An Election Form Guide


Today, all eyes are on the U.S. presidential election. There’s certainly plenty to follow, and in my latest Wall Street Journal column, David Rothschild and I begin by highlighting the most interesting markets to track. The big one is obviously who will be the next president:

Barack Obama‘s stock has risen in recent days, and he is now rated a 90 percent chance to win [update: now 93 percent], after holding steady at about 85 percent for most of the week. But we think this understates the true chances. Given the bias these markets typically exhibit against favorites, he may be as much as a 98 percent chance to win.

Incidentally, this forecast coincides with Nate Silver’s updated numbers at

When is the right time to start watching the markets? Throughout the day you will see the markets respond to early indications of voter turnout, to rumors, and to leaked exit polls. While the markets may give more false signals, the benefit is that they will keep you ahead of the televised talking heads:

Expect the networks to be reluctant to “call” a state on the basis of exit polls that are anything other than conclusive. By contrast, traders will respond immediately to any news. … In the 2006 election, we observed the prediction markets moving sharply for the winning Senatorial candidates at least half an hour before the major networks discussed the trends.

And for those in need of an hour-by-hour guide:

Approaching 7 p.m., Virginia, Georgia, and Indiana will be closing their polls. Virginia currently rates Sen. Obama an 86 percent chance to win this large state. Virginia is a useful bellwether for the validity of pre-election polls, and an Obama hiccup in this state could provide the first hints of a Bradley effect. By contrast, an Obama victory in Indiana, or even keeping the election close in Georgia (where he is rated a 25 percent chance to win), would send Sen. Obama’s stock soaring. If this occurs, the networks — looking to build tension — won’t call the election, but traders will call the race immediately, potentially allowing market-watchers to begin celebrating, or drowning their sorrows.

Approaching 8 p.m., Pennsylvania’s polls will close. If all of the states are lined up from most likely to go Obama to least likely, Pennsylvania is the pivotal state that puts each candidate over the line. [It is] just over a 90 percent chance for Obama. … Any sharp movement in the Pennsylvania market will be paralleled by an equally sharp shift in the chances of an overall Obama victory.

Those election-watchers who want the tension to continue past 8:30 p.m. should be cheering for an uptick in Sen. McCain‘s stock in Pennsylvania. If this doesn’t occur, then the only real question will be whether we are looking at an Obama landslide. … Markets currently rate Missouri a tossup at around 50 percent. If Sen. Obama wins, we expect him to ultimately prevail by as many as 200 electoral college votes.

Approaching 9 p.m., attention will focus on Colorado [which] is currently rated a 90 percent chance to vote Democratic. Unless Sen. McCain sweeps every state in which he is currently favored, then Colorado is a must-win for him.

You can read my full column here. You can listen to an interview I did on Marketplace, here (or read it here), and there’s an interesting piece in Forbes, here.

Tonight I’ll be joining the Journal’s all-star pundit cast in live-blogging the election at; stay tuned throughout the evening for updates.

Oh, and the Freakonomics vs. Economix bet? Economix responded today — and it’s on.

Finally, here are some links to the sites that I’ll be tracking:,, the Daily Kos election map; Drudge is a good place to look for exit-poll leaks, as are Wonkette and Slate; in 2006, The New Republic and National Review were among the first to leak these numbers. Expect the first reasonable guesses to hit at around 6 p.m.

And while you wait, here are the early-voting numbers and polls of early voters.


"Expect the networks to be reluctant to “call” a state on the basis of exit polls that are anything other than conclusive."

Yes, something about voting machines, Diebold, and a past election pretty much fried the networks exit poll strategy...

I can safely bet we don't hear about any alternative parties today- unless some serve as "spoilers" (the media's definition) to the business-owned, and media-fawned-over two parties...


Uh...what exactly does that DailyKos map show right now? Anyone?

Mike B

Another prediction market thought could be that people are using them to hedge against an improbable Obama loss as a form of emotional insurance, rather than being over confident in a McCain victory.


Okay, they fixed it. It was showing a McCain victory, with numbers that were different from previous results. Odd.

Tor Hershman

It’s your philosophical duty not to participate in the theocratic-oligarchies’ Punch & Judy Show Elections.

Economics = Terrorists yelling, “What do ya think, apples grow on trees?”

Stay on groovin' safari,

Tor Hershman