It’s been two weeks since I put out a challenge to proponents and practitioners of computable general equilibrium (CGE) models. As you may recall (or if your short-term memory falters, reread), I pointed out that, in light of well-known issues in general equilibrium theory, and after 30 years of experience, the time had come for those who believe in CGE to step forward with performance data. Is there any evidence that predicted CGE results hold in the real world, or that CGE methods are seen as useful in the marketplace? Or is this simply a self-perpetuating enterprise, feeding off academic and government contracts in a world where results don’t matter?
In addition to its thunderous presence on EconoSpeak, my challenge was mirrored in several other venues. Thanks to the magic of Google, I was able to monitor the comments posted around the blogosphere, and here is what I got: nada. Oh, I read several lame arguments that it is in the nature or purpose of CGE that real-world performance data are impossible to come by or that they wouldn’t matter anyway, but no one had even a shred of evidence to offer.
So here we go again, and I will be more explicit. Unless anyone can demonstrate otherwise, I suggest that CGE is an academic bubble whose intellectual value is grossly inflated, and which will crash to earth once the idea dawns that conjectures are no substitute for results. Those who currently buy this sort of analysis, not to mention those who make the career decision to invest years in learning its technical arcana, are throwing away their time and money. There’s no there there. Show me why I’m wrong.