The Silver Ceiling

An analysis of 10,213 movie listings reveals the monetary cost of misogyny for women who direct, and star in, major films.

By Declan Bradley

It was November 2021, and Chloé Zhao was facing disappointment. The Academy Award Winning director had garnered widespread excitement for her latest film: Marvel Studios' The Eternals. Yet the movie underperformed dramatically at the box office, earning only about 160 million USD — the second lowest of any Marvel film to date, and not enough to make back its budget.

Zhao's case was not unique. Since 1980, films directed by women have earned significantly less at the box office than films directed by men, a divide that has only steepened in recent years, according to more than 10,000 film records scraped from the rankings site Rotten Tomatoes.

Films directed by women earn less at the box office than films directed by men.

Data scraped from Rotten Tomatoes. Genders approximated from directors' first names using U.S. Social Security Administration (SSA) baby name data.

This profit gap persists despite the fact that audience reviews of the same films remain largely uninfluenced by director gender, varying by only a handful of percentage points.

Films directed by women receive similar ratings to films directed by men.

Data scraped from Rotten Tomatoes.

Yet this is only half the story. A director may guide the artistic vision of a film, but its stars are the face of the project. Since Rotten Tomatoes does not list whether a film is primarily about men or women, we'll need to take a closer look at the film's descriptions to approximate that gender divide. Consider these synopses for The Bourne Identity (2002) and The Hunger Games (2012). Both descriptions provide subtle cues we can use to approximate the gender data we need.

The story of a man (Matt Damon), salvaged, near death, from the ocean by an Italian fishing boat. When he recuperates, the man suffers from total amnesia, without identity or background... except for a range of extraordinary talents in fighting, linguistic skills and self-defense that speak of a dangerous past. He sets out on a desperate search-assisted by the initially rebellious Marie (Franka Potente) - to discover who he really is, and why he's being lethally pursued by assassins.

— The Bourne Identity (2002)

In what was once North America, the Capitol of Panem maintains its hold on its 12 districts by forcing them each to select a boy and a girl, called Tributes, to compete in a nationally televised event called the Hunger Games. Every citizen must watch as the youths fight to the death until only one remains. District 12 Tribute Katniss Everdeen (Jennifer Lawrence) has little to rely on, other than her hunting skills and sharp instincts, in an arena where she must weigh survival against love.

— The Hunger Games (2012)

Not all synopses so clearly categorize themselves as 'the story of a man', nor do all include the main character's first name. But most do include a pronoun — 'her life' or 'his story' — and usually the first pronoun mentioned will refer to the film's primary lead, if one exists. By extracting those first pronouns from each synopsis, we can compare the profitability of films about both men and women.

Films about women earn less than those about men.

Data scraped from Rotten Tomatoes.

In the 2010s, films about women grossed 64 cents at the box office for every dollar grossed by films about men. And since 2020, that number has dropped to only 41 cents on the dollar, the lowest since 1980.

Those also directed by women earn even less.

Films by women about women earn 43 cents on the dollar.

Data scraped from Rotten Tomatoes.

Films made by female directors about female protagonists earn less than every other category — grossing less than half the average earnings of films made by men about men.

Films made by male directors about women earn 10 million dollars less, on average, than those about men. While such films are less common, they still make up nearly 30% of the available data for male-directed films.

The majority of films with available data are by men about men.

Data scraped from Rotten Tomatoes.

By contrast, female directors – their work already underearning – face an additional loss of only about 2 million dollars for projects starring women. That such a significant profit gap exists even for male directors suggests that the gender of a film's protagonist has a significant effect on its profitability, regardless of the creative team behind the project.

Yet ultimately the most significant profit gap remains for a film's director. For films starring men – the most profitable category – female directors can expect to earn an average of 20 million dollars less than their male colleagues.

Women who direct consistently earn less than their male colleagues in every category.

Data scraped from Rotten Tomatoes.

That profit gap can have real economic consequences. Of the 778 films in the dataset released since 2020, only about a third described women in their synposes, and only one in five were directed by women. Without more enthusiasm from audiences, such films may find it difficult to get off the ground.

Methodology

Data collection with rvest and analysis with R and RStudio. Page design with Coolors, Flourish, and GitHub Pages.

RottenTomatoes.com is a movie rankings site which stores individual films under unique URLs. Each film page contains information about the film, including critical (Critics %) and popular (Audience %) reviews, a synposis, the film's U.S. gross in USD, and a list of five related films.

A spider scraper was fed an initial list of 300 seed urls taken from Rotten Tomatoes editorial list of the best movies of all time. Upon visiting each page, the scraper saved relevant information about a film to a local disk. It then collected the URLs of that page's five 'related' films and added them to the bottom of the stack for further scraping. After each iteration, the stack was refined to remove URLs that had already been visited.

This method produced a 'spiderweb' growth pattern that organically visited thousands of films reachable from the initial set before eventually converging when the remaining stack reached 0. While this method did not guarantee an even pass over all Rotten Tomatoes pages, it provided a large sample of films within the general breadth of culturally relevant films.

Data analysis was performed in R, with directors' first names converted to genders using the SSA method. The original 10,723 records were reduced to 10,213 rows with usable data.