Will Ohio State’s football team decide who wins the White House?

Here is my recent Slate.com piece with Kevin Grier, excerpt:

Just how irrational are voters? It is statistically possible that the outcome of a handful of college football games in the right battleground states could determine the race for the White House.

Economists Andrew Healy, Neil Malhotra, and Cecilia Mo make this argument in a fascinating article in the Proceedings of the National Academy of Science. They examined whether the outcomes of college football games on the eve of elections for presidents, senators, and governors affected the choices voters made. They found that a win by the local team, in the week before an election, raises the vote going to the incumbent by around 1.5 percentage points. When it comes to the 20 highest attendance teams—big athletic programs like the University of Michigan, Oklahoma, and Southern Cal—a victory on the eve of an election pushes the vote for the incumbent up by 3 percentage points. That’s a lot of votes, certainly more than the margin of victory in a tight race. And these results aren’t based on just a handful of games or political seasons; the data were taken from 62 big-time college teams from 1964 to 2008.

The good news, we suppose, is that sports really can cheer us up and make the world seem like a brighter place. The sports fan is left happier and more satisfied all around, not just on the gridiron. When you are feeling upbeat and happy, you feel more satisfied with the status quo in general. And feeling satisfied with the status quo makes you more likely to vote for the incumbent politician, even if that’s totally irrational.

The study’s authors control for economic, demographic, and political factors, so the results are much more sophisticated than just a raw correlation. They also did a deeper analysis that took into account people’s expectations. It turns out that surprise wins are especially potent, raising local support for incumbent politicians by around 2.5 percentage points.


If true, there should be an incentive to schedule "easy" home games in swing states pre-election. Does this happen?

Perhaps, the government ought to reinstate the tradition of Roman festivals.....

An incentive for whom? The politicians standing for election don't control the football schedules, so without knowing the political proclivities of the people who do set the schedule, we can't say which way the effect should go. Most likely it's mixed, since many people in many different jurisdictions are involved in the scheduling.

Some politicians are involved on the board of directors. I've pondered the fact that a layman might consider a 10 and 2 football coach to have done a good job when in fact it could be rather mediocre. Boards and ADs like winning records and coaches like to protect the incumbent coach.

Or 'trustees' rather, as it they are known in highfalutinland.

Sitting legislators? That seems like it would be pretty rare. Can you cite some examples? At any rate, even the Board of Visitors (that's what they call it here) doesn't set the football schedule. At most they have some influence over the hiring and firing of the AD, but even the AD has only limited control of the schedule because he must coordinate with the team's conference and with other schools. Thus, the idea that incumbent politicians could control the schedule enough to use this effect to their advantage seems far-fetched.

You underestimate the importance of informal chains of command.

So, your contention is that multiple, interconnected "informal chains of command" staffed by people with divergent agendas can somehow coordinate reliably enough to produce a detectable effect? That is an unbelievable stretch. It's amazing, the intellectual knots some people will tie themselves into in order to avoid having to say, "You're right; I hadn't thought of that." Try it just this once. I promise it won't kill you.

Anyway most late October-November games are conference games, which are set by (you guessed it) the conference, not the individual ADs. Absent some huge nationwide conspiracy, I just don't think this theory holds water.

I am also willing to bet that the politicians will reject this hypothesis out of hand. Do you really think if Obama wins his team will say things like "well, we won because there was an uptick in employment numbers, which we had no control over, the weather was nice where we needed it to be nice, thus increasing voter turnout in some key cities, and Florida voters were in a good mood because the Noles and Gators won?" Of course not. They (both of them) genuinely believe that if they are voted in, it's because they deserve it, that the American people buy into their program wholeheartedly, and that they'll lead us to the promised land. One of the primary qualifications of being a successful politician is self-confidence to the point of arrogance.

"The study’s authors control for economic, demographic, and political factors, so the results are much more sophisticated than just a raw correlation."

Which means the study is too wonkish for Progs/Libs.They are only capable of taking median female wages divided by median male wages and claiming women are paid 70 cents on the dollar of men FOR THE SAME WORK.

Can we keep the discussion level above "you're a doodle-head!" "no, _you_ are.", please?

With OSU playing Illinois at home, Romney might be in trouble...

Seriously - I can't believe Tyler didn't go in this direction in the article. He mentions the PedState game that's 10 days before the election, instead of the one against OBAMA'S HOME STATE just three days earlier. Granted, PedState has a better chance of winning than does Illinois, but that just means an Illinois victory would be a much bigger upset that would put voters in a much fouler mood.

If Ohio State loses to either Penn State or Illinois, it would be natural for Ohioans to feel that there's something wrong with the world -- especially at home to inept Illinois.

Documentation for "inept": http://www.fightingillini.com/sports/m-footbl/sched/ill-m-footbl-sched.html

Can't imagine Obama could get much bounce from a win over Illinois. Buckeye fans (rightly) expect to win that one. OTOH a loss would be pretty good news for Romney, if we believe this study.

Florida, Iowa, Penn St. & Notre Dame will all be heavy favorites to win on Nov 3. Wisconsin has the week off. I dont know who people in Virginia root for...

Just one more indictment of democracy.

This is hard to swallow. Sounds like GNP Granger causing sun spots to me.

This is completely plausible. There is strong evidence that a big victory boosts local consumer spending in the week after the game, which is almost certainly a result of a temporary improvement in the people's mood or confidence. And it's perfectly rational that a voter in a good mood (for whatever reason) is more likely to vote for the incumbent, because that voter will feel like "things are going pretty well."

It also comports with my personal experience, as I am completely irascible and bitter after a loss. I don't know if that feeling lasts a full week (Saturday to next Saturday election) but it will probably last until Tuesday. On an unrelated note, why the hell did they decide to have Presidential elections on Tuesday?

There is a boost in local consumer spending in the week after a big game, win or lose. I have been to several OSU vs Michigan games (a pretty big game, in particular the OSU#1 vs UM#2 game several years ago) and there are literally at least 300,000 people drinking beers and otherwise consuming outside the stadium in addition to the 100,000 in the stadium. Many of them have traveled and so stay a few days. Thus, a big increase in spending in the local economy.

As well, silly as it sounds, my mood for a few days is no doubt affected when my team loses a big game. As a cleveland/ohio sports fan, my mood is not often made better. That said, I have yet to observe even one person I know come out of a game (win or lose) with a different favored candidate. I realize thats a small sample, but I just dont see it happening. If this result were replicated a few times, perhaps my mind would change. For now, however, I will error on the side of experience.

That's all true, with respect to ticket sales/hotels/restaurants who rely on gameday business, but I'm talking about the indirect effects of victories, which go well beyond the actual day of the game.


Fair enough. People win and the celebrate, people lose and they go home. Changing temporary discretionary spending, sure. Changing who they vote for, not so much.

On an unrelated note, why the hell did they decide to have Presidential elections on Tuesday?

Because people have important things to do on the weekend?

I'd be curious to know whether anyone's quantified the extent to which having the election on a workday increases the relative voter turnout of the AARP block, for whom "weekday" and "weekend" exist in theory only.

According to NPR, it couldn't be Monday, because people might need to travel to the polling site and wouldn't do so on the Sabbath. And they needed to be back on Wednesday, which was farming towns' market day.

Well that explanation remains perfectly relevant in today's society!

That's the day bars take in the least amount of money, its all about economics; isn't it.

Interesting stuff. There are new analytics being done in football that actually determine the probability of winning for each team on a play-by-play basis. So you could theoretically watch the election unfold by charting win probability for the most popular teams in the swing states. See AdvancedNFLStats.com, FootballStudyHall.com, FootballOutsiders.com, etc for reference on the football side (note: I'm not affliated with any of these other than being an avid reader). They already determine the most important plays of the games. Wouldn't it be interesting to say that "Steinkowski's fantastic block on the defensive end here opened up the running lane, securing another four year run for the Obama administration?" Ha!

Somebody should do a study on the impact of Cleveland Browns football on voting behavior.

I am skeptical of fooball's relationship with voting behavior, but I am certain Cleveland Browns football is causally related to making me want to vomit week in week out.

So the comically awful Colorado team is of national importance now?

And they happen to be playing Oregon so it is certain to be another landslide defeat for Colorado. On the other hand, I presume all the losing has already had near it's maximum effect on CU fans by this point in the season so the marginal effect should be pretty small. Or in other words, their mood can't get much worse.

Oops, the Oregon game is this weekend, not the weekend before the election... so just replace "Oregon" with "Stanford" and the rest of the comment stands.

I honestly don't know how supposedly realist invocations of median voter theorem and other voting theories based on rational self-interest can survive studies like this. A great addition to Ilya Somin's work on political ignorance.

I think the effect could be even greater in countries where national teams are highly significant. I know that in New Zealand, a shock loss in the semi-final of the Rugby World Cup to France three weeks before the 1999 election was considered by many to be the nail in the coffin of the incumbent National government. Conversely, winning the 2011 World Cup (against France in the final) helped the incumbent government win re-election despite a bad economy. I'm sure similar results can be observed in soccer-mad countries in Europe and Latin America.

I approve of this message.


I'm not sure this evidences any more irrationality than voting based on economic performance when the elected official(s) in question have little control over the economy. What's interesting about this study is that performance of sports teams is so important to fans.

For many people, the utility of following a winning team far exceeds the utility of (say) a 2% marginal income tax cut. Hell they probably spend more than that on tickets.

A little quibble: Per your description of the results, the incumbent gets a boost in a state if that state's team wins. Georgia is settled; Florida is in play. Why would Romney have "some advice for how the Gators can bottle up Georgia’s running game"? Wouldn't Romney want to help Georgia, so that Florida loses and Floridian voters feel less well disposed toward the president?

The Labour landslide in the 1966 UK general election, and their loss to the Tories in something of an upset in 1970 are often attributed to the English victory in the World Cup final right before the first election, and their loss in the final right before the second.

Don't we run into problems with the rise of early (and increase in absentee) voting, particularly in states like Ohio? If a substantial number of people vote prior to the game in question, wouldn't the impact be smaller?

This seems important, especially since there really is only 1 swing state.


I think you are misreading the football schedule. PSU plays at Purdue the week before the election. The election is November 6, not October 30. So stop trying to get us to root against PSU. Evil Tyler.

We don't have to wait for next week for the games to be impactful. From the study:

"The effect may be somewhat stronger for the games occurring the week before the game than for the games immediately preceding the election, although the effects are not statistically distinguishable (P = 0.21)."

Does this apply only to college teams, or do Obama's hopes also hinge on the success of the Browns and Bengals?

When I saw the title, I thought this was going to be a post on how Ohio wouldn't vote for a self-professed Michigan Wolverines fan like Romney.

This piece of writing will help the internet viewers for
creating new blog or even a blog from start to end.

My spouse and I absolutely love your blog and
find many of your post's to be what precisely I'm looking for.
Does one offer guest writers to write content for you?
I wouldn't mind publishing a post or elaborating on a lot of the subjects you write related to here. Again, awesome weblog!

PSU plays at Purdue the week before the election. The election is November 6, not October 30. For more info visit: http://timesharescamconsultant.com and go to the blog then the post of the law of large numbers and the Social Science.

Comments for this post are closed