Patri Friedman

What about Futarchy, but to replace elections rather than legislation? Ie select govt officials by using prediction markets on their performance. Ideally using objective metrics, but as a fallback just bet on ex post voter satisfaction. @robinhanson

Correct me if I've misunderstood, but doesn't any market-based system for elections reduce to "the rich can select the candidates of their choice"? Folks with food insecurity won't participate at all, while the rich can treat it as an expense, not a wager.

I think you misunderstand the mechanism. These are prediction markets. So people are betting on, say, what will the income of bottom decile be if we appoint A vs B vs C. Then we pick whoever is predicted to maximize income of bottom decile. Prediction markets have the property...

..that attempts to manipulate them makes them stronger. The more money is fed in by rich people trying to force a selection by making false bets (ie betting that A will increase income for poor by $6,000/yr when really it will be $2,000/yr), the more quants earn by fixing.

Sure, but what if Candidate B includes as part of their platform "I'll cut taxes on prediction markets"? A quant's rational self-interest may not focus *solely* on the outcome of a particular bet.

Also, maybe this is a solved question somehow, but I don't see how you'd *resolve* a bet like the one you described. How do you determine which candidate's plan actually maximizes income for the poor so you know which bet pays out?

You make make markets whose prices represent estimates E[X|A] if you will later know X and A. You don't need to later know anything else.

So do the E[X|B] bets just get cancelled if A wins? I feel like there are some unrecouped costs there, and a very real risk that manipulative bets on losing alternatives wouldn't be punished.