Legal Commentary

I just read a really great paper by Jonathan Berk and Eric Hughson called "Can Boundedly Rational Agents Make Optimal Decisions? A Natural Experiment."  The Berk and Hughson examine a sample of games on the game show The Price is Right to see whether contestants make optimal bidding decisions. 

The two games they consider are Contestants' Row (where you bid on the price of the prizes - whoever gets closest to the price without going over wins).  The second is the Wheel Spin - whoever gets closest to 100 wins.   You must win on Contestants' Row to get to the Wheel Spin.

Their results are interesting - they find that a large percentage (nearly 50%) of relevant bidders on Contestants' Row bid suboptimally - in other words, they don't bid $1 over one of the other bidders.  This result is interesting, because that strategy is pretty easy to figure out.  At the same time, about 95% of Wheel Spinners spun optimally, despite the fact that the Wheel Spin decision is a much more complex decision.  In otherwords, a large portion of contestants were irrational in the easy game and highly rational in the difficult game.

Some comments on the implications and on the reasons follow the jump.

The implications of the study are important for  the study of behavioral economics (or law & economics).  If bounded rationality is divorced from the difficulty of the decision, it makes it much more difficult to model behavior in any predictable manner.  I suppose that the non-economist answer to that is "Duh!"

Berk and Hughson discuss possible reasons for the phenomenon they observe, such as learned behavior (both in the course of the game and from watching on television) and irrational focus on the prize value in the bidding game.

I have a couple of comments that they don't address.  First, while the discuss the Wheel Spin bonus if you hit exactly 100 and dismiss it as something that doesn't affect optimal behavior, they don't discuss the $500 bonus one wins if he or she bids the exact prize value on Contestants' Row.  In other words, it may be optimal to bid more than $1 over the next closest bidder when shooting for the bonus.

This leads to a second possible explanation, though I'm not sure how I would categorize it.  Contestant's Row is a guessing game that involves human estimation skills.  As a result, participants (at least some of them) may have faith in their estimation prowess in hitting the exact number.  As a result, they may not view these suboptimal decisions as irrational; instead, they believe they can find the right answer.

The Wheel Spin, however is all random and everyone knows it.  Contestants have no illusion that their special spinning skills will somehow get them to 100.  I suspect that if any modicum of skill were involved in the Wheel Spin, we would see a lot more suboptimal behavior in people trying to go for the bonus.

In any event, this is an interesting short article worth a look.

H/T: Legal Theory Blog