In thinking about the fall of Larry Craig, I went back and re-read Dan Popkey's Idaho Statesman story from last week. Popkey's story makes it clear that rumors had been dogging Craig on this question for years, of not decades.

Craig is clearly not the only politico that carried around the whiff of scandal before it actually hit. My Louisiana contacts tell me the same thing was true of David Vitters. And, Lord knows, everyone knew Bill Clinton had a problem before a story broke.

So here's my question to economists and political scientists. If there are prediction markets for elections, why isn't their a prediction market for politicians and scandals? Admittedly, elections have a clear end date and (hopefully) a clear winner. Still, one could devise several market outcomes on which to bet: a Washington Post story about a scandal, a Nexis count of news stories about a scandal, or even an actual resignation. Contracts could be limited to, say, 3-month or 6-month time windows.

This sort of thing could have the potential to be a useful indicator (admittedly, it would also be ripe for manipulation by mischief-makers; but so are election markets) for media and politicos -- it could create a metric for off-the-record, on-the-qt-and-very-hush-hush kind of information.

My question to Tyler Cowen: is there are markets in everything, why isn't their a Scandal Pool?