#1. Patri Friedman is a Google employee who is part of their prediction market team, headed by Bo Cowgill.
#2. Patri Friedman wrote yesterday a blog post on Overcoming Bias under the title, Dangers of political betting markets. ["Dangers"?? I'm so frightened. 
]
#3. His blog post has nothing to do with “overcoming bias”.
#4. The term “betting markets” is used two times, whereas the term “prediction markets” is used only once. [The blog post is filed in the "prediction markets" category, though.] Patri Friedman follows Robin Hanson’s footsteps —the Master prefers using the term “betting markets” for common markets, and the term ” prediction markets” for the markets that he rates as more prestigious. [Absurd. We should have one single term that tells what it is, and, if needed, we should add an adjective that rates that particular kind of prediction markets, and thus making many variations with one single term and many adjectives. Same idea with the term "prediction exchange".]
#4. Patri Friedman opened by stating that “I’m a big fan of prediction markets, and I think that in most cases, they do good by creating more accurate estimates.” The truth is exactly the opposite of that. In most cases, the prediction markets aggregate the information that we already know about and the predictions that they generate do not surprise us. All the work of a prediction market consultant in a firm willing to employ internal prediction markets should be to determine the specific (and rare) instances where they can make a big difference —e.g., project schedule forecasting and pre-demand forecasting.
#5. The body of Patri Friedman’s blog post (Dangers of political betting markets) deals with a science-fiction hypothesis —a political candidate betting/speculating against himself/herself. It is a totally remote and uninteresting hypothesis.
#6. It makes Overcoming Bias looks like a blog on science-fiction, disconnected from the real issues —a useless read.
#7. Robin Hanson adds a comment, saying that he wishes more “markets that give the consequences of electing any particular candidate“. That is an interesting idea, which has pros and cons.
UPDATE: Mike Giberson has a long comment…
[...] He worries that the ability of special interests to hedge their attempts to buy political favors will make that particular market more efficient, and therefore increase the amount of effort that goes into “the coercive business of politics.†[...]
NEXT: What Robin Hanson meant
Chris, I think you misread the post, which in any case I found more interesting than expected given your commentary.
Your point about well-purposed internal markets is a good one, but somewhat beside the point of Friedman’s post. Political betting markets are generally conceived as public markets.
However, the “dangers” he conjures are more apparent than real. He worries that the ability of special interests to hedge their attempts to buy political favors will make that particular market more efficient, and therefore increase the amount of effort that goes into “the coercive business of politics.” However, donating to politicians to gain favors is competing for attention, where the relative amount rather than absolute amount matters. Making the political donations market more efficient doesn’t increase the amount of attention that the politician has, or the influence available to dispense.
The only “science fiction” element here, from the point of view of someone subject to U.S. law, is the legal limits placed on gambling on elections. The world conjured up in Friedman’s post must be in the far distant future (or maybe the U.K.).
One other thought suggested by Patri Friedman’s post. Maybe that big “manipulative” infusion of cash into the Hilary Clinton contract at Intrade was actually someone hedging against the possibility of a HRC win?
Does the publicly available information allow a conclusions as to the motives behind the trades (investment, hedging, manipulation, liquidity supply/demand)?
“Your point about well-purposed internal markets is a good one, but somewhat beside the point of Friedman’s post. Political betting markets are generally conceived as public markets.” … in your first comment…
My point also holds for public prediction markets. Many of them do not surprise us, when we are well informed. It’s good that they give us an objective probability for an outcome, but to the well-informed persons, they don’t bring us surprising results.
So, in matters of public prediction markets, too, we should “determine the specific (and rare) instances where they can make a big difference”.
I didn’t see anything dangerous in what he suggested, either: what he’s saying is, essentially, “In a more efficient market, big government is a bad deal.” Which is classic protectionism applied to a new constituency.
But I don’t think it’s science fiction to claim that politicians hedge their bets — Lieberman did it in 2000 by running for his Senate seat even though he was a VP candidate. Prediction markets make the common explicit. If the result is surprising, it’s evidence that markets are more informative than we thought.