Online Reality

The Law of Demand must be hidden      here somewhere

Experimental verification in social sciences is maddeningly difficult: imagine, for a second, that you manage to convince a whole society to go around its business under carefully controlled conditions, in order to test an economic theory. As a control, you convince a second, equal-sized society to do likewise, this time with a critical variable altered.

And then, at the end—well, this is all experimental, you see—your experiment proves inconclusive. You have no alternative but to tweak a couple of variables here and there and run it again in order to see if anything changes. So, please, folks, we'll start all over again, OK? We go back to square one. Now move.

Splitting protons or tailoring bacteria is kid's play by comparison.

So, historically, hard sciences have tended to look down the nose upon the so-called soft ones, such as economics or sociology: not scientific enough, they sniff. But wait a second: for quite a while now we have been creating virtual worlds populated by millions of people who interact with each other on a daily basis. Such worlds span the globe. They are long-lasting. They have a currency and social conventions. They are hugely popular. Examples of these synthetic worlds, which form the stage for what is known as "massively multiplayer online role playing games", are World of Warcraft or EverQuest. Couldn't we run our social experiments there?

Edward Castronova, the acknowledged authority in the realm where virtual worlds intersect our real one, thinks this could well be the case. And the findings he reports in his latest CESifo Working Paper seem to support his thesis.

The paper describes the experiment he and his collaborators ran to elucidate a nagging small matter that has to be cleared up before virtual worlds qualify as economics laboratories: does virtual-world behaviour mimic real-world behaviour? After all, for most players their participation in the virtual world is "just a game", a stage on which they can experience things that are not possible in the real world. It could well be that when a person is an elf or an orc, the value of money, for one, would be perceived quite differently compared to the real world: in the virtual one it does not matter whether a person has enough money to "live" on, for instance, or whether certain activities are completed on schedule.

So Mr Castronova and his team, with the support of the MacArthur Foundation and the makers of the Neverwinter Nights videogame and the Aurora virtual-world-creation toolkit, created the game Arden II: London Burning, set it up in two exactly equivalent worlds, which differed solely on the price of a single good (50% cheaper in one of them than in the other), assigned a total of 43 players randomly to either world, and set out to test whether economics' Law of Demand holds in cyberspace. In other words, they wanted to test whether human behaviour in such worlds is economically "normal".

Well, it turns out that it is. Just as the Law of Demand predicts, the higher price in one world of the good in question—a health potion— drove demand down by 43.1 percent there, reflecting a price elasticity well within the bounds of what is expected for normal economic agents.

This, in turn, suggests that fantasy gamers may well be economically normal. This finding adds weight to Mr Castronova's observations in a previous experiment that showed that co-ordination game theory holds its water in online worlds (read article here).

On top of that, online games have served as a venue to study the spread of a virtual epidemic: a couple of years ago, a virtual disease called "Corrupted Blood" was accidentally let loose in World of Warcraft. The game's administrators had introduced the disease as a challenge for some high-level players. What they didn't expect was that it would break out of the caves and into the virtual world's cities and towns. Despite all efforts to quarantine the infected, the disease ravaged the player population and gave World of Warcraft its first virtual-world pandemic.The way it spread throughout the synthetic world, by stupidity as well as pets and instant travel, provided invaluable real-world insights to the researchers who were alerted of its existence before the game makers decided to wipe out the corresponding malignant computer code.

It thus appears that these huge worlds can be used to study human behavior on a vast scale. The Castronova study and other examples suggest that the presence of fantastical creatures and odd roles within these environments need not invalidate them as research sites. It may be true that players enter a game world only because they want to hunt dragons; but if they need health potions to do this, it can be expected that the market for those potions will be a pretty normal market, with ordinary supply and demand curves and ordinary equilibrium conditions.

So, who said doing research in economics wasn't much fun?

Edward Castronova: A Test of the Law of Demand in a Virtual World: Exploring the Petri Dish Approach to Social Science, CESifo Working Paper No. 2355

• Other CESifo Working Papers dealing with synthetic worlds
• Other CESifo Working Papers by Edward Castronova
• Download game environments used for Arden II: London Burning
• A review of Mr Castronova's book "Synthetic Worlds: The Business and Culture of Online Games"
• Edward Castronova's website
• CESifo Working Papers published in the past three months