Photo by Alexander Abero on Unsplash
Photo by Alexander Abero on Unsplash

When a movie is led by female actors, reviewers dial up the sexism

Embargoed until: Publicly released:
Peer-reviewed: This work was reviewed and scrutinised by relevant independent experts.

Reviews of films with a female-dominated cast are more likely to be sexist than reviews of male-dominated films, according to international researchers. The team developed an AI system to detect two types of sexism; 'benevolent' sexism that reinforced patronising stereotypes of women, and 'hostile' sexism which involves hostility and negativity toward women. The researchers tasked their AI with reading over 17,000 film reviews and compared the levels of sexism against the female involvement in the movies the reviews were about. They say movies with female-dominated casts had a 149% higher magnitude of hostile sexism in their reviews and a 44% higher magnitude of benevolent sexism. The researchers say the gendered nature of this criticism is likely impacting the advancement of women in the film industry.

Journal/conference: PLOS ONE

Research: Paper

Organisation/s: American University of Beirut, Lebanon

Funder: This project was generously funded by the University Research Board (URB). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Media release

From: PLOS

Reviews of movies with female- versus male-dominated casts found to contain more sexist language

Linguistic study of 17,165 reviews highlights how critics may contribute to biases in film industry

In a new linguistic analysis, reviews of movies with female-dominated casts were found to have significantly higher levels of sexism than reviews of movies with male-dominated casts. Jad Doughman and Wael Khreich of the American University of Beirut, Lebanon, present these findings in the open-access journal PLOS One, on January 29, 2025.

Prior research suggests that negative movie reviews can affect actors’ finances, career paths, and mental well-being, while also influencing the broader media landscape. However, studies of gender bias in reviews have traditionally relied on movie ratings or box-office earnings, without considering the words used by reviewers.

To address that gap, Doughman and colleagues combined the Movie Review Data dataset containing 17,165 film review transcripts with metadata from the Open Movie Database on the films’ first actors, first writers, and first directors. Then, they used a previously validated AI-powered gender bias detection system to detect gender bias in the reviews. Their analysis distinguished between well-established forms of gender bias, including “benevolent” sexism, which reinforces idealized or patronizing stereotypes of men as dominant and women as needing help, and “hostile” sexism, which is expressed with negativity and aggression.

On average, reviews of movies with female-dominated casts were found to have a 149 percent higher magnitude of hostile sexism and a 44 percent higher magnitude of benevolent sexism than reviews of movies with male-dominated casts. Benevolent sexism in reviews was found to be highest for movies in the genres of Family and Music, while hostile sexism in reviews was found to be highest for Romance movies. Compared to their male counterparts, female first actors, directors, and writers were found to receive greater degrees of both benevolent and hostile sexist criticism.

The analysis also underscored the industry’s representation bias: 72 percent of the first actors, 91 percent of the first directors, and 86 percent of the first writers were male.

These findings suggest that critics may judge films differently depending on the genders of those involved. Thus, the researchers suggest, critics could play a key role in promoting a more equitable and unbiased film industry. Future research could address this study’s limitations, including its primary focus on films and reviews from Western studios and critics by including more data from non-Western sources. In addition, this study relied on name-based predictions of gender, which did not recognize non-binary or gender-diverse identities.

The authors add: “Our AI-driven analysis of 17,000 professional film reviews reveals that movies with female-dominated casts receive up to 149% more hostile sexism and 44% more benevolent sexism in their reviews compared to male-dominated films. This systematic bias, combined with severe underrepresentation—only 28% of lead actors, 9% of directors, and 14% of writers are female—creates significant barriers to women's advancement in the film industry, impacting both their earnings and well-being.”

Attachments:

Note: Not all attachments are visible to the general public

  • PLOS
    Web page
    The URL will go live after the embargo ends

News for:

International

Media contact details for this story are only visible to registered journalists.