Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
When Economists Watch The Oscars…

February 27th, 2012 · 5 Comments
Fun With Data · Just For Fun

For those of you who weren’t already aware, the Oscars were on television last night. I’m pretty sure that most people watch the Oscars in order to see the beautiful people and beautiful dresses and suits that parade across the television screen. I will admit that I am not immune to this phenomenon, but I am probably somewhat unique in that, rather than wondering “who are you wearing?” I want to know “how much did your dress contribute to U.S. GDP?”

I’m also pretty curious as to how winning an Oscar contributes to the financial success of the movies, directors, producers, and actors involved. Apparently I’m not alone in this, since a line of research on this topic does, in fact, exist. The first paper in the series is entitled “What’s an Oscar Worth?” and it focuses specifically on the financial impact of the award on the box-office revenues that follow. (This may seem weird to you, but it’s important to keep in mind that Oscar-nominated films are strategically released later in the year, which means that they are reasonably likely to still be in the theaters around Oscar time. In addition, some theaters will put a movie back on its screens once it’s been nominated for an Oscar.) The main results table of that paper looks like this:

As the table indicates, revenue impacts exist both because more people are going to see the movies and because movies are more likely to stay in the theaters longer, and the impacts are bigger for wins than for nominations. In addition, as seems reasonable, the impact is bigger for Best Picture and Best Actor awards than for supporting awards.

It’s important to note, however, that this dataset uses movies from 1978-1987. On one hand, it’s impressive that the authors found statistically significant effects with this relatively small sample. On the other hand, it calls into question whether the results are representative of a longer time frame. Fortunately, we have some follow-up data to work with from a paper entitled “For Oscar Glory Or Oscar Money?” that uses a sample of movies from 1990-2000. Contrary to the original study, this new paper states that the main financial box-office benefit comes from the nominations and not from the awards themselves. Their main table is the following regression output, though I will warn that it is not for the faint of heart (or faint of context!):

Overall, it certainly seems like there is a tangible benefit to being nominated and/or winning a prestigious award, above and beyond the reward for just being a better movie/actor/whatever in general. *falls over in shock* In the future, I would like to see more work on the overall financial benefits, including DVD sales, etc. and also the future financial returns to the actors and actresses that are associated with the awards or the award-winning films.

For the record, I was not rooting for Moneyball to win…mainly because I was annoyed that it didn’t talk enough about the math behind the story. (I can feel the eye-rolling from here, thanks- I got plenty of it in person from my roommates.) I am also disappointed that I didn’t think to take a picture with my friend’s Oscar* and an econ book, since I distinctly remember thinking “that would be funny, but what are the chances I would ever have a reason to use it?” My bad. To make up for my lapse in judgment, I will play with Photoshop tonight and get back to you.

* To clarify, the Oscar was not won by my friend, it is instead a family heirloom type thing. That said, I fully expect all of my friends to eventually be extraordinarily successful.

Tags: Fun With Data · Just For Fun

5 responses so far ↓

  • 1 EconoNerd // Feb 27, 2012 at 4:54 pm

    Wait, are you telling me that Moneyball didn’t emphasize the cosmic sphere of awesomeness that Sabermetrics resides within? Was it just some biopic or something?

    If so, you can cross it off my must-see list.

  • 2 Chris // Feb 27, 2012 at 5:35 pm

    Very interesting results. Is this a logit model? Can you post some diagnostics? Did you check for serial correlation left in the errors, etc?

  • 3 Nick // Feb 27, 2012 at 8:53 pm

    I just wonder why Nate Silver didn’t build a model this year… muzzled by the NYT?

  • 4 Dave // Mar 2, 2012 at 11:48 am

    I’ve looked at the DVD rental data and it suggests that there are two types of movies that do well in the rental market: Academy Award nominees and horrible movies (Good Luck Chuck, anything with “The Rock” or Ice Cube in it, etc.). In both cases, viewers likely intend to only watch once, so there’s no need to buy.

  • 5 bdbd // Mar 2, 2012 at 4:50 pm

    speaking of which,

    (what ballerinas do, think they do, are thought to do by others etc)

Leave a Comment