What was PAM?
The Policy Analysis Market or PAM was designed by a host of people, including Net Exchange and the Intelligence Unit of the Economist for DARPA. A number of independent scientists helped with the design as well, mostly through Net Exchange which apparently maintains close ties with the academic community.
What did PAM trade in?
While the original announcement on July 28 by US Senators Ron Wyden of Oregon and Byron L. Dorgan of North Dakota claim that this was a terror market that traded in assassinations and terrorist attacks, the truth was that this market did not trade on those kind of events. What apparently occurred was that the company, Net Exchange had a couple of ill-advised examples. From a Washington Post article on Poindexter's resignation letter:
The letter contained no acknowledgement of personal error. In the case of the futures trading plan, he said, an unauthorized decision by an outside contractor -- the small California firm Net Exchange -- to post "some extremely bad examples" on the program's Web site gave critics ammunition to distort the effort as a proposed market in terrorism. The examples included the possibility of betting on the assassination of Palestinian leader Yasser Arafat or the overthrow of Jordan's monarchy.
Dr. Robin Hanson, an economics professor at George Mason University, and one of the architects of the PAM system had this to say:
The web page had an example of a regime change, which is
very different from a terrorist attack. Typical claims
would have been about the level of economic growth in
Egypt in 2nd quarter 2004, or the level of civil instability
in Saudi Arabia in the 3rd quarter of 2004, or the level
military activity in Iran in the 1st quarter of 2005.
Later, he added as a clarification:
There was such an example on the web page, unfortunately. But the idea was
that it was an example of regime change, and it is little like a typical
individual terrorist attack. That is the sort of event that could have big
repercussions in the middle east, and so the sort of events people would
want to include in their scenarios they are forecasting.
In any case, PAM didn't trade in terrorist events or assassinations, but it did trade in things that the US and incidentally the target country would be deeply interested in.
OK, What Made PAM Special?
There were a number of things. First, PAM bet real money on real issues. While not everyone is interested in the Middle East, the point is that PAM would have traded on the likelihood of events that were important to the US. This is the first time that markets were attempted as a way of accumulating information in the intelligence community. And there are strong reasons to think that PAM would have worked. The key is that markets are the best way discovered so far for collaborating the knowledge of participants. Sometimes they are astoundingly accurate (as when assigning blame for the Challenger Space Shuttle accident in 1986) and sometimes not (the Dot Com Bubble), but what works better? The markets allow one to express their degree of confidence (by purchasing little if they're not very confident to a lot if they are), and they reward accurately those who are right and penalize those who are wrong. The simple truth is that we haven't found anything that does those simple things better.
Second, PAM was an special kind of an Idea Futures market. It's a betting market that trades in outcomes of events like one would trade stocks. PAM went much further. In additional to trading on these basic claims, you could trade a mix of claims at once. For example, you could assign probabilities to the huge number of possible outcomes for a large number of claims. If there were five states that Jordan's economy could be in, and two states for whether or not Jordan experienced regime change, then one could assign a probability to each of the ten possible outcomes (that is, an economic state crossed with whether or not regime change occurred). As I understand it, this could have be done for many claims yielding thousands or even millions of possible outcomes each with its own probability. Further, this was done in such a way that even incredibly low liquidity (say one or two people trading on these outcomes) would generate useful information. So add neat technology to the list.
Third, the market wasn't limited to government employees. How many programs in the history of government have built in, non-government participation. Even extreme views (like communists or libertarians) could participate in this market. And most controversially, the radically destructive elements like terrorists and assassins could trade here as well. I consider this a good thing (more on this later) because how else are you going to hear about attempting regime changes? Even though the civilian world isn't thought of as a knowledgeable pool concerning the Middle East, the simple truth is that there's a lot of people out there with knowledge. This market would have brought them together.
Finally, all trading would be anonymous. That means that the government employees who are forced to slant their reports the right way could still accurately report their views via the market. The insiders would trade, again improving the accuracy of the markets.
Now to answer some common questions about these markets. What's wrong and how it can be dealt with?
The bad guys could trade and profit.
Since the market was an experimental market, initial investment was limited to $100. Another little issue that the opposition never noticed. So in the short term, no one is going to fund their coups or assassinations off of this market. Even years later, PAM would be dwarfed by the stock markets which as in the 9/11 attacks have already been used to profit off of terrorist attacks. Further, this is an example what is called "moral hazard". That is, the buyer of a good engages in risky or dangerous activity due to the purchase. This is a known problem and it's dealt with in the stock markets, insurance markets, gambling markets, and many other areas every day. For example, billions of dollars are gambled on the Superbowl or the World Soccer Cup. However, we don't have trouble with cheating. Nor terrorism, assassinations, or any of the other behavior being associated with PAM. These risks are still present, but they have been managed to an acceptable level so that gambling on the games is uninhibited by it.
Insider trading is perhaps another concern. But PAM dealt with it in the proper way. It would reward those who trade, and hence who bring the information into the market. Also, insider knowledge is a vague term. A casual visitor to Jordan might end up having "inside" knowledge through some accidental observation. As a prediction market, it can't make a best possible prediction if this information is prohibited from entering the market.
Markets Don't Predict the future that well.
The recent example is the dot com bubble. The question shouldn't be why didn't the markets predict accurately the price of the dot com stocks, but rather who did better? The media didn't do better. The government didn't do better. Academics didn't do better. Public statements from companies, stock analysts, mutual funds, venture capitalists, and other self-interested parties certainly didn't do better. When you look at the aggregated sources of information, no source handled the Bubble very well.
Markets are efficient aggregators of information, but if that information is incomplete or wrong, then the market will be wrong as well.
PAM was just gambling.
Some people consider gambling immoral. I don't. For what its worth, PAM would have a special legal status (through negotiations with the SEC and IRS resulting in limitations like the $100 investment ceiling, I think) that would put it above usual gambling laws. But why should gambling on a sports game where it simply doesn't matter who wins be better than betting on how much a country's GDP goes up? At least with the latter, you get quality information on the future economic state of that country.
Also, in all markets you have speculation. Purchases with the expectation that the value of the purchase will increase in the short term. Generally, speculation is thought of as gambling. Sometimes, the level of speculation is much greater than the hedging transactions (purchasing the security to hedge a risk). Yet we don't treat this activity under the gambling laws. The trading in PAM is small and experimental, but if it were extended to large volume, could be used to hedge all sorts of business dealings in the Middle East.
Further, here's something to think about. By putting money in PAM or a similar market, you are saying something about the future and risking something at the same time while trying to be right. I think that sort of activity should be allowed.
PAM was US-centric. Why should the rest of the world care?
First, you may reside in a part of the world that would be covered by PAM. That information would be good for you to know too. Second, chances are that you live in a country that would be affected by any negative events in the Middle East, hence the knowledge that was so useful to the US would also be useful to you. Finally, the ideas behind PAM can be applied locally and nationally. Perhaps, your company needs to evaluate the success of a marketing project or of a new product. Perhaps, your town wants to determine the likelihood of certain natural disasters or project the future economic prosperity of the region, and your country most certainly would have national interests.
Finally, we have the opportunity to settle in a rational way debates on the relative harm of various global policies. That is, how harmful is the US's CO2 (Carbon Dioxide) production? The current ways are embarrassing. Basically, there's a pro-business side that states global warming is a chimera, there isn't a problem, and jobs are at stake! While the environmentalist side states that we're on the edge of global catastrophe and CO2 limits (or perhaps harsher regulations) should be imposed. The facts such as they are, don't really support either side. Yes, there's significant CO2 production and substantial build-up in the atmosphere, and yes, there is global warming, but what it really means and what we do about it is undecided. Or I should say, undecided by the masses, but not those with a prior agenda.
What's still out there?
Well, there's the new protest site, the American Action Market. No word on their technology and how it compares to PAM, but with such claims as when the next lie "breaks" in the White House, you know it's juicy. Currently, the winner in the news has been TradeSports which trades a growing number of non-sports claims in finance, entertainment, politics, and world events. Further on is the reputation betting site, the Foresight Exchange which trades on an extremely eclectic set of claims including some terrorism claims in the US and nuclear war. It's an old favorite of mine. There's a bunch more out there both real money and not. So don't take this as a comprehensive list.