Quantifying the Oscars' Woman Problem

A large Oscar statue is seen in the Dolby Ballroom during the 86th Oscars Governors Ball press preview in Hollywood, California February 20, 2014. REUTERS/Fred Prouser

Every awards season prompts a flurry of articles asking whether the Oscars have a "woman problem." Usually that conversation centers on behind-the-camera categories - such as directing, writing and producing - in which women are a perennial minority. But often someone will notice on-screen disparities as well. "Only three of the nine films nominated this year even have women in leading roles: American Hustle, Gravity and Philomena," writes feminist critic Holly L. Derr of this year's slate. Two years ago, Anita Sarkeesian pointed out that a mere two of the 10 nominated films passed the Bechdel test. Indeed, a quick scan of recent best picture nominees and winners shows a dearth of female-driven stories.

Focusing on particular years and films, though, can easily devolve into tiresome debates full of anecdotes and speculation. What specific female-centric films, someone might ask, are obvious replacements for specific male-centric films this year? Did Bridesmaids fail to get nominated two years ago because it's about women, or because it's a comedy? (Or because it was a sometimes very gross comedy?) Cue some pointing out how the singing ladies of 2002's Chicago beat The Lord of the Rings: The Two Towers, and we fade out....

This is why I like data. Being a professional numbers nerd, I decided to cross-reference 80-plus years of data on acting nominations and best picture nods to see if any patterns emerged. The gender differences were surprisingly distinct. First, I noticed that actresses, unlike actors, are more likely to be nominated for performances in films that are not nominated for best picture. Films with meaty roles for women are, by and large, considered lower caliber by the Academy.

Meanwhile, within the universe of movies that do get nominated for best picture, women-centric films are much less likely to win. Zero Dark Thirty and Black Swan, for instance, were good enough to make it to the final round in their respective years. But they joined a long line of actress-driven films that failed to win. In other words, the problem is not just that Hollywood produces too few films about women (and therefore smaller raw numbers of potential nominees). It's that even among those select films that meet the Academy's nomination threshold, the success rate for female-centered films is dramatically lower than the success rate for male-centered films. Women's stories, the data show, are not particularly valued by the Academy.

Data Snapshot: What Films Do Actors and Actors Get Nominated for?

With a few exceptions, 10 actors and 10 actresses* are nominated for Oscars each year (combining the lead and supporting categories). Often, those acting nominations come from films that have also been nominated for best picture. This is no surprise - the quality of a film depends in part upon the strength of its featured players. Overall, around 46 percent of all acting nominations come from best picture-nominated films. Break that figure out by gender, however, and differences start to emerge.

* For clarity I use actor and actress in this article, though generally I like the trend of referring to all thespians as actors, regardless of gender.

Only 40 percent of women's acting nominations come from best picture-nominated films, as opposed to 52 percent of men's acting nominations. Nearly twice as many actors (14 percent) as actresses (8 percent) were nominated for films that eventually won best picture. In other words, men more often get recognized for their performances in films of the highest perceived quality.

We could derive various hypotheses from this initial finding. Perhaps the relationship is causal. Maybe men are just better at acting than women, and their superior acting skills help elevate a film into best picture territory. This seems unlikely and also unprovable.

More likely, given the wealth of previous evidence, is that simply too few meaty roles for women exist in the majority of movies - including the best films of any given year. Gravity has room for only one woman, for instance, and films like The Wolf of Wall Street don't exactly provide key roles for the ladies. By this theory, actresses are reacting logically to a scarcity problem by seeking high-quality parts in lower profile films, rather than competing for the small number of well-developed women's roles in the year's greatest films.

But what if our whole definition of great film is itself gender-biased? So far we've been looking at the acting prizes through the lens of best picture, but let's flip that focus. Can the acting categories instead help us glean information about the types of films the Academy respects the most?

What Stories Are Considered Best Picture-Worthy?

Around 500 films have been nominated for best picture. Did these films tend to focus more on men, women or a mix of both? Watching them all would take months. Instead, I used acting nominations as a convenient indicator of the film's focus. Stories that are predominantly about one gender tend to have more nominated performers of that gender. Think of 1954's On the Waterfront, with four nominated actors and one nominated actress, versus 1950's All About Eve (four actresses, one actor). Of course, in many cases the balance is less skewed, so it's harder to be certain without qualitative information.

That's why we can learn a lot by comparing the two most skewed subsets of films: Best picture nominees where only actors or only actresses were nominated. If Oscar had no gender bias, we'd expect similar stats for these two extremes of the spectrum.

In the history of Oscar, 83 of the films that have been nominated for best picture also had nominations in at least one of the women's acting categories but zero nominations in any of the men's acting categories. Let's make the assumption that these films are, on the whole, more centered on female characters. Out of these 83 female-centric best picture nominees, only six have actually won the big prize - 7 percent.

Now let's look at the alternative scenario. There have been 146 films nominated for best picture that featured at least one nominated actor but zero nominated actresses. Again, we'll assume these films tend to be centered on male characters. It's already noteworthy that this pool of nominees is so much larger than the pool of female-centric films. The kicker is that, of these 146 male-centric nominees, 29 went on to actually win best picture - 20 percent, which is nearly triple the rate for the actress-heavy films.

The difference becomes even more glaring when we further restrict these pools to best picture nominees with two or more acting nominations. Multiple great performers must mean it's a pretty great movie, right? Well, yes, if those performers are men. Out of all the best picture nominees with multiple nominated actors (but zero nominated actresses), 44 percent ended up winning the big prize. Only 7 percent of the best picture nominees with multiple nominated actresses (but no nominated actors) went on to win. That is worse than the 9 percent win rate for films with no acting nominations at all.

There are also 44 films that never even got nominated for best picture despite having two or more actress nominations (but no actor nominations) - this list includes films like 1975's Carrie and 1991's Thelma & Louise. Only 22 films with multiple nominated actors (sans actress noms) failed to get a best picture nomination.

Finally, films where both actors and actresses are nominated seem to do better when men are the dominant force. Films with two or more nominated actors and a nominated supporting actress (but not lead actress) were much more likely to succeed than the reverse situation. Recent examples include 2010's The King's Speech, which featured a trio of acting nominees: Colin Firth as the King, with Geoffrey Rush and Helena Bonham-Carter supporting. That film won best picture. The 2008 film Doubt - with a lead actress nomination for Meryl Streep and supporting nominations for Amy Adams, Viola Davis and Philip Seymour Hoffman - was not even nominated.

Time Is Not a Cure

My data set covers the entirety of the Academy's 80-plus years of awards. But don't be tempted to place blame on an earlier, more sexist era for skewing the data. I split the data set into two halves and re-analyzed, and the results were disappointing.

The modern era has shown an even stronger skew away from female-dominated films. In fact, 1997's Titanic is the only modern best picture winner to feature nominated actresses and no nominated actors. By contrast, 17 post-1970 best picture winners have nominated actors and no nominated actresses.

Of course, that list includes such acclaimed films as The Godfather, Schindler's List, and No Country for Old Men. I don't mean to imply that any given best picture winner is unworthy. But it's hard to look at the lists of winners and losers and not see evidence of Hollywood's devaluation of female-centric stories. I love 2006's The Departed, but if male and female stories were given equal respect, might The Queen have won instead? How many films have missed out on nominations, let alone wins, because they focused on women and were dismissed as unserious? How many films never got made because Hollywood didn't think the stories were worth telling?

The Road Ahead

The "actors only" and "actresses only" best picture nominees represent extremes on a spectrum, and unfortunately those extremes are not balanced. The good news is that a plurality of best picture nominees throughout history have featured nominated actors and nominated actresses, and those films also have the highest rate of actually winning best picture. Two of this year's frontrunners fall into this middle ground: American Hustle and 12 Years a Slave both have male and female nominees in their casts. In the grand distribution of things, it would be ideal if stories full of interesting male and female characters took an even greater share of the pie. If, however, some acclaimed films focus more on one gender than the other, it would be nice if that didn't predominantly mean men. Who knows - maybe Gravity or Philomena will pull an upset and help balance Oscar's ledger a bit.

On the other hand, I strongly suspect that a data set with race variables would show an appalling bias towards stories about white people, which 12 Years a Slave could help counteract. In 2012 the Los Angeles Times revealed not just how overwhelmingly male the Academy Awards voters are, but how white and aged they are as well. The composition of Academy voters - which has a median age 62 - is "nearly 94 percent Caucasian and 77 percent male" according to the Times. It shows. Regardless of which film wins best picture this year, Oscar voters have a long way to go towards proving they can identify with people who don't look like themselves.

Of course, the Academy does not exist in a vacuum. It reflects cultural biases in addition to contributing to them. We all subconsciously incorporate some cultural preferences into the types of stories we applaud or dismiss. I'm sure many people will roll their eyes if I suggest that 1995's Sense and Sensibility is a better film than best picture winner Braveheart, but how much of that eye roll is about quality and how much is a culturally absorbed feeling that Jane Austen is frivolous and war epics are important? Even if women's stories were only very slightly undervalued relative to men's - the archetypal "A" grade versus an "A-plus" - it would be enough to dramatically skew rankings in aggregate. The Academy, however, is in the unique position to codify that skew into nominations and gold statues. Was last year's winner Argo definitely better than Zero Dark Thirty or Beasts of the Southern Wild? How did the poorly rated Extremely Loud and Incredibly Close get nominated two years ago but Bridesmaids and The Girl With the Dragon Tattoo did not? Was the hero at the heart of 2000's Gladiator really more compelling than the heroine (and metaphorical gladiator) in Erin Brockovich?

These are all entirely subjective questions. And in any given matchup, it's difficult to single out gender as the sole reason why one film triumphs over another. But, by taking the long view, the data can help reveal deeper biases. We need more parity in the films that get added to our cinematic pantheon. It's not enough to just beef up roles for women in the types of stories Hollywood already esteems. We need a broader definition of what makes a great story in the first place.

Amelia Showalter (@ameliashowalter) is a quantitative political consultant living in Washington, D.C.