Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
On Why The New York Yankees Aren’t Like Groundhogs…

November 4th, 2009 · 9 Comments
Fun With Data · Sports

First, a reminder on correlation versus causation here. For those of you that are too lazy to read that link, let me summarize: CORRELATION DOES NOT IMPLY CAUSATION.

Now that we are reminded of that important concept, let me introduce what I find to be a particularly amusing article on the relationship between Yankee success and economic health:

Wall Street might want to root for the Yankees in Game 6 of the World Series on Wednesday night in the Bronx. That’s not just because the Yankees represent New York, but because a coincidental sports indicator shows that a Yankee victory in the series has historically been bullish for the stock market.

For the record, this article is from the New York Times, not from some random no-name publication. To its credit, the article is careful to point out that it’s a “coincidental” indicator. In econ-speak, we call that a spurious correlation. We could have also made it simpler and called it “correlation without causation” instead. In this case, the relationship probably holds because some external force is affecting both Yankee success and economic performance. (For example, it could be the case that positive expectations allow the Yankees to spend more money, which makes them do better. I am playing around with the numbers here and will let you know if I find anything.)

The people over that the Wall Street Journal seem to be thinking along the same lines:

If the Yankees win the World Series the economy will have a nice bounce back in 2010 but if the Phillies prevail it will be a long slog to recovery, according to a Real Time Economics analysis of gross domestic product following Yankees and Phillies World Series victories since 1930 (which is as far back as the Commerce Department’s GDP numbers go). Okay, so it’s a stupid calculation, but just for fun let’s take a look at the numbers.

There is also the matter of correlation versus causation. Just because the stats point in one direction doesn’t mean there’s any causal link between Yankees victories and growth. Or is there? “Perhaps the Yankees thrive in years in which their rate of outspending other teams surges, and that their investments stimulate the economy. Or they only outspend when their staff economists forecast economic growth,” Mr. Bialik wrote in an e-mail. He added: “This Mets fan says any argument that suggests rooting for the Yankees is inherently flawed.”

Unfortunately, whoever wrote the headline “Yankees World Series Victories Boost Economic Growth” is less enlightened on the matter. The people writing about the negative relationship between Phillies wins and economic performance don’t do much better:

Historically, the truth is difficult to pinpoint, but one thing is certain – the triumphs of Philadelphia baseball always coincide with the devastation of the U.S. economy. You can’t have one without the other.

Had A’s owner/manager Connie Mack not created a dynasty, FDR and Churchill likely would not have been called upon to save the world.

Okay, the article takes the argument far enough to get me to believe that it is actually satirical. (I hope so, at least, but the commenters seem confused about the article’s tone, so perhaps the author should have made it more clear.) Serious or not, the article bring about the notion of the counterfactual, which is basically the “what would have happened if the world was the same except that this one thing didn’t happen?” scenario. If we are going to show that one thing causes another, in this case that the Yankees winning causes an uptick in the economy, we have to establish a counterfactual where everything is the same but the Yankees don’t win and show that the uptick doesn’t happen in this case. (Sadly, that would be a price I am willing to pay.)

Okay fine, maybe this is just a long-winded way of saying “Go Phillies.” 🙂 At least if the relationship does turn out to hold, I have a reasonable hedging strategy going…

Tags: Fun With Data · Sports

9 responses so far ↓

  • 1 Bponders // Nov 4, 2009 at 9:39 pm

    Why are the Yankees not like Groundhogs? Easy… they aren’t cure and furry.

    Oh, and Go Phils!

  • 2 Christopher Scott // Nov 4, 2009 at 11:59 pm

    Kiss the rings!

  • 3 Alex Rodriguez // Nov 5, 2009 at 12:32 am

    i’m going to go out on a limb and say that karma is the reason for this “coincidental” indicator. and yes i say “coincidental” because if we apply karma to the situation:

    the yankees won the world series = a great evil has occurred in the world.

    karma balances out all of these negative energies by kicking the economy into over drive.

    so if you believe in karma, then the yankees winning does cause a bull market just to balance the world out.

    (i’m not some big karma supporter, i just have to believe that if the yankees won, it was for a reason)

    GO SOX.

  • 4 dWj // Nov 5, 2009 at 10:58 am

    Correlation doesn’t even imply correlation; you say “the relationship probably holds because some external force is affecting both Yankee success and economic performance”, but it seems to me that noise is just as likely. What’s the p value assuming serial independence, is serial independence a reasonable assumption, and how many data were mined in the discovery of this relationship? (cf.

    In economics, these relationships often get trickier, as the belief in future correlation can have a causal effect that causes the correlation. If people expect the economy to correlate with sunspots, it will; if phases of the moon help coordinate collusive activity, the activity will correlate with phases of the moon. Each of these can start as a spurious correlation and become causal.

  • 5 hh // Nov 5, 2009 at 11:59 am

    Maybe this is an obvious question, but wouldn’t it be best for the economy if there were actually 7 games played instead of just 6, from all of the spending and business that takes place due to an event like this? Not just locally, but nation-wide, TV and radio, advertising, etc… By that logic, it seems as if it would be best if the Yankees won more slowly and dragged out the horror show…

    On another note, from a purely psychological and behavioral standpoint, I can see how any team winning will stimulate the economy because people generally feel better when their team wins, which means they’d celebrate, etc. (Side note: How does the theory of shopping-therapy conflict with this…?) Just looking at the fact that there are more people in NY (who are presumably happier when the Yankees win) than there are in Pennsylvania, the uplift makes sense (again, ignoring shopping-therapy). But what I’m realizing as I’m writing this, and what I can’t seem to reconcile just yet, is that it seems as if so many people outside of NY simply hate the Yankees, and that should counteract the effect… Then again, Wall Street is in NY…

    Either way, I’m rooting for the economy and staying out of it!

  • 6 econgirl // Nov 6, 2009 at 2:18 pm

    @hh: Not necessarily…if people are only willing to give the World Series a certain amount of attention/resources/whatever, they would just be spreading that attention over a longer time span. 7 games versus 6 probably wouldn’t have a particularly negative effect on attention per game, but if your logic held overall then why not play 100 games? In that comparison is is more clear that each game would be less lucrative…

    Part of the key to the logic of 7 versus 6 being attractive is that people don’t know ahead of time how many games there are going to be…and I’m confdent that there are people that don’t start paying attention until a game could be a clincher in one direction or the other. So you could contrive a setup in order to maximize attention, but simply lengthening the series wouldn’t do the trick.

  • 7 Joshua // Nov 10, 2009 at 5:22 pm

    I love watching people confuse correlation with causation over the stock market and the economy as a whole.

    Does the NYSE do better when the Yanks win, or does the economy? Because, though correlated, there isn’t necessarily causation here, esp. if you are looking at monthly/daily blips in the market.

    If it’s the NYSE, I’d be willing to bet there is a Master’s thesis out there for behavioral economics in NYC.

  • 8 rslakmackf // Nov 20, 2012 at 6:48 pm

    I continue to look for a solution to my problem. ,karen millen outlet onine

  • 9 vmware training in pune // Dec 24, 2016 at 2:35 am

    nice one…

Leave a Comment