) may be a member of this group. The math becomes so monstrous that it dwarfs the current state of the art. Economists aren’t admitting that they’re working on coherent if tiny building blocks of a ten-thousand piece puzzle here. If you’re dubious about the math, just think about the non-linear dynamics that are necessarily involved; consider the inexpressibility in closed form of the solution of the three-body problem in simple Newtonian gravitation. Just as an off-the-cuff example, consider that the fixed-preferences model may be absurd in a world that is full of feedback; where we adjust our preferences based on the observed result of our earlier actions.
Free markets fail to work in surprising ways; simple single-candidate yes/no democratic voting can produce disastrous results; models that Fit the World are Scary and economists, even with modern genius-level work, are barely getting started, just as meteorologists are.
Addiction is a case where local optimization pins one to a badly deteriorating situation. The addict begins by taking the drug to feel good, and then continues even after it becomes the main cause of his pain, because stopping it feels even worse.