Two weeks after the US presidential elections, postmortem results continue to flow forth both in America and abroad. While most are busy examining the weaknesses of the Kerry campaign or the hidden strengths of the Bush campaign, a few are still trying to figure out the bizarre discrepancies in exit polling data.
The reasons for the exit polling misdiagnosis of Ohio, Minnesota, and others earlier this month (or of Florida in 2000) are numerous, and most stem from the intricacies of the human mind. Meanwhile, the mainstream media has largely ignored a tool which proved amazingly accurate: the futures market.
Tradesports.com is one of many “futures markets” online. It essentially provides would-be betters with a platform for betting on anything from the viability of Iraqi elections to more mundane sporting events. For the 2004 presidential election, Tradesports set up a national popular vote contract which would close on election day valued at Bush’s final national popular vote percentage. So one would buy a “contract” for the current market price (i.e., estimated percentage), and either make the profit or lose the difference on the final percentage.
I’m actually on the road as I type this, and I didn’t have the foresight to note the closing price on the national presidential contract. It doesn’t matter anyway; the US uses the electoral college system, by which a candidate who wins a state’s popular vote receives the votes of the number of electors (equal to the total number of legislators a state elects to Congress). A candidate must win at least 270 electoral votes to win.
Thus, we turn out attention to the individual state contracts. Political junkie Dennis Johnson provides an electoral map of the US based on the closing prices on November 1st. At the time, political experts were predicting anywhere between a 320-vote Bush win to a similar Kerry win. Even pollster James Zogby infamously predicted a solid Kerry win. Look at the map: each state voted as the contract price predicted, resulting in an electoral vote total matching the exact end result.
The individual percentages were not necessarily accurate. New Hampshire was much closer than its closing price would indicate, but this was largely because Kerry had taken a lead larger than the margin of error in most polls, and even the Bush campaign seemed to have let the state go. Thus, the market was off.
The overall result was incredibly accurate, proving once more the power of markets as a source of information. Each person involved in the market brings his own knowledge to the table, and in composite condense more knowledge than even the largest poll with seemingly foolproof methodology…and possibly at lower cost.
The US Department of Defense tried to harness this power in 2003 by creating a futures market for world events. Such events as a possibility of an overthrow of the Jordanian monarchy or the assassination of the then-healthy Yasser Arafat were to be traded as rolling contracts with a closing price analogous to probability. The plan was quickly scrapped after widespread concern–especially from the left–over the moral implications of such a market.
One may say that the functioning of such a market is simply a matter of “putting one’s money where one’s mouth is”, as they say, but proponents will still have to wait some time before the public is accustomed to the idea of prices as information rather than a shopkeeper’s tool of oppression.