It’s not always easy to do repeatable experiments in economics. But a few economists have been exploring the idea that virtual worlds might offer a good laboratory for doing such experiments. A recent article in the magazine of the Federal Reserve Bank of Richmond describes some of the work that’s going on.
The existence of such clear economic behavior has convinced Castronova [an ecoomist at Indiana University] that virtual worlds may — but don’t always — provide venues for economists to learn things about economic activity that they otherwise couldn’t. Traditionally, economists have relied on 1) theoretical models that require perhaps imprecise abstractions and assumptions about human behavior 2) statistical regressions of past economic activity, which may fall short because changing the rules of the game will probably mean changes in future behavior, rendering the lessons from the past moot, and 3) experiments with groups of people in random and control groups, which tend to suffer because of the small sample sizes and unrealistic environments.
[…]
“Given this level of control [in virtual worlds], an easy yet breathtakingly powerful research strategy almost immediately leaps to mind,†Castronova wrote in a 2005 paper. “Build several synthetic worlds in exactly the same way, except for some difference in a variable of interest … attract people into the worlds, sit back, and watch what happens.â€
But what about the fact that virtual worlds in which some crucial variable differs might attract different kinds of people, introducing all kinds of (real-world biology/sociology-based) uncontrollable variables?
Georg – Unsurprisingly, that question seems to be a quite significant source of contention. Having fundamental questions like that open is what makes something an interesting nascent research program.