FWIW, I’m pretty sure the biggest long-term obstacle is the illegality. Considering the immaturity of the technology, user-interface, question-writing process, methods to integrate these forecasts into actual decision-making for big firms, and so on. These would probably all improve in a world with vibrant, legal, let-it-rip prediction markets.
Take the early Covid forecasts as an example. It’s easy for the authorities to ignore unknown nerds on Metaculus, who turned out more accurate. And we saw PolyMarket hosting questions about Covid too, but they got fined a million dollars. They still run, but their potential will be hamstrung to a mere shadow, I’m sure kept on a very tight leash.
It involves letting that stuff continue to develop, to establish and expand. Thousands of instruments trading, on all sorts of pathogen-related questions.
At some point, it would be untenable for epidemiology authorities to ignore it. It would be a bit like central bankers ignoring the stock market, or farmers ignoring the commodity futures market.
The illegality may not directly stop innovation, and doesn’t stop companies from internally dabbling in it. But it has an overwhelming chilling effect on the endeavor. Imagine a startup founder trying to bake in a “Fire the CEO” market into the company’s structure. They would of course be fined by the CFTC, once they’re big enough to be noticeable.
Thus, prediction markets are relegated to illegitimate crypto projects, which no big firm will want to touch. Or internal experiments that die as soon as the enthusiast takes a position somewhere else, instead of being the invasively-growing market that it probably would be. Or prestige-points like Metaculus, good for hobbyists, bragging rights, and maybe a resume for an unusual think tank position. Or developing countries with a tiny fraction the potential value and talent. Or possibly good for a tax writeoff, assuming Manifold pulls that off without getting shut down or fined at some point.
All the points you raised are very illuminating, and sobering for enthusiasts like myself. I’m just surprised that when people speculate about why prediction markets haven’t taken off, they seem to gloss over the one crucial, long-term problem. They’re illegal, you get fined or shut down.
Thanks Jotto, I agree that evaluating prediction market potential with reference to what it has achieved is messy because it has been stymied by regulations. Note though that it has been less stymied in the EU.
Strong upvote.
FWIW, I’m pretty sure the biggest long-term obstacle is the illegality. Considering the immaturity of the technology, user-interface, question-writing process, methods to integrate these forecasts into actual decision-making for big firms, and so on. These would probably all improve in a world with vibrant, legal, let-it-rip prediction markets.
Take the early Covid forecasts as an example. It’s easy for the authorities to ignore unknown nerds on Metaculus, who turned out more accurate. And we saw PolyMarket hosting questions about Covid too, but they got fined a million dollars. They still run, but their potential will be hamstrung to a mere shadow, I’m sure kept on a very tight leash.
It involves letting that stuff continue to develop, to establish and expand. Thousands of instruments trading, on all sorts of pathogen-related questions.
At some point, it would be untenable for epidemiology authorities to ignore it. It would be a bit like central bankers ignoring the stock market, or farmers ignoring the commodity futures market.
The illegality may not directly stop innovation, and doesn’t stop companies from internally dabbling in it. But it has an overwhelming chilling effect on the endeavor. Imagine a startup founder trying to bake in a “Fire the CEO” market into the company’s structure. They would of course be fined by the CFTC, once they’re big enough to be noticeable.
Thus, prediction markets are relegated to illegitimate crypto projects, which no big firm will want to touch. Or internal experiments that die as soon as the enthusiast takes a position somewhere else, instead of being the invasively-growing market that it probably would be. Or prestige-points like Metaculus, good for hobbyists, bragging rights, and maybe a resume for an unusual think tank position. Or developing countries with a tiny fraction the potential value and talent. Or possibly good for a tax writeoff, assuming Manifold pulls that off without getting shut down or fined at some point.
All the points you raised are very illuminating, and sobering for enthusiasts like myself. I’m just surprised that when people speculate about why prediction markets haven’t taken off, they seem to gloss over the one crucial, long-term problem. They’re illegal, you get fined or shut down.
Thanks Jotto, I agree that evaluating prediction market potential with reference to what it has achieved is messy because it has been stymied by regulations. Note though that it has been less stymied in the EU.
I think CFTC has no authority over play-money internal prediction markets, so that undercuts illegality a bit.
I guess one might even experiment with structuring them as real money markets, e.g., by paying winnings as “bonuses.”