Whitman news since 1896

Whitman Wire

Iowa Electronic Markets best poll data in accuracy

Emily Percival

Hang on for a minute...we're trying to find some more stories you might like.


Email This Story






I am a person who hungers for the day my telephone rings and a pollster is on the line, asking me if I have a few minutes to answer some questions. Polls are used to gauge the sentiments of America, and the idea that one can reliable do this based on the opinions of a sampling of a few hundred people is fascinating. Even more so is the accuracy of political betting markets such as the Iowa Electronic Markets (IEM), where you can place bets on candidates and elections months in advance. Whereas polls can tell the current attitudes of the populace, the IEM can, with startling accuracy, predict the outcomes of elections.

A study done by the University of Iowa College of Business Administration titled “Results from a Dozen Years of Election Futures Markets Research” outlines some of the reasons for the impressive track record of the IEM. According to the report, “Traders are not a representative sample of likely voters; they are overwhelmingly male, well-educated, high income and young.”

Why, then, is the IEM a useful forecaster? The report says it is because traders are thinking differently about elections and politics than those answering poll questions. They are putting their money on the line, creating an incentive for thinking critically about the current political environment. They are less likely to ask themselves who they’d like to win and more likely to ask who other people would like to win: then place their bets on the latter judgment.

The study offers a complete analysis of the presidential election of 1996 as an example of the long-term accuracy of the IEM. Clinton’s margin of victory over Dole comes to around nine percent; just after Super Tuesday (Mar. 15) the IEM was predicting a Clinton victory by a margin of seven to eight percent. The IEM stayed in the five to nine percent range through Apr., while Gallup polls were predicting a Clinton victory by 13 to 22 percent, and Harris polls were predicting a 22 to 30 percent margin of victory. The IEM markets were significantly more accurate.

As the campaign progressed, the IEM fluctuated around the nine percent mark, but never over- or under-shot by more than six percent. Polling data spent months off by an average of 12 percent (with outliers much above that and a little below).

The IEM prediction for the 1996 election “diverged from the correct outcome in the final days close to midnight on eve of the election with prices further from the election outcome than they had been since the Super Tuesday primaries in March.” The study cites a “large cash influx by new traders” placing last-minute bets as the source of the inconstancy.

Last-minute inaccuracy is not the norm. In the 1988 Presidential Election, the IEM had a 0.2 percent error the night before the election, whereas the polls had a 2.7 percent error. The IEM also bettered polls on the eve of the 1992 presidential election with a 0.4 percent error to an average poll error of 4.4 percent.

The IEM can be cautiously used as a predictor for election outcomes over both the long- and short-term. Now, then, the all-important question: What is the IEM saying about the current election? As of this writing, the favored Democratic presidential nominee is Hillary Clinton, at 65.80 dollars, followed Barack Obama at 20.70 dollars. John Edwards trails in third place with 7.80 dollars. On the Republican side, Rudy Giuliani takes the lead with a price of 31 dollars even, followed by Mitt Romney with 26.40 dollars and Fred Thomson with 24.40 dollars.
Best of all, the IEM currently favors a Democratic presidential takeover 59.60 dollars to 40.30.

Print Friendly, PDF & Email

Leave a Comment

If you want a picture to show with your comment, go get a gravatar.




Whitman news since 1896
Iowa Electronic Markets best poll data in accuracy