Can the Information Problem Be Solved By Prediction Markets?

Userpic
Matt Arnold
October 21, 2008

The last post discussed the information problem and hinted that there may be a solution. Bear with me-- I hope I don't turn off every smart person reading this, by revealing at the beginning that the solution may lie in a type of "market". It's not the kind you think. I think anarchist libertarians need to acknowledge that corporations are a form of government, so this post is not a post about unbridled free markets.

Now, that having been said: markets are known to be just incredibly predictively correct most of the time. Correct in the service of evil, sometimes, granted. But often just amazingly prescient. The problem with corporations, and markets in general, is that the only metric they are currently set up to achieve is to deliver money to stockholders. That's why corporations, when considered as a giant super-organism hive mind person, are sociopaths. Because the only happiness-stimulus to that system's hive mind is financial gain.

What if you kick-start that into achieving measures other than sales and profits and shareholder financial return? Political Action Groups and other fundraising organs should be required to publicly bet in the government's official prediction market.

The one-sentence summary of the prediction market concept is "Vote on values, bet on beliefs." There would be a regularly-occurring referendum set up for the predictive market. Let's refer to it as the Priorities Referendum. Through voting in the Priorities Referendum, the populace prioritizes the importance of different measures of national welfare. This vote is used not as law-- not directly-- but as the official definition of national welfare. To have a say in what kind of nation you would like to live in (for instance you would like more forested areas, or higher grades, or better health) you just go into the voting booth and rank them in importance with all the other things that all kinds of other people want to see around them.

By contrast, to have a say in matters of factual expertise, you are claiming to know how the priorities would best get done, and you must put your money where your mouth is.

Each measure of welfare is like a horse race, in which competing policies are the horses. At the end of the year, if your existing policy you bet on has performed less well than you predicted in your bet, you get hosed, and the people who bet smarter than you have taken you to the cleaners.

I would love to read a story about someone in such a nation, who attempts to

(1) perform a massive sabotage of national welfare to win on the prediction market,

(2) and then attempts to contaminate the measurements of national welfare,

and whether or not they get caught.

Challenge (3) is getting Joe Public voting in the Priorities Referendum to understand the difference between ends and means. There is also a challenge (4) to be overcome, which I will discuss in a subsequent post.

The bottom line is that we must provide a higher reward to the most canny among us for telling us the truth about science and economics than they would receive for lying to us about it. The vulnerabilities in the current system are well-known, exploited, and largely unpatched.

Comments


drew4096 on Oct. 22, 2008 9:40 PM

This is a good idea in principal. But how do you impliment multiple policies at the same time in order for them to compete, especially if they conflict with one another? (ie, Bailout Freddy and Fannie versus letting them die.) Or is this your 4th challenge?

And if you only impliment one such policy (like, the one with the most money "bet" on it), who's to say that the other one wouldn't have had "better" or "worse" results?

Leave a Comment

Enter your full name, maximum 100 characters
Email will not be published
Enter a valid email address for comment notifications
Enter your comment, minimum 5 characters, maximum 5000 characters
Minimum 5 characters 0 / 5000