Reason Foundation

http://www.reason.org
http://www.reason.org/news/show/are-markets-smarter-than-polls

Reason Foundation

Are Markets Smarter Than Polls?

George Passantino
November 1, 2004, 6:29pm

For the last several weeks, we have been inundated with polls about who will win the Presidential Election tonight. While many of the polls that make their way onto the evening news look at the aggregated or popular vote, these are quite deceptive. In fact, as you will recall, Al Gore won the popular vote but lost in the Electoral College where the real action is. With that said, I couldn't help but wonder how futures trading markets would compare to traditional polls in terms of predicting the outcome of the election. If you are not familiar with the concept of elections futures trading, click here for basic information. Much like oil futures, elections future contracts are fully functioning markets where people can buy and sell contracts on whether or not someone will win a particular race...they typically trade on a range of 0 to 100. To illustrate, on Tradesports.com, you can purchase a futures contract on Bush winning Florida (and all of the other states). If the price of this Florida future is 97, the market is essentially saying it is a sure thing that Bush will win. If, by contrast, the price is at 3, the contract is a real long shot. If a contract is trading at 52, it is a very tight race. In a highly unscientific fashion, I haven taken these futures' contract closing prices from yesterday. With that, I have calculated the projected Electoral College outcomes and then compared them to state-by-state Electoral College tallies projected by polls from two sites, the Los Angeles Times (which has a great interactive electoral poll feature) and RealClearPolitics.com, which has their own averaging model. My Electoral College calculations ignore margins of error and the presence of undecided voters (which in some case could theoretically reverse an election outcome). They also do not account for Maine and Nebraska that allocate one of their electoral votes separately from the statewide vote. Here is what the results say before the actual vote tallies start coming in. Los Angeles Times: If you give Bush the Red States and Kerry the Blue States and then go with the latest polling data in the "toss-up" category (again, ignoring MOE and Undecided Voters), it shows Kerry at 269 and Bush at 265. 270 are needed to win. The reason that there is no winner is that Hawaii is tied. If Kerry wins Hawaii under this scenario, he wins. If Bush wins, you have a tie of 269 to 269. In that case, the House of Representatives votes–one vote per state delegation. Most likely Bush would win, given the Republican Majority. The VP is selected by the Senate. And, at this point, it is unclear who will control the Senate. Real Clear Politics. If you give Bush the Red States and Kerry the Blue States and then use the Real Clear Politics Rolling Average for the battleground states, Bush would receive 300 electoral votes and Kerry would receive 238. Bush would be the clear winner. John Fund. I greatly respect John Fund's political wisdom. Today on OpinionJournal.com he predicted Bush will receive 296 electoral votes to Kerry's 242. TradeSports.Com. If you take the states with "Bush Win" contracts valued at more than 50 and give Kerry the Bush Contracts valued at less than 50 (since such a contract is more of a long shot), Bush wins with 281 Electoral College Votes to Kerry's 257. There is also a separate contract on the total number of Electoral College Votes Bush will win. It closed last at 276. Tonight we will get to see how they stack up to reality. SITE LINKS RealClearPolitics.com Los Angeles Times TradeSports.com John Fund

George Passantino is Senior Fellow


Print This