Used to be, many ads appeared in local newspapers, which competed for attention via their local news coverage. This induced such papers to fund local investigative journalism, often looking for dirt on local politicians, which induced such politicians to avoid looking dirty. Which was good.

Alas, in the last few decades local ads have become disconnected from local news, contributing to a great fall in local investigative journalism, and likely a big rise in dirty politicians. But I see a simple fix:

Assume you will vote for 1 of 2 candidates for US Mayor, Governor, or House of Repr. Only 1 has posted $1M bond payable to anyone who proves their misbehavior re lies, bribes, or other cheating. On average, what is chance you vote for this candidate?

— Robin Hanson (@robinhanson) November 17, 2022

If typical voters are anywhere this responsive to such a move, candidates would face strong incentives to offer such bonds. They’d ask their donors to give them money to create such bonds, or to more directly back such bonds. Which would fund more journalists to seek such dirt, again inducing politicians to avoid looking dirty. Which would again be good.

Here is a more detailed vision, in time order.

A) Groups of apparently-politically-neutral investigation-experts team with financial orgs to offer bond holding and judging services. Each such service J would declare the kinds of dirt topics T they feel qualified to judge, and the kinds of assets they can bond. For each such topic T, they declare an evaluation fee $E.

B) A politician P approaches such a service J, deposits asset $X of appropriate form, and chooses a topic set S. Judge J then creates a bond that pays $X to anyone who, by deadline D, proves dirt on P re any topic T in set S.  Politician P can now publicly announce the existence of this bond, to induce voter confidence.

C) Any investigator who thinks they have found dirt on P of such a type T in S can approach service J by deadline D, pay appropriate evaluation fee $E, and then present their evidence in support of this dirt claim. This induces J to evaluate this claim. If J decides that the claim is valid, this fact is announced, and the $X asset is transferred to this investigator.

D) If deadline D is passed, no investigator has been awarded the bond, and no investigator evaluation remains in progress, then the asset $X is returned to the politician or their backer.

Prediction markets on whether any investigator will ever win a particular bond could offer voters more refined confidence indicators on a candidate.

This system design should handle most cases, though I could imagine problematic cases in which an investigator’s evidence raises many suspicions, but doesn’t quite rise to the level of persuading the judges J. Then a subsequent investigator might collect more evidence, at which point the judges J are persuaded. Perhaps in this case the judges might split the asset $X across these investigators.