What Economists (and Epidemiologists) can learn from an Online Game Epidemic

I find this absolutely fascinating.

Epidemiologists used a virtual epidemic in the World of Warcraft online game to study the behaviour of people in cases of epidemics. They say that the behaviour of the 'virtual' people is quite close to the real behaviour: from the altruistic, trying to help others while exposing themselves, to the jerk who tries to contaminate as many people before he dies.

One reason why the epidemiologist interviewed in the article was excited, is because people studying disease dynamics are limited to 'observational and retrospective studies' - because you can't launch an epidemic in a human population just to study people's reaction. (yes, you can't, really.)

Hmmm. Don't we have the same problem in economics?
The joke 'an economist is someone who will be able to tell you tomorrow why what he predicted yesterday did not occur today' exists for a reason: we, too, function on mathematical models for predictions, and retrospective studies for explanation (and further predictions).

We can't decide to create hyperinflation in a country or devaluate its currency, for the sake of experimentation. We can't decide to fully liberalise a country's economy and foreign trade to see the effect on growth - that's why we have CGEs and Gravity Models and stuff.
But that's also why some theories that seem outlandish managed to survive from the 17th century until today.

So what if we could actually model economics - in a virtual world? Imagine a SimCity, 2nd Life or whatever (or World of Warcraft! :) where people actually take economic decisions - production, consumption, trade, and so on. Next, do something - a drought, free trade, an oil shock, whatever your heart fancies. And see people's behaviour.

Sure, it won't be optimal - but i really think the idea deserves some digging. It could totally be a yearlong project in a university.

Anyone out there looking for a dissertation topic? :)