Menus Subscribe Search
intrade

Intrade, We Hardly Knew Ye

• March 19, 2013 • 4:00 AM

The prediction market that shut down last week was fun, but it wasn’t any better at predicting politics than polls are.

If you’re a gambler, you might have noticed that the prediction market known as Intrade shut down last week. Theories abound as to why that happened. It’s not clear what will become of Intrade—perhaps it will be back up next week, perhaps we will never see it again. But I wanted to mention some of Intrade’s contributions to politics. (At the risk of being presumptuous, I’ll be referring to Intrade in the past tense, although I’ll be happy to be wrong about that.)

If you never used it, Intrade was really a clever gambling device, allowing you to bet on the probability of an event occurring. You establish an account and put some money in it. Then you look among hundreds of different betting “contracts” (who’s going to win the presidency, who’s going to win American Idol, will US GDP growth average above 3 percent this year, etc.) and decide where to place your money. Surer things were more costly bets. But you could buy and sell a contract any time prior to the event. So, for example, if you had bought an Obama re-election contract on May 1, 2011 (then valued at about 60) and then sold it on May 3, 2011, the day after Bin Laden was killed (when the contract was valued at about 70), you’d have turned a tidy profit.

Beyond the gambling fun, Intrade was believed to be very useful for those interested in forecasting elections. Polls have the disadvantage of being meaningful and accurate only within a few months—or weeks—of an election, since most voters aren’t well aware of the candidates or issues until then. By contrast, people betting on political events are a self-selected group of highly informed political observers, people with enough interest and confidence in their knowledge to actually put down money. In theory, at least, these folks should be able to provide much more reliable predictions much earlier in the race. And, indeed, some studies (like this dissertation by Ian Saxon) have found early political betting markets to be better predictors than early polls. Robert Erikson and Christopher Wlezien, however, note that markets and polls aren’t actually measuring the same thing, and that when properly compared, polls tend to do a better job forecasting.

Part of the problem with using markets like Intrade to forecast elections is that they may be subject to short-term manipulation, which compromises their predictive value. For example, back before banking laws made it prohibitively difficult for Americans to place bets on Intrade, I made a few. I made a bit of scratch (i.e., enough to buy a high-end Frappuccino) by betting on Hillary Clinton to win the 2008 Democratic nomination right before the Ohio primary (which she was expected to win) and then selling that contract the day after. I did the same with Mitt Romney right around the Michigan primary that year. (I also lost a few dollars here and there.) Notably, I did not think that either Clinton or Romney would win their respective parties’ nominations that year. I was just betting that enough other people would think that to briefly raise the value of their contracts, and it worked.

And that, of course, was part of the problem, if you were trying to use Intrade to actually estimate a candidate’s chances of winning a contest. My purchases slightly distorted the value of the contracts, and if enough other people were doing that at the same time, it just ended up inflating the candidates’ apparent chances of winning after a day of good news, even though their chances really hadn’t improved.

More generally, it turned out that people who bet on Intrade were subject to the same biases and impressionability that people surveyed in polls were. Or worse: as Josh Tucker notes, bettors may have been just relying on polls in placing their bets. Take an example from the 2012 Republican presidential nomination contest. Below is a chart from Real Clear Politics showing the poll standings of the Republican candidates. It was an odd contest in that almost every candidate saw some time as a front-runner. Note in particular Newt Gingrich (in green), who had two poll surges, one in late 2011 and the other in early 2012:

gingrich polls 2012

Now note Gingrich’s Intrade contract from the same time period. Basically, he sees the exact same double-surge pattern.

gingrich intrade

Now what, in fact, were Gingrich’s actual chances of winning the Republican presidential nomination in 2012? Virtually zero! He had almost no endorsements, most of the Republicans who served with him in the House were not only not backing him but actively opposing him, he had a history of flip-flopping on key Republican issues, he hadn’t served in office for over a decade, he had loads of personal baggage, etc. Either he never had a shot, or everything we knew about presidential nominations was wrong. Ideally, the markets should have reflected that, and should not have been so easily impressed by a few weeks of good press for him or a strong debate performance. But the markets proved as fickle as poll respondents.

So, if Intrade comes back on line, or if you want to use another site like the Iowa Electronic Markets, by all means have fun with it, but just know that you’re not necessarily getting any inside dope on the next election by watching people bet.

Seth Masket
Seth Masket is a political scientist at the University of Denver, specializing in political parties, state legislatures, campaigns and elections, and social networks. He is the author of No Middle Ground: How Informal Party Organizations Control Nominations and Polarize Legislatures (University of Michigan Press, 2009). Follow him on Twitter @smotus.

More From Seth Masket

A weekly roundup of the best of Pacific Standard and PSmag.com, delivered straight to your inbox.

Recent Posts

September 12 • 4:00 PM

Life, Liberty, and the Pursuit of Plastic Bags

California wants you to pay for your plastic bags. (FYI: That’s not an infringement on your constitutional rights.)


September 12 • 2:00 PM

Should We Trust the Hearts of White People?

On the 50th anniversary of the Civil Rights Act, revisiting a clip of James Baldwin on the Dick Cavett Show.


September 12 • 12:00 PM

Big Government, Happy Citizens?

You may like to talk about how much happier you’d be if the government didn’t interfere with your life, but that’s not what the research shows.


September 12 • 10:00 AM

Whispering in the Town Square: Can Twitter Provide an Escape From All Its Noise?

Twitter has created its own buzzing, digital agora, but when users want to speak amongst themselves, they tend to leave for another platform. It’s a social network that helps you find people to talk to—but barely lets you do any talking.


September 12 • 9:03 AM

How Ancient DNA Is Rewriting Human History

We thought we knew how we’d been shaped by evolution. We were wrong.


September 12 • 8:02 AM

Give Yourself a Present for the Future

Psychologists discover that we underestimate the value of looking back.


September 12 • 8:00 AM

I Walked Through the Financial Crisis

Why are former Wall Street employees guiding tourists around the Financial District? Paul Hiebert signed himself up and tried to find out.


September 12 • 7:05 AM

Scams, Scams, Everywhere


September 12 • 6:17 AM

In Soccer as in Art, Motifs Matter

A new study suggests a way to quantitatively measure a team’s style through its pass flow. It may become another metric used to evaluate potential recruits.


September 12 • 4:00 AM

Comfort Food Is a Myth

New research finds that, contrary to our beliefs, such foods don’t have any special ability to improve our moods.



September 11 • 4:00 PM

Reading the Camouflage Uniforms in Ferguson: ‘You Are Now Enemy Combatants’

Why are police officers wearing green or desert camouflage in a suburban environment?


September 11 • 2:00 PM

Wage Theft: How Two States Are Fighting Against Companies That Categorize Employees as Independent Contractors

New York and Illinois have passed hard-nosed laws and taken an aggressive tack toward misclassification.


September 11 • 11:03 AM

Yes, I’m a Good Person. But Did You Hear About Her?

A new study tracks how people experience moral issues in everyday life.


September 11 • 11:00 AM

Searching for Everyday Morality

Experimenters use text messages to study morality beyond the lab.


September 11 • 8:00 AM

The Geography of Uber

If it continues to grow—and there are few reasons to think it won’t—will Uber transform the infrastructure of cities or glom onto what’s already there?



September 11 • 6:05 AM

One Man’s Search for an Orgasmic Life Force

It remains unclear what “orgone” actually is, but Wilhelm Reich thought you could find it by sitting inside a box.


September 11 • 4:03 AM

Jack the Ripper’s DNA: Was Aaron Kosminski Behind the Whitechapel Murders?

Russell Edwards says he’s solved the mystery. His proof might be a little threadbare.



September 10 • 4:00 PM

The Average White American’s Social Network Is Just One Percent Black

And three-quarters of white Americans report that they haven’t had a meaningful conversation with a single non-white person in the last six months.


September 10 • 2:00 PM

Eye on the Fly

The tiny fruit fly has been beloved by developmental biologists for more than a century. Turing patterns may yet explain its shape.


September 10 • 10:02 AM

Why Do Women Earn Less as Mothers and Men Earn More as Fathers?

For women, becoming a parent means you can expect to earn even less over your lifetime—unless you’re Marissa Mayer.



September 10 • 7:00 AM

Is Back Pain Ruining Your Sex Life?

You might be doing it wrong.


Follow us


Big Government, Happy Citizens?

You may like to talk about how much happier you'd be if the government didn't interfere with your life, but that's not what the research shows.

Give Yourself a Present for the Future

Psychologists discover that we underestimate the value of looking back.

In Soccer as in Art, Motifs Matter

A new study suggests a way to quantitatively measure a team’s style through its pass flow. It may become another metric used to evaluate potential recruits.

Searching for Everyday Morality

Experimenters use text messages to study morality beyond the lab.

Is Back Pain Ruining Your Sex Life?

You might be doing it wrong.

The Big One

One country—Turkey—produces more than 70 percent of the world's hazelnuts. September/October 2014 new-big-one-2

Copyright © 2014 by Pacific Standard and The Miller-McCune Center for Research, Media, and Public Policy. All Rights Reserved.