Vice President Cheney has promised that Iran will not get nuclear weapons.
If we had a very liquid current events prediction market then the U.S. government could bet, say, $20 billion that over the next 10 years Iran will not explode an atomic weapon.
Such a bet would make it far more credible that the U.S. would stop Iran from going nuclear. It might thus discourage Iran from trying to acquire atomic weapons.
The disadvantage of the bet, of course, would be that those on the other side of the U.S. on the bet would have a large incentive to help Iran get atomic weapons.
However, this sort of behaviour would make the prediction market less accurate. Whatever the effect of $20 billion dollars would have on the US and Iraninan behaviour, it is unlikely to reduce the chances of Iran getting the bomb by precisely twenty billion dollars worth of prediction markets.
And it removes one major player - the US government - from participating accurately in the prediction market.
Also, say the U.S. targeted the $20 billion offer - say it only takes bets from Iranian citizens.
It sounds as if the best idea would be to combine this with members of congress and presidential candidates taking large personal bets that Iran won't get the bomb.