Robin Hanson as Prospective Economics Nobel Laureate
Michael Strong April 22nd, 2007
Contra Robin’s recent claim that he has failed both the market and the academic test,
I’ll take credit for creating some ideas the world has found useful, but I have completely failed both the market test and the academic test. That is, I can’t convince any business to let me join them to deliver my ideas at scale, and I can’t convince any top journal to publish my ideas.
I continue to see prima facie evidence that Hanson’s work will likely change the world in significant ways and that he is a reasonable candidate for a Nobel Prize. To be fair, I assume that Hanson is referring to the slow pace at which his work on decision markets has received attention and/or been implemented. But given his contributions to both prediction and decision markets, his remark seems myopic.
Many economists have won Nobel prizes largely for the creation of productive new fields of inquiry, including Gary Becker (sociological applications of economics), Vernon Smith (experimental economics), James Buchanan (public choice theory), and Ronard Coase (law and economics). While Hanson is not the sole creator of the prediction markets field, he is certainly a key founder whose “Could Gambling Save Science?” is, I predict, still in its early stages in terms of its professional recognition as a source of fruitful ideas. (Assuming that this paper eventually gets the recognition that it deserves, his greatest challenge with the Nobel committee will then be that he had the poor sense to write it before he received a Ph.D., thus potentially putting him in the same category as the great Gordon Tullock: Not certified to win a Nobel prize).
Masse argues persuasively that Hanson’s decision markets concept is ignored by real world decision makers for much the same reason that most adolescent males would not be interested in new and improved dildos. I suspect that “Could Gambling Save Science?” has received less than is its fair share of attention from academics for similar reasons. In both cases, Hanson is proposing a disruptive innovation that has the potential to undermine the work and lower the status of many current stakeholders and gatekeepers while potentially allowing many interlopers in who have not proven themselves in the existing status systems.
This delayed recognition need not be a matter of conspiracy, but rather merely incentives combined with the cognitive biases that Hanson articulated in his original paper. Hanson’s premise is that the algorithms by means of which academic decision-making takes place are themselves systematically flawed. Given his keen understanding of the failures of existing decision-making in both academia and the real world, it is premature for him to suggest that he has failed in any sense whatsoever.
Between cognitive biases, information and search costs leading to institutional inertia, legal obstacles, and adverse incentives faced by many current gate-keepers and decision-makers we could probably come up with multiple plausible stories regarding why Hanson’s ideas have not been adopted more quickly. (I would hope that Bee Hossenfelder’s difficulty in understanding Hanson’s ideas in a recent conversation at his “Overcoming Bias” blog illuminates for him the magnitude of the cognitive biases and cognitive costs that are preventing more rapid transmission even among very “smart” people who actually spend time trying to understand him.) That said, if the ideas are indeed better, in terms of adding greater value in their respective domains, then over time I would expect them to be adopted. They will often be adopted in the first instance by new entrepreneurial entities which are designed from the ground up with a novel set of internal institutional incentives. As someone who has both started, and promoted, such new entitites it is obvious to me that, to state the banal, it takes time to do so.
Relative to 1990, prediction markets have come a long way in terms of both academic recognition and public implementation. Decision markets, too, will gradually be implemented at some scale, somewhere, largely after prediction markets have become far more familiar and ubiquitous. If both prediction markets and decision markets, twenty or thirty years from now, turn out to be conceptual innovations with a public impact as large as the conceptual innovations introduced by Becker, Smith, Buchanan, or Coase, then Hanson is likely to be a serious candidate for a Nobel prize.
- All Guest Authors's Posts , Analysis (Industry) , Analysis (Meta)
- Comments(2)
Thanks Michael for the nice words, but Nobel prizes don’t tend to be given out for economic inventions, but more for theories or data. They also are rarely given out for ideas that didn’t hit the top journals when they came out, no matter how well they are later regarded.
Your “theory” is that prediction markets will improve the production and dissemination of accurate intellectual judgment as compared to the existing system of academic publication. Empirical evidence of the validity of your theory will arise as more people compare circumstances in which prediction markets accurately identify valid intellectual judgments as compared to circumstances in which the publication paradigm only slowly and haphazardly identifies valid intellectual judgments.
Hitherto your theory has not been clearly and unambiguously verfied by empical evidence in this sense. But I could imagine clever academics and prediction market-makers creating compelling comparisons that were widely regarded as empirical validation of your theory. Such empirical verifications will be more compelling to the Bee Hossenfelders of this world than is all the persuasion you can muster.
The notion that thirty years from now Nobel committees will still be as attached to reliance on publication in top journals as they are today assumes that no new credible technologies for the identification of excellent intellectual judgment will be developed between then and now. But your ideas provide the theoretical basis for just such improved technologies. Thus it remains a possibility that such improved technologies will be developed that will result in the legitimation of broader criteria in the awarding of such prizes. (And I do think that you under-estimate the importance of the “later regarded well” aspect in the existing identification process - despite the fact that at present they look for a fig-leaf of formal academic legitimation, and thus a Tullock can, at present, be screwed by the process.)
Not a done deal by any means, but it is important to encourage the notion that innovation is possible even in the realm of academia as an institution. The alternative is to concede that we are stuck with a radically sub-optimal knowledge-generating and knowledge-disseminating institution in perpetuity.