Saturday, September 29, 2012

LA Cars and Santa Fe Bars: Carmageddon 2 and the El Farol Problem

Think fast: what do the freeways of Los Angeles have in common with a bar in far-away Santa Fe, New Mexico? This weekend, it turns out, quite a lot.

Today and tomorrow, drivers in Los Angeles face Carmageddon 2, a nascent traffic nightmare stemming from the closure of the 405 freeway for construction. During the original Carmageddon a year ago, Angelenos heeded officials’ dire warnings and mostly stayed home. That led to much speculation this year — would drivers stay home again, or would last year’s smooth sailing tempt them back into their cars?*

What was that about some bar in the Land of Enchantment?

What L.A. drivers face this weekend is called the El Farol Problem, named for a bar down the street from the old offices of the Santa Fe Institute, or SFI. Here’s the problem in a nutshell. You want to go to a cool bar — or go driving on the streets of Los Angeles — but so does everyone else in town. If everybody heads out, it’ll be crowded, noisier than usual, and generally no fun — the bar, I mean. But if nobody goes, you and your friends will have the whole bar to yourself, and you’ll have a grand old time.

When SFI economist Brian Arthur cooked up the problem, he wasn’t interested in the conventional explanation of the problem, which is simple, if implausible — the standard prediction is that everyone decides completely at random.

Instead, Arthur wanted to ponder how real people would think through the problem, and that’s where Carmageddon 2 comes in. Will Angelenos will stay home in fear, or will they learn, so to speak, from last year and venture out to what they anticipate will be empty streets? Will some drive as an act of defiance? Will anyone actually roll dice? Probably all of the above, but the truth is, we still don’t really know.

Like many SFI social scientists, Arthur thought conventional economics swept the interesting details of decision making under the rug, favoring instead elegant but sometimes inaccurate theories based on the assertion that people are rational decision makers. The unfortunate truth is that, while we have made progress, we still don’t understand how people decide well enough — not well enough, for example, to understand who will vote in the upcoming election, and not well enough to prevent another economic crisis like the one we had in 2008.

Nor, I’m afraid, well enough to say how long it’ll take to get to Santa Monica tomorrow. Good luck with that one.

*Actually, things are looking okay for the moment, despite a glitch in the construction timeline.