HOME / press box: Media criticism.

When Wall Street Bet on Elections"Roosevelt New York Odds Drop to 7 to 5" and other headlines from a century ago.

By Jack ShaferPosted Friday, Sept. 26, 2008, at 6:21 PM ET

(Continued from page 1)

RSS

Related in Slate

Slate has a long interest in the Iowa Electronic Markets, going back to 1996, when we routinely published their data against our "pundit index" in a feature called "The Horse Race." This year, we toyed around with data in a new feature called "Political Futures." Dan Gross assessed political futures in this 2004 piece. James Ledbetter knocked prediction markets in this 2008 article for being slow, too small, and too low-stakes to be of value.

Compiling the historical data and analyzing it, Rhode and Strumpf discovered that markets did a fine job of forecasting elections in this era before scientific polling. "In only one case did the candidate clearly favored in the betting a month before Election Day lose, and even state-specific forecasts were quite accurate," they write.

The accurate results were probably not a fluke produced by easy-to-predict markets. Over the period studied, "Republicans won eight of the elections in the Electoral College and Democrats seven; the party in power won eight, the opposition seven," they write.

Campaign betting was an extraordinarily large business, too.

"For brief periods, betting on political outcomes at the Curb Exchange in New York would exceed trading in stocks and bonds. Crowds formed in the financial district—on the Curb or in the lobby of the New York Stock Exchange—and brokers would call out bid and ask odds as if trading securities," they write.

One criticism of the predictive power of modern election markets such as Intrade is that the bettors are merely parroting the wisdom found in the scientific polls. That criticism doesn't apply to the markets from the last century. Scientific polling didn't arrive until the mid-1930s, so these gamblers—I mean, investors—weren't pinching poll data to make their bets. To borrow James Surowiecki's insight, crowds betting their hard-earned money proved remarkably adept at aggregating political information. (Did the posted odds influence election outcomes? My gut says it's unlikely.)

Rhode and Strumpf attribute the demise of election betting to a number of causes. The law generally looked down on such wagers, and it grew tougher over time. Newspapers started giving less coverage to the wagers as they discovered "ethics." The legalization of pari-mutuel betting on horse races in New York in 1939 provided a much better product for gamblers. And stock exchanges started barring members from gambling on elections.

The final nail, Rhode and Strumpf write, was the advent of the scientific poll, especially Gallup's successful call of the 1936 election, which supplied newspapers with a "ready substitute for the betting odds, one not subject to the moral objections against gambling."

As our culture's moral objections against gambling approach zero, what excuse do daily newspapers have for ignoring the trove of political intelligence contained in its prices posted by Intrade, Iowa, and others? As indicators of future results, how much worse can they be than Zogby? They'd sure make better reading.

******

Greg Mankiw, FiveThirtyEight, and others are wondering this week if Intrade is out of line because its results aren't lining up with Iowa and Betfair, a U.K. gambling-futures site. Place your bets at . (E-mail may be quoted by name in "The Fray," Slate's readers' forum; in a future article; or elsewhere unless the writer stipulates otherwise. Permanent disclosure: Slate is owned by the Washington Post Co.)

Track my errors: This hand-built RSS feed will ring every time Slate runs a "Press Box" correction. For e-mail notification of errors in this specific column, type the word Strumpf in the subject head of an e-mail message, and send it to .