Valve's new economist-in-residence will publish notes on the political economy of games

[Wow. Could virtual worlds finally push economics into being a true science? Seems plausible to me. -egg]
Valve's new economist-in-residence will publish notes on the political economy of games:
Yanis Varoufakis, "an academic economist," recounts the story of how his widely read writings on the Euro crisis led to a job offer from Valve software, who were contemplating the creation of a shared virtual currency between two worlds. Valve founder Gabe Newell was contemplating balance of payments when he realized "this is Germany and Greece," and he wrote to Varoufakis to ask for his consulting help.

Varoufakis is now Valve's economist-in-residence, and he has set out "to forge narratives and empirical knowledge that (a) transcend the border separating the ‘real’ from the digital economies, and (b) bring together lessons from the political economy of our gamers’ economies and from studying Valve’s very special (and fascinating) internal management structure."

He will blog his findings on the brand new Valve Economics blog. The inaugural post tells the tale of his recruitment, and tantalizes with the prospect of more to come.

For the first time since I switched from mathematical statistics to economics (around 1982), I saw an opportunity for scientific research on some really existing (albeit digital) economy. For let’s face it: Econometrics is a travesty! While its heavy reliance on statistics often confuses us into believing that it is a form of applied statistics, in reality it resembles computerised astrology: a form of hocus pocus that seeks to improve its image by incorporating proper science’s methods, displays and processes. Is this not too harsh a judgment on econometrics?

Not in the slightest. Econometrics purports to test economic theories by statistical means. And yet what it ends up testing is whether some ‘reduced form’, an equation (or system of equations), that is consistent with one’s theory, is also consistent with the data. The problem of course is that the ‘reduced form’ under test can be shown to be consistent with an infinity of competing theories. Thus, econometrics can only pretend to discriminate between mutually contradictory theories. All it does is to discover empirical regularities lacking any causal meaning. To put it bluntly, it is impossible to avoid absurd conclusions such as “Christmas is explained by a prior increase in the demand for toys”. And when we do (avoid them), it is only by accident (or because of a good hunch), as opposed to scientific rigour.

IT ALL BEGAN WITH A STRANGE EMAIL

(via MeFi)