Gender Inequality in the Movie Industry: A Big Data Analysis of Female Underrepresentation
Sebastián Cole Poma-Murialdo Student ID: 11351446
Master Thesis
Graduate School of Communication Research Master in Communication Science
University of Amsterdam
Supervisor: dr. Jeroen Lemmens
Abstract
The current study aimed to evaluate gender inequality in the movie industry. The gender of the cast and top crew positions from all 4885 feature films that received a wide release in United States between 1982 and 2017 were evaluated. Data was collected by scraping online movie databases. In the first part of the study, three types of representation were considered. First, the analysis of the numerical representation confirmed that there is high gender inequality in both cast and crew. Second, the examination of the quality of these representations revealed that, while all genres are male dominated, comedy, drama, romance and music have a higher proportion of women that the other genres (female genres). An analysis using the Bechdel test, used as a measure of female independence, showed that most movies from female genres pass the test, while movies from other genres fail it. Third, the analysis of the centrality of the representations showed no significant difference between big and small studio size, as both were male dominated. The relationship between cast and crew indicated that a higher proportion of women in the crew increases the proportion of women in the cast, especially for female genres. Finally, the second part of the study evaluated the relationship between the cast and crew’s gender and movie success. Overall, more women in the cast is associated with lower online review scores, awards and tickets sold, while the effect for female crew is positive on online ratings.
Gender Inequality in the Movie Industry
Over time, the movie industry has achieved a highly economic and cultural importance among the entertainment markets (Eliashberg, Elberse & Leenders, 2006). Hollywood is
considered the oldest, most profitable, and most influential of the film industries. As this industry is mostly considered a project and reputation-based system, formed by “a multiplicity of organizations and individuals” (Bielby, 2009; Hadida, 2009, p. 229), it is relevant to study the current demographics of such organizations and individuals. Specifically, it is necessary to understand the characteristics of the industry that produces and reproduces culture and
symbolic representations of gender both in movies and among fandoms (Bielby, 2009; Erigha, 2015; Toffoletti, 2014; Cohen, Atwell Seate, Anderson, & Tindage, 2017).
Hollywood can be considered as a tool for soft power for the U.S. culture and
ideology, as well as a transmitter of its social and political values (Aydemir, 2017; De Zoysa & Newman, 2002). This means, as previous studies on media effects have suggested, that gender under-representations or misrepresentations in movies may have consequences for viewers, possibly reinforcing or contributing gender stereotypes (Lauzen & Dozier, 2005). On the other hand, debates unfolding in Hollywood around issues such as gender pay gap
(Lawrence, 2015; Smith, 2015a; Carlin, 2016; Demaria, 2017) and sexual abuse and
harassment scandals (Smith, 2015b; Davies & Khomami, 2017; Cooney, 2018) suggest that consciousness around gender representation in the industry may be rising. All in all, research regarding gender representations can contribute to such awareness, by providing new insights on the topic.
The disparity of earnings and employment between men and women has been persistent since the 1980s, when data on specific movies became available (Bielby, 2009; Lauzen, 2012; Lincoln & Allen, 2004). The first aim of this study is to address the topic of gender inequality in the movie industry, both on-screen and behind-the-scenes, by reviewing
the representation of women. Three main types of gender representation are considered in the current study: 1) numerical representation, 2) the quality of the representation and 3) the centrality of the representation. Together, these determine how well a social group is represented in a specific industry. Furthermore, the relationship between the gender representation of the crew and the cast members will be evaluated. The second aim is to explore the influence of gender representation on film performance and success. Previous studies have suggested mixed results regarding how stars, and members of the production team influence success in terms of popular and critic appeal, such as award nominations and theatrical attendance (Elberse, 2007; Hadida, 2009). The current study aims to clarify the influence of the gender representation of the cast and crew on movie success, as most studies regarding movie performance do not consider gender distributions as a factor that may influence online ratings, ticket sales and award accomplishments.
The above elements are tested in the current study using an automated content analysis of websites. This allows the analysis of a larger dataset than previous studies where manual data collection was applied, and the scope was limited to specific years or genres. Scraping allows the use of a bigger sample to expand, update and refine existing knowledge on the topic. Overall, this study will review gender representation in all 4885 movies that received a wide release in United States between 1982 and 2017 by following Erigha’s (2015) three elements of representation: numerical, quality and centrality, and review the relationship between cast and crew and the effect of such representations on movie success.
Theoretical Framework Representation of Cast Members
Cultural products influence opinions, attitudes and behaviours of the people who consume them, while they reflect their creator’s ideology and opinion (Aydemir, 2017). Therefore, it is necessary to evaluate who produces such products and decides how men and
women will be presented in their stories. Erigha (2015) provides an overview of the
representation of gender in Hollywood’s film production and distinguishes three main aspects of representation: 1) numerical representation, 2) quality of representation and 3) centrality of representation. These three types of representation refer to both on-screen and behind-the-scenes positions and, as they are interrelated, provide a full image of the marginalization of social groups.
First, the numerical representation refers to the presence or absence of the social group or minority (Erigha, 2015). Women are underrepresented in movies, as numerous studies have found that male characters in popular movies outnumber female characters (Lauzen & Dozier, 2005; Lincoln & Allen, 2004; Acevedo Caradeux & Gil Salom, 2013; Coyne, Callister, & Robinson, 2010; Anderson & Daniels, 2016). There is also a decrease in the portrayal and presence of female characters over time (Lauzen, 2018a; Hunt, Ramón, & Tran, 2016). For instance, in the top grossing movies from 2017, only 24% of protagonists were female, which is a decrease when compared to the previous year (29%; Lauzen, 2018a). A tool by Google and the Geena Davis Institute to determine on-screen character’s gender and speaking time in movies confirmed that not only were there fewer female characters than male characters on-screen, but they also had fewer speaking roles (The women missing, 2017; Smith, Pieper, Granados, & Choueiti, 2010; Smith, et al., 2015a). Moreover, female
characters are completely absent or used as background characters more often than male characters, possibly increasing the generic perception of anonymity and powerless
characteristics of their characters (Acevedo Caradeux & Gil Salom, 2013). In sum, women are underrepresented in all areas, thus it is expected that:
H1: Since 1982 there has been women underrepresentation in wide released movies.
Second, the quality of representation refers to the roles that members of a social group are assigned to. In addition to being present, social groups in both on-screen and
behind-the-scenes positions seek for complex, multi-dimensional roles in a diversity of genres, instead of stereotypical roles and being typecasted (Erigha, 2015). However, not all roles in movies achieve such diversity and complexity. Studies suggest that, for example, actresses are more likely to be sexualised than male actors and top grossing movies in general have fewer female characters in leading roles (Smith, et al., 2015a; Lauzen, 2018a).
In line with the above, women are more likely than men to be portrayed in traditional roles, such as being a parent or in a relationship (Smith et al., 2010; Lauzen & Dozier, 2005), suggesting, that women on-screen are still, to some extent, dependent on men. This can be tested using the Bechdel test. To pass the test, a movie must have 1) at least two women, 2) who talk to each other, 3) about something other than a man (Bechdel Test Movie List, 2018; O’Meara, 2016). While the Bechdel test is not a strictly academic measure and has some limitations, it is still considered a valid test to measure the presence (numerical
representation) and independence of women (quality of representation) in film in a somewhat quantitative manner (Lindner, Lindquist, & Arnold, 2015; Lindner & Schulting, 2017). It has been used in both academic and non-academic studies as a measure of gender inequality in movies, revealing that, most movies fail the test (Friedman, Daniels, & Blinderman, n.d.; Sharma & Sender, 2014; Hickey, 2014; Lindner, et al., 2015; Lindner & Schulting, 2017). In line with this, it is expected that most movies in the sample will fail the Bechdel Test,
regardless of its low threshold.
H2: Most wide released movies between 1982 and 2017 fail the Bechdel Test.
The quality of representation of actors and actresses is also related to their typecasting into specific roles and genres. A movie genre is an abstract concept used to classify certain patterns or conventions of elements that are repeated in movies (Feuer, 1992). There is no clear definition, as genres are mostly subjective, socially negotiated and dynamic, and they may change depending on societal, temporal and cultural perspectives. Yet, existing studies
have shown that there are both structural and content differences among different genres (Buckingham, 1993; Grabe & Drew, 2007; Gottfried, Vaala, Bleakley, Hennessy, & Jordan, 2013). Studies report that while there are more men in all genres, they are more likely to have leading roles in action, adventure and animation, while female characters are more likely to appear in comedies and dramas (Lauzen, 2018a; Smith, et al., 2015a). In line with previous studies, typecasting is expected to be more pronounced in the same genres:
H3a: Male cast members have a higher presence in all genres, and especially in action, adventure and animation movies
H3b: Female cast members have a higher presence in comedy and drama movies compared to other genres
Third,the centrality of the representation indicates to what extent social groups are working in an industry’s core institutions and cultural products (Erigha, 2015; Erigha, 2018). These two elements are related, as bigger studios and institutions are expected to have bigger budgets, even if this does not mean that every movie in a big studio will receive the same high budget (Erigha, 2018). Participating in movies from bigger studios with bigger budgets
provides the cast and crew a better opportunity of gaining fame and an influential position in the industry. Today, there are six main studios that can be considered as core institutions in Hollywood. These Hollywood studios produce most worldwide blockbusters, and compete against each other, while forming an oligopoly and keeping the main competition among themselves (Gomery, 2008). As of today, and mostly since the beginning of the 21st century, the ‘Big Six’ are: Warner Bros., Paramount, Twentieth Century Fox, Sony, Disney and
Universal (Gomery, 2008; Schatz, 2009). Smaller studios usually produce independent, niche
genre, and smaller budget products (Schatz, 2009).
The size of the studio must be considered when examining the centrality of the representation. Elberse (2007) suggests that actors and actresses with high star power and
reputation are cast by big studios due to their possible impact on the success of the movie. However, press interviews suggest that big studios don’t hire women even for smaller projects (Dargis, 2014). On the other hand, Lauzen’s (2018a) study on 2017’s top films showed that movies with female protagonists are more likely to be produced in smaller (65%) than in bigger studios (35%), while sole male protagonists and ensemble casts were more likely to appear in movies produced by bigger studios. In line with the above, it is expected that bigger studios will have higher male representation than independent studios.
H4: Female cast members have a higher presence than males in movies produced by smaller studios, when compared to bigger studios.
Representation of Crew Members
The same three types of representation will be used to study the crew’s gender representation, starting with the numerical representation. Recent reports suggest that Hollywood is still dominated by men in behind-the-scene positions (Lauzen 2018a; Lauzen, 2018b; Lauzen & Dozier, 2005; Friedman, et al., n.d.). Three top crew positions will be taken into account in the current study, as these make decisions about the other members of the cast and crew.
The producer of a movie can be considered its leader, as it is their job to oversee the entire project, including all creative and organizational factors. In general, the producer identifies ‘properties’ that could be developed, selects the main people to develop it (such as directors and writers), determines budgets and locations, and contacts studios, sponsors and other organizations that are involved (Hadida, 2009; Erigha, 2015). The cast in the movie is also selected by the producer and to a lesser extent, the director and writing team.
While the producer mostly manages the business aspects of the project, the director guides its artistic and filming aspects. In this sense, the director oversees the artistic
While the producer also has a say in the selection of the cast and other creative aspects, the director has a larger say over who is selected as a contributor in the project (Hadida, 2009).
Third, writers are considered to have a strong influence on the product, as there wouldn’t be a final product without a script (Bielby, 2009). They create the characters, their personalities, behaviours and genders. In other words, writers decide, to some extent, the amount of male and female characters in the script.
Historically, there were more women in crew positions in Hollywood’s earlier days, but it became dominated by men as it became more lucrative (Bielby, 2009). Since then, women’s under-employment as screenwriters and directors has persisted (Bielby, 2009; Lauzen, 2012). Specifically, numerous reports show that women had fewer creative positions (producers, directors and writers) than men between 2007 and 2014 (Smith, et al., 2015a; Smith, Pieper, & Choueiti, 2015b; Hunt, et al., 2016). Therefore, in the current study an overall underrepresentation of women in top behind-the-scenes professions is expected:
H5: Since 1982 there have been more men than women working as producers, directors and
writers in the crews of wide released movies
The quality of representation of the crew follows a similar structure as the cast’s quality of representation. While producers, directors and writers avoid being classified into specific genres, female crew members seem to be limited to female genres, such as comedy, romance and documentaries, suggesting that gender-classification still exists (Erigha, 2015). Women directors have been found to be limited to specific genres, as female-directed movies from Sundance Film Festivals between 2002 and 2014 were focused on genres like drama, comedy and romance, when compared to male-directed films (Smith, et al., 2015b; Lauzen, 2012). Similarly, studies on the independent film industry found that while there are generally more male-directed films, women are more present in documentary films than in fictional films (Lauzen, 2018c). Writers have also been assigned to work on similar genres (Bielby,
2009). Overall, female crew members are expected to be found in the same genres, as identified by previous research:
H6: Female crew members have a higher presence in comedy, drama, romance and
documentary movies when compared to other genres
Gender and racial representations also may vary by movie studio (Erigha, 2018), as female-directed movies are more likely to be distributed by an independent studio, while male-directed movies were more likely to be distributed by larger studios with a larger reach (Smith, et al., 2015b; Dargis, 2014). However, studies suggest that women are also under-represented in behind-the-scenes positions in the independent film industry (Lauzen, 2018c), and limited to genres with smaller budgets (Erigha, 2015). A report on the Sundance Film Festival, that focuses on independent movies, suggests that male directors outnumbered female directors between 2002 and 2014 (Smith, et al., 2015b). While men still dominate all behind-the-scene jobs in the industry, female representation is expected to be higher in smaller studios:
H7: Female crew members have a higher presence than males in movies produced by smaller studios, when compared to bigger studios.
Interrelation Between the Cast and the Crew
As the top members of the crew produce and design the movie, they might also
influence the on-screen representations and other crew members (Erigha, 2015). For instance, films with female-directors had more female writers, editors, and cinematographers in 2017 (Lauzen, 2018c). Similarly, Lauzen and Dozier (2004) add that writers and directors are more involved in creating higher quality roles for both male and female characters, while producers influence the project in a more global and numeric way. Previous research suggests that the gender composition of the production team influences the on-screen representations. For instance, studies on movies show that the presence of at least one female director or writer
increases the presence of female characters as protagonists, major characters and in speaking roles, when compared to an all-male producing team (Lauzen, 2018a; Lauzen & Dozier, 2005; Smith, et al., 2015b). In line with this, a relationship between the behind-the-scenes and on-screen numerical representations of gender is expected, meaning that the crew has an influence on the cast. This is explored with the following hypothesis:
H8: A higher proportion of women in the crew increases the proportion of women in the cast
Gender Representation and Film Reception
The movie industry is considered to be highly risky, uncertain and unpredictable, as it is impossible to determine with precision what will make a movie successful (Teti, 2013; De Vany & Walls, 1999). Nevertheless, the movies’ storyline, plot, genres, and cast have been suggested to determine its U.S. revenue (Hadida, 2009). Also, in line with previous studies that show that a movie’s budget may be related to its theatrical success (Litman, 1983), Elberse’s (2007) blockbuster strategy suggests that entertainment industries should invest both time and money in products that they consider that might become a hit, for example, by casting big stars (Elberse, 2007).
There are contradicting results regarding if directors or actors contribute more to the success of the movie (Hadida, 2009). Nevertheless, the relationship between the cast and crew and the success of the movie is still unclear, as “it is difficult to draw conclusions about the direction of causality” (Elberse, 2007, p. 102). Although most studies regarding movie performance don’t take into account the cast or crew’s gender, studios can be assumed to make decisions depending on the expected success of the movie. Thus, the
underrepresentation of women suggests that studios don’t associate female cast or crew with higher revenues or critical success. Similarly, a study showed that female acting is less associated with outstanding movies in terms of awards, suggesting that men playing leading roles have a higher impact than women in the same roles (Simonton, 2004). Therefore,
motives for female underrepresentation are examined, as well as the influence of the cast and crew’s gender on the movie’s success:
RQ1: To what extent does the cast and crew’s gender representation influence the film’s performance in terms of awards, ratings and revenue?
Method Sample and Procedure
The current study consists of an applied quantitative automated content analysis, bringing a new approach to studies on gender roles in the movie industry. The current study analyses information from multiple online databases about all 5048 movies that have received a wide release (≥ 600 screens) in the United States between 1982 and 2017. All movies in the current study are considered feature films, as they meet the minimum 40-minute length rule required by the Academy of Motion Pictures, Arts and Sciences award regulations (AMPAS; Rule Two, n.d.). After removing duplicates in the form of re-releases, behind the scenes or other editions of previously released movies, the final sample included 4885 movies. These movies had an average running time of 106.73 minutes (SD = 17.42, range = 63-229). The movies had a wide release in United States each year between 1982 and 2017, with an average of 132 movies per year (44.7% of movies before the year 2000). Movies with all age ratings were considered. While considering that PG-13 was not instated until 1984 (History of Ratings, n.d.), the ratings of the movies in the sample included, R (40.2%), PG-13 (35.9%),
PG (20.0%), G (3.2%), NC-17 (0.2%; n = 1) and missing (0.5%).
In order to collect all relevant data, a Python code was used to scrape movie
information from internet databases. The list of movies with wide releases between 1982 – 2017 came from Box Office Mojo, an online database with box-office data, that provides reliable box-office information for all wide releases from 1982 onwards. Next, the API from the Open Movie Database (OMDB), which provides information identical to the Internet
Movie Database (IMDB), was used to collect most general information, such as release year, genres, production studio, revenue, ratings and the IMDB ID. The IMDB ID is a unique number used to identify movies in IMDB, but is also listed in other databases, providing a connection between them. The IMDB ID was used to access the Bechdel Test Movie List API to scrape information regarding the Bechdel test score and The Movie Database (TMDB) API to scrape the gender of the cast and crew members. While IMDB and the Bechdel Test List have been used before in previous studies (Gosselt, Van Hoof, Gent, & Fox, 2015; Lindner, et al., 2015), OMDB and TMDB have not. However, they have been considered good substitutes to IMDB by non-academic publications (Basu, 2018; Casciato, 2016).
Measures
Cast. The numerical representation of the cast members was measured through the data collected from TMDB. The cast considered in the current study included every actor or actress who had a credited role in the movie and was in the casting list (M = 24.66, SD = 18.41). This means that cast members labelled as ‘extras’ or ‘uncredited’ were not considered in the study. On average, the total cast in the sample was 53.5% male (SD = .17), 26.4% female (SD = .14) and 20.1% unknown (not in the database; SD = .17). Proportions of men (M =.67, SD = .16) and women (M = .33, SD = .16) excluding unknown cast were used in the analysis.
Crew. The numerical representation of the crew members was also measured through data collected through TMDB. Influential positions whose role involves taking major
decisions on who will be hired and how the characters will be represented were analysed: producers (n = 4858, M = 2.18, SD = 1.65), directors (n = 4858, M = 1.06, SD = .33) and writers (n = 4858, M = 1.16, SD = 1.13). First, only general producers were considered (male:
M = .71, SD = .34; female: M = .12, SD = .24; unknown: M = .17, SD = .28), thus excluding
‘location manager’. Second, all credited directors were considered, including multiple
directors when credited (male: M = .88, SD = .32; female: M = .05, SD = .21; unknown: M = .07, SD = .25). Third, regarding writers, only screenplay writers were considered (male: M = .76, SD = .37; female: M = .08, SD = .24; unknown: M = .16, SD = .32). This means that writers credited for the story, characters, comics or books that inspired the screenplay were not included in the analysis. A gender proportion excluding the unknown crew members was calculated for the analysis, showing that most producers (M = .85, SD = .28), directors (M = .95, SD = .22) and writers (M = .91, SD = .26) in the sample were male.
Genres. The first measure for quality of representation is the movie’s genre, which were collected via the OMDB API. As in other databases, several genres are assigned to each movie, which are organised alphabetically (McGregor Olney, 2013; Hsu, Negro, & Perretti, 2012). In the current study, up to eight genres were scraped from OMDB, meaning that a movie could be described with up to eight genres (M = 3.7, SD = .92). From the 23 genres in the sample, drama (n = 2226), comedy (n = 2211), thriller (n = 1399), action (n = 1280), romance (n = 1093) and adventure (n = 1047) were the most common. The other genres had less than 1000 movies each. All genres were used for further analyses, except News because there was only one movie with this genre.
Bechdel test. The second measure for quality of representation was the Bechdel test, that was used as a measure of female character’s independence from male characters. Previous studies have also used the Bechdel test as measures of independence and female representation (Lindner, et al., 2015; Lindner & Schulting, 2017). To obtain the Bechdel test score, the Bechdel Movie List was scraped to identify which movies passed or failed the three Bechdel questions (Are there two women? Who talk to each other? About something other than a man?). Users update the information on the website, which is moderated by the website’s webmaster. Then, the website gives a score to each movie from 0 to 3, depending
on how many questions the movie passes (M = 2.16, SD = 1.03). From the sample in the current study, 54.3% (n = 2655) movies had information on the Bechdel test.
Production studio. The centrality of the representation was measured by the size of the production studio. The bigger and smaller studios were separated into two groups. The Big Six were considered as the big studios, while all the other studios were considered as the smaller studios (n = 2658; 54.3%). While some of the smaller studios are owned by one of the Big Six, they were considered in the current study to be part of the small studio group, as they produce films independently. For example, DreamWorks was coded as a small studio, even though it is owned by Universal. If two studios shared credit for a movie (n = 100), the first one listed was considered to be the main one.
Film success variables. Film success measures were collected through OMDB. First, the amount of Academy Awards (Oscars) wins and nominations. From the movies in the sample, only 5.6% won an Oscar (M = 0.12, SD = 0.64, range = 0-11) and 10.8% were nominated and didn’t win (M = .21, SD = .80, range = 0-11). Second, the estimated visitor count, which was calculated by dividing the total revenue by the average ticket price in the released year (Adjusting for ticket price inflation, n.d.). On average, movies had 8.88 million total visitors (SD = 10.83 million, range = 19,660 - 111.11 million). Third, the online review score provided by users. The IMDB score (M = 6.21, SD = 1.03, range = 1.1 - 9.3) was used for this, as it gives a score out of 10 based on user voting. Fourth, the online review score provided by critics, as measured on Metacritic and Rotten Tomatoes. Metacritic’s metascore (M = 51.35, SD = 17.70, range = 0 - 100) is a weighted average of professional critics without considering the audience, while Rotten Tomatoes’ tomatoscore (M = 47.76, SD = 27.60,
range = 0 – 100) provides the percentage of positive reviews given by professional critics
Stegner, 2018). In summary, six measures to film success were used: Oscar wins, Oscar nominations, total visitors, IMDB rating, metascore and tomatoscore.
Results Cast Representation
Numerical representation. The first hypothesis, that postulated that there is female underrepresentation in the movies, was tested with a paired samples t-test using the proportion of female versus male actors in the cast. This showed a significant difference between the proportion of male (M = .67, SD = .16) and female (M = .33, SD = .16) cast members, t(4847) = -74.35, p = .000, 95% CI [-.35, -.33]. Furthermore, only 12.3% of the movies (n = 597) had more women in the cast than men. Figure 1 and Figure 2 show that although the average proportion of female cast members has slightly increased since 1982 (r = .15, p = .000), the inequality in gender representation has persisted since the 1980s.
Figure 1. Total and average proportion of numerical representation of men and women in the
cast per decade.
Moreover, an ANOVA (F(3,4844) = 32.56, p = .000) showed that there was a significant difference for the proportion of female cast members between the ‘80s (M = .30,
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0 5000 10000 15000 20000 25000 1982-1989 1990-1999 2000-2009 2010-2017 Cas t Pro p o rtio n To ta l Cas t
SD = .15), ‘90s (M = .31, SD = .16), ‘00s (M = .34, SD = .17) and ‘10s (M = .36, SD = .15).
In general, there is an increase of the proportion of female cast across the decades (r = .14, p = .000). A Bonferroni Post Hoc test confirmed significant group differences for all except between the ‘80s and 90’s (p = .610) and the ’00s and ‘10s (p = .065). Hypothesis 1 was confirmed, as women are underrepresented in the cast of the analysed movies.
Figure 2. Total and average proportion of numerical representation of men and women in the
cast per year
Quality of representation. First, a series of independent t-tests using the proportion of female cast as the dependent variable were computed to test the gender differences in each genre and measure the existence of typecasting. As seen in Table 1, each genre was male dominated as they all have more male than female roles. Genres with a significantly higher proportion of female cast, compared to other genres, were categorised as female genres. These were romance, comedy, drama, horror, music, musical, and animation. Female genres had female proportions between .30 and .40. On the other hand, the male genres were action, adventure, crime, war, thriller, history, Sci-Fi, western, sport, documentary and biography. Family, fantasy and mystery were excluded from the genre categorisation because they did not have significant differences. Male and female genres were grouped into two dummy
0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0 500 1000 1500 2000 2500 198219841986198819901992199419961998200020022004200620082010201220142016 Cas t Pro p o rtio n To ta l Cas t
variables for further analysis. Overall, from the genres hypothesised to be male or female, only animation was expected to be male but categorised as female. Even if all genres are male dominated, the genres hypothesized to be female genres did have a higher proportion of women, when compared to other genres, therefore hypothesis 3a and 3b were accepted. Table 1
T-test Results Using Proportion of Female Cast as Dependent Variable
Movie genre N (genre) M SD t Mean difference F P Pass Bechdel test % Drama* 2215 .34 .17 -5.52 -.03 23.58 .000 58.2 Comedy* 2204 .36 .16 -10.99 -.05 2.20 .000 58.8 Thriller 1394 .29 .14 9.89 .05 30.14 .000 48.5 Action 1276 .25 .12 22.77 .11 126.10 .000 41.4 Romance* 1088 .40 .16 -18.08 -.10 4.67 .000 69.1 Adventure 1043 .28 .13 12.74 .07 88.34 .000 48.7 Crime 967 .28 .13 11.57 .07 54.98 .000 40.0 Family 688 .33 .15 0.495 .00 16.07 .620 59.3 Fantasy 688 .33 .14 -0.71 -.00 29.32 .481 58.1 Sci-Fi 597 .29 .12 7.40 .05 73.28 .000 49.0 Horror* 518 .37 .15 -5.78 -.04 3.34 .000 70.9 Mystery 518 .34 .15 -1.47 -.01 3.42 .142 56.0 Animation* 274 .31 .14 2.05 .02 17.82 .041 53.8 Biography 259 .29 .15 3.89 .04 1.17 .000 48.5 Sport 218 .26 .15 6.40 .07 5.60 .000 38.4 Music* 204 .38 .16 -4.26 -.05 .01 .000 68.8 History 151 .23 .14 8.13 .11 7.34 .000 43.9 War 138 .19 .14 10.70 .15 2.93 .000 31.6 Musical* 127 .37 .17 -2.90 -.04 .66 .004 70.7 Western 69 .21 .11 6.30 .12 11.57 .000 32.4 Documentary 43 .20 .25 5.57 .14 17.09 .000 33.3 *Female genres
Next, the analysis of the Bechdel test showed that 30.2% (n = 1477) of the movies with Bechdel test information available (n = 2655, 54%) passed all three questions of the test, while 3.6% (n = 176) failed all three questions, 14.4% (n = 705) passed only the first question and 6.1% (n = 297) passed the first two questions. Overall, as most movies with Bechdel information passed (55.2%), the second hypothesis, which expected most movies to fail the Bechdel test, was rejected. Furthermore, as seen in Table 1, the percentage of movies that pass each question from the Bechdel test varies per genre, where more movies within male
dominated genres tend to fail the test and movies with female genres tend to pass it. Most movies with a female genre pass the Bechdel test (58.5%) while only 48.8% movies with a male genre pass it. A chi-square test showed that the association between female genres and passing the Bechdel test was significant, χ2 (1, N = 2655) = 47.15, p = .000, as well as the association between male genres and failing the test, χ2 (1, N = 2655) = 116.81, p = .000.
Centrality of representation. An independent-sample t-tests was used to test the centrality of the effect of studio size on the proportion of female cast, as stated in hypothesis 4. This test showed no significant difference between big (M = .33, SD = .15) and small (M = .33, SD = .16) studios, t(4846) = 1.29, p = .198, 95% CI [-.00, .01]. Therefore, hypothesis 4 was rejected.
Crew Representation
Numerical representation. The fifth hypothesis stated that there are more men than women working as a) producers, b) directors and c) writers in the behind-the-scene teams. This was tested with paired-samples t-tests. Overall, there was a significantly higher
proportion of male producers (men: M = .85, SD = .28; t(3905) = -79.95, p = .000, 95% CI [.72, .69]), directors (men: M = .95, SD = .22; t(4549) = 138.81, p = .000, 95% CI [.91, -.88]) and writers (men: M = .91, SD = .26; SD = .26; t(2914) = -82.96, p = .000, 95% CI [-.83, -.79]). Furthermore, only 4.7% of the movies (n = 231) had more female directors, 4.2% had more female writers (n = 202) and 6.0% had more female producers (n = 289). As for the cast, Figure 3 and Figure 4 show that even though the average proportion of female crew has increased since 1982 (r = .10, p = .000), gender inequality has also persisted in these positions since the 1980s.
Figure 3. Total and average proportion of numerical representation of men and women in the
crew per decade
Furthermore, three ANOVAs showed that there was a significant difference between the proportion of female crew members between each decade for producers, F(3,3908) = 8.70,
p = .000, writers, F(3,2911) = 3.63, p = .012, and directors, F(3,4546) = 4.72, p = .003.
Correlations also suggest an increase of female crew across the decades (r = .09, p = .000). A Bonferroni Post Hoc test confirmed significant group differences between the ‘80s and all later decades for female producers, between the ‘80s and ‘00s and ‘10s for writers, and between the ‘80s and ‘00s for directors. Overall, as demonstrated by the analyses above, hypothesis 5 was confirmed, as there are more men than women in the crew of the analysed movies. 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 0 1000 2000 3000 4000 5000 6000 1982-1989 1990-1999 2000-2009 2010-2017 Cre w Pro p o rtio n To ta l Cre w
Figure 4. Total and average proportion of numerical representation of men and women in the
crew per year
Quality of representation. The crew’s quality of representation was tested in the same way as for the cast’s quality of representation. Multiple independent samples t-tests using the proportion of female crew as dependent variable were conducted to test whether the male and female crew is assigned to specific genres. As for the cast, Table 2 shows that all genres are male dominated, but some more than others. Genres with a significantly higher proportion of women when compared to other genres were categorised as female genres. Overall, female genres for the crew had female proportions between .10 and .20. These were romance, comedy, drama, and music. On the other hand, the male genres wereaction, crime, thriller, horror, mystery, western, and documentary. Adventure, sci-fi, sport, history, war, family, fantasy, animation, biography and musical were excluded because they were not significant. Male and female genres were grouped into two dummy variables for further analysis. Overall, most male and female genres were the same for cast and crew, except for horror, mystery and biography, that were female for cast and male or not significant for the crew. Moreover, from the genres expected to be female in hypothesis 6, documentary was categorised as a male genre. Furthermore, while all genres were male dominated, the genres
0,00 0,20 0,40 0,60 0,80 1,00 1,20 0 100 200 300 400 500 600 700 800 198219841986198819901992199419961998200020022004200620082010201220142016 Cre w Pro p o rtio n To ta l Cre w
hypothesized to be female did have a higher proportion of women, therefore accepting hypothesis 6.
As for the cast, movies with a female genre tend to pass the Bechdel test (56.8%), while more movies with a male genre tend to fail it (49.0% pass). A chi-square test showed that both the association between female genre and the Bechdel test, χ2 (1, N = 2655) = 4.28,
p = .039, and between the male genre and the Bechdel test, , χ2 (1, N = 2655) = 57.44, p =
.000, were significant. Table 2
T-test Results Using Proportion of Female Crew as Dependent Variable
Movie genre N (genre) M SD t Mean difference F p Pass Bechdel test % Drama* 2178 .12 .21 -6.40 -.04 89.89 .000 58.2 Comedy* 2162 .12 .22 -4.81 -.03 78.87 .000 58.8 Thriller 1384 .07 .15 7.42 .05 173.20 .000 48.5 Action 1264 .07 .15 7.05 .04 154.76 .000 41.4 Romance* 1074 .16 .24 -11.02 -.07 209.99 .000 69.1 Adventure 1017 .09 .17 1.68 .01 16.87 .093 48.7 Crime 954 .07 .15 6.71 .05 126.19 .000 40.0 Family 663 .12 .20 -1.83 -.01 6.53 .067 59.3 Fantasy 669 .11 .19 -0.79 -.01 0.359 .432 58.1 Sci-Fi 593 .09 .16 1.86 .02 20.12 .063 49.0 Horror 512 .07 .14 4.08 .04 50.06 .000 70.9 Mystery 516 .08 .15 3.01 .03 36.86 .003 56.0 Animation 259 .10 .17 0.35 .00 1.47 .727 53.8 Biography 252 .12 .20 -1.44 -.02 1.02 .151 48.5 Sport 212 .08 .16 1.66 .02 4.71 .097 38.4 Music* 203 .14 .22 -2.42 -.03 10.56 .015 68.8 History 147 .09 .16 0.810 .01 3.57 .418 43.9 War 136 .08 .16 1.16 .02 4.81 .247 31.6 Musical 122 .11 .19 -0.08 -.00 0.01 .933 70.7 Western 68 .06 .13 2.07 .05 13.37 .038 32.4 Documentary 38 .03 .08 2.48 .08 23.12 .013 33.3 *Female genres
Centrality of representation. The crew’s centrality of representation was tested in the same way as the cast’s centrality of representation. The independent samples t-tests showed that the relationship between studio size and gender was not significant for producers, t(3904) = .12, p = .908, 95% CI [-.02, .02], and directors, t(4548) =.09, p = .926, 95% CI [-.01, .01]. On the other hand, a significant difference was found for writers, t(2913) = -2.19, p = .029,
95% CI [-.04, -.00]. The average proportion of female writers is higher in bigger (M = .11, SD
= .27) than in smaller studios (M = .08, SD = .25). Overall, as there is no difference for
producers and directors, and the proportion of female writers is not higher in smaller studios, hypothesis 7 was rejected.
Interrelation between the Crew and the Cast
As proposed in hypothesis 8, the crew’s gender is expected to influence the cast’s gender. To test the relationship between the crew and the cast, Pearson’s correlation coefficients were calculated. As expected, there was a positive relationship between the proportion of male producers (r = .10, p = .000), directors (r = .18, p = .000), writers (r = .25,
p = .000) and the proportion of male cast. This suggest that more men in the crew results in
more men in the cast, confirming the expected relationship proposed in hypothesis 8. These relationships were also tested with regressions, where male and female genres were included as moderators. New variables were created for male and female genres, where only genres that were categorised equally for both cast and crew were included. Male genres (M = .49, SD = .50) were action, crime, thriller, western, and documentary. Female genres (M = .79, SD = .41) were comedy, romance, music and drama. Other genres were excluded because they were not considered as male or female genres for both crew and cast.
A regression was computed in SPSS Process using the proportion of female crew as independent variable, the proportion of female cast was used as dependent variable, and the male and female genres as moderators. This model was significant, F(3,4748) = 81.41, p = .000, but showed a weak strength of prediction, R2 = .05. There was a positive and significant interaction between female crew and cast, such that for female genres this relationship was stronger, b* = .11, t = 3.13, p =.002, 95% CI [.04, .18]. The moderation of male genres was also significant, b* = -.08, t = -3.45, p =.001, 95% CI [-.13, -.04], meaning that the
will result in more women in the cast, and this effect is stronger for female genres and weaker for male genres, further confirming hypothesis 8.
Gender Representation and Film Reception
To explore and test the relationship between the cast and crew’s gender and the film’s reception, correlations were computed between the proportion of male cast and crew and the movie ratings (metascore, IMDB and tomatoscore), the film’s number of visitors and the number of Oscars wins and nominations.
As seen in Table 3, all correlations showed significant negative relationships between female cast and the film success variables, suggesting that more women in the cast will result in lower success, measured as online critic and audience reviews, tickets sold, and awards. On the other hand, correlations between the proportion of female crew and the success variables was positive, suggesting that more women in the crew would increase the movie’s metascore, tomatoscore and IMDB rating. While the associations between female crew and Oscar wins and nominations and total visitors show a similar trend, these were not significant. When the behind-the-scenes positions are tested separately, the same relationship was seen for female producers. However, the relationship between success, female directors and writers is negative or not significant. Overall, this suggests that while more men in the film’s cast increases both audience and critics reception, more men in the crew decreases them. Table 3
Correlations between crew and cast gender distributions and success variables
Proportion Female Cast Proportion Female Crew Proportion Female Producers Proportion Female Directors Proportion Female Writers Metascore -.08* .07* .07* -.01 .02 Tomatoscore -.05* .06* .06* -.00 .02 IMDB rating -.12* .03*** .03*** -.02 -.01 Total Visitors -.12* -.00 .01 -.05*** .01 Oscar Wins -.07* .00 -.01 -.02 .00 Oscar Nominations -.04** .02 .00 -.02 -.00 Note: * p < .001, ** p < .005, *** p < .05
Discussion
The current study provides an overview of the gender representation in the cast and crew in the movie industry between 1982 and 2017 and explore the effect of these
representations on the movie’s success. The gender of the cast and crew of 4885 movies was examined with an automated content analysis to determine three dimensions of representation: numerical, quality and centrality. Furthermore, the influence of the crew’s gender on the cast was also tested, as well as their influence on the movie’s success.
The first aim of the study was to address gender inequality in the movie industry. Multiple t-tests and ANOVAs confirmed the prevailing gender inequality in all forms of representation, as identified by recent reports (Lauzen, 2018a; Smith, et al., 2015a; Hunt, et al., 2016). Specifically, the analyses showed that all cast and crew professions are dominated by men, regardless of genre or studio size. Moreover, while an overall increase in female representation since 1982 was found, the most recent decade shows lower proportion of female representation, thus confirming previous reports that have suggested a decrease in female representation in recent years (Lauzen, 2018a; Smith et al., 2015a).
When looking at the quality of the representations, the analysis also confirmed that, although men dominate all genres, both men and women are assigned into specific genres. In the current study, men were especially present in action, crime, thriller, western, and
documentary movies, while a higher proportion of women was found in comedy, drama, romance and music than in other genres. Except for documentaries, that have been previously associated with female crews, these results are in line with previous reports (Lauzen, 2018a; Smith et al., 2015a).
Additionally, the Bechdel test was used as a measure to evaluate the women’s independence from men. To pass the test, a movie has to pass all three questions (Are there two women? Who talk to each other? About something other than a man?). In the current
study, the majority of movies passed the test (55.2%). This finding is similar to results from previous reports of movies passing the test (53%, Hickey, 2014; 57.3% Bechdel Test Movie List Stats, 2018), although other studies report movies failing it (36%, Sharma & Sender, 2014; 44%, Lindner, et al., 2015). As suggested by O’Meara (2016), different types of movies had different Bechdel results. Specifically, female genres (comedy, drama, romance and music) passed the test more often than male genres, as they have a stronger female representation in both cast and crew.
Next, the effect of the crew’s gender on the cast was examined. Men, who are dominant in behind-the-scene positions, were found to cast more men in movies. This
suggests that, to some extent, the previously identified gender inequality is part of a cycle. As there are more men in all top crew positions, there will also be more men in acting positions. This is in line with previous studies, suggesting that more women in the producing team have increased the representation of women on-screen (Lauzen & Dozier 2004; Lauzen & Dozier, 2005).
The second aim of the study was to investigate the relationship between the cast and crew’s gender representation on the movie’s success. A higher proportion of female cast members was found to decrease the movies’ success in terms of both audience (ticket sales, IMDB rating) and critical evaluation (tomatoscore, metascore, awards). These results support Simonton (2004)’s ‘Best Actress Paradox’, that suggests that movies with excellent female performances are less likely to be considered outstanding films than movies with excellent male performances.
This finding suggests that the previously mentioned inequality cycle, where men hire more men, is not an irrational decision, as more men in the cast resulted in higher success. While recent reports suggest that movies with female casts are more successful at the box office than movies with male cast (Buckley, 2018; Female-led films, 2018), the current study
found that more women in the cast reduces the film’s total amount of tickets sold. Therefore, as there are more men in the crew, there will be more men in the cast, especially since the audiences reward it. Similarly, due to the influence of the crew over the cast, it makes sense that there is typecasting for both crew and cast in the same genres.
However, these findings suggest that the inequality cycle can be broken, as a higher presence of women in the crew increases online critic and user scores. Once more women obtain behind-the-scenes positions, the number of women in the cast should also increase, especially for female genres. Future studies could explore the reasons and motivations of both audiences and critics to approve or disapprove specific movies. Due to the large number of external factors that influence the success of a movie, linkage studies (de Vreese, et al., 2017) are required in future research to connect the specific cast, crew or content to an audience or critic rating to establish causality and provide new insights.
Moreover, the limited scope of the current study did not allow the examination of other movie-related information that is also available online. While the focus of the current study was on the gender representation of the cast and crew, future studies could also consider scraping the movies’ plot or scripts to identify topics and a more detailed quality of
representation. This could also be used to improve the Bechdel test as a measure, as it currently does not measure the context of the representation or if it is positive, negative or stereotypical (Lindner & Schulting, 2017; O’Meara, 2016). Future studies could aim to add new and more complex layers to the Bechdel test, in order to raise the standard that is now expected from movies regarding female representation. Similarly, the quality of the
representation should be considered when studying abrupt changes in female numerical representation, as there may be fewer female cast members, but they could have more leading roles than in previous years.
Reference List
About Rotten Tomatoes. (n.d.). Rotten Tomatoes. Retrieved from https://www.rottentomatoes.com/about#whatisthetomatometer
Acevedo Caradeux, D., & Gil Salom, L. (2013). Social representation of gender in award-winner short films in Spain. Procedia - Social and Behavioral Sciences, 95, 126–135. https://doi.org/10.1016/j.sbspro.2013.10.631
Adjusting for ticket price inflation. (n.d.). Box Office Mojo. Retrieved from: https://www.boxofficemojo.com/about/adjuster.htm
Anderson, H. & Daniels, M. (2016). Film dialogue from 2,000 screenplays, broken down by gender and age. The Pudding. Retrieved from: https://pudding.cool/2017/03/film-dialogue/
Aydemir, E. (2017). Use of Hollywood as a soft power tool in foreign policy strategy of the United States of America. International Journal of Humanities and Social Science
Invention, 6(11), 79–83.
Basu, S. (May 23, 2018). The best alternative to IMDB is… The Movie Database. Make Use
Of. Retrieved from https://www.makeuseof.com/tag/best-alternative-imdb/
Bechdel Test Movie List. (2018). Retrieved December 2018, from http://www.bechdeltest.com.
Bechdel Test Movie List Stats. (2018). Retrieved December 2018, from: http://bechdeltest.com/statistics/
Bielby, D. D. (2009). Gender inequality in culture industries: Women and men writers in film and television. Sociologie Du Travail, 51(2), 237–252.
https://doi.org/10.1016/j.soctra.2009.03.006
Buckingham, D. (1993). Sorting out TV: Categorization and genre. In Children Talking
Buckley, C. (2018, December 11). Movies starring women earn more than male-led films, study finds. The New York Times. Retrieved from:
https://www.nytimes.com/2018/12/11/movies/creative-artists-agency-study.html Carlin, S. (2016, April 28). Has Jennifer Lawrence's essay on equal pay changed anything in
Hollywood? Bustle. Retrieved from https://www.bustle.com/articles/154211-has-jennifer-lawrences-essay-on-equal-pay-changed-anything-in-hollywood
Casciato, M. (2016, December 22). No IMDb API? Check out these options. Third party alternatives. Medium. Retrieved from https://medium.com/@mcasciato/no-imdb-api-check-out-these-options-75917d0fe923
Cohen, E. L., Atwell Seate, A., Anderson, S. M., & Tindage, M. F. (2017). Sport fans and sci-fi fanatics: The social stigma of popular media fandom. Psychology of Popular Media
Culture, 6(3), 193-207. https://doi.org/10.1037/ppm0000095
Cooney, S. (2018, September 17). Here are all the public figures who’ve been accused of sexual misconduct after Harvey Weinstein. Time. Retrieved from
http://time.com/5015204/harvey-weinstein-scandal/
Coyne, S. M., Callister, M., & Robinson, T. (2010). Yes, another teen movie - Three decades of physical violence in films aimed at adolescents. Journal of Children and Media,
4(4), 387–401. https://doi.org/10.1080/17482798.2010.510006
Dargis, M. (2014, December 24). In Hollywood, it’s a men’s, men’s, men’s world. The New
York Times. Retrieved from
https://www.nytimes.com/2014/12/28/movies/in-hollywood-its-a-mens-mens-mens-world.html
Davies, C. & Khomami, N. (2017, May 25). Harvey Weinstein: the women who have accused him. The Guardian. Retrieved from
https://www.theguardian.com/film/2017/oct/11/the-allegations-against-harvey-weinstein-what-we-know-so-far
De Vany, A., & Walls, W. D. (1999). Uncertainty in the movie industry: Does star power reduce the terror of the box office? Journal of Cultural Economics, 23(4), 285–318. De Vreese, C. H., Boukes, M., Schuck, A., Vliegenthart, R., Bos, L., & Lelkes, Y. (2017).
Linking survey and media content data: Opportunities, considerations, and pitfalls.
Communication Methods and Measures, 11(4), 221–244.
https://doi.org/10.1080/19312458.2017.1380175
De Zoysa, R., & Newman, O. (2002). Globalization, soft power and the challenge of Hollywood. Contemporary Politics, 8(3), 185–202.
https://doi.org/10.1080/1356977022000025678
Demaria, M. (2017, April 7). These actresses took a stand for equal pay. Refinery29.
Retrieved from https://www.refinery29.com/2017/04/149108/female-actresses-equal-pay
Elberse, A. (2007). The power of stars: Do star actors drive the success of movies? Journal of
Marketing, 71(4), 102–120. https://doi.org/10.1509/jmkg.71.4.102
Eliashberg, J., Elberse, A., & Leenders, M. A. A. M. (2006). The motion picture industry: Critical issues in practice, current research, and new research directions. Marketing
Science, 25(6), 638–661. https://doi.org/10.1287/mksc.1050.0177
Erigha, M. (2015). Race, gender, Hollywood: Representation in cultural production and digital media’s potential for change. Sociology Compass, 9(1), 78–89.
https://doi.org/10.1111/soc4.12237
Erigha, M. (2018). On the margins: Black directors and the persistence of racial inequality in twenty-first century Hollywood. Ethnic and Racial Studies, 41(7), 1217–1234. https://doi.org/10.1080/01419870.2017.1281984
Female-led films outperform at box office for 2014-2017. (2018). Shift7. Retrieved from: https://shift7.com/media-research
Friedman, L., Daniels, M., & Blinderman, I. (n.d.). Hollywood's gender divide and its effect on films: Examining the gender of writers, producers, and directors who make films that fail the Bechdel Test. The Pudding. Retrieved from:
https://pudding.cool/2017/03/bechdel/
Feuer, J. (1992). Genre study and television. In R. C. Allen (Ed.), Channels of Discourse,
Reassembled: Television and Contemporary Criticism (pp. 138-159). London:
Routledge.
Gomery, D. (2008). The economics of Hollywood: Money and media. In A. Alexander, J. Owers, R. Carveth, C. A. Hollifield, & A. N. Greco (Eds.), Media economics: Theory
and practice, 3rd edition, (193-206), New Jersey, USA: Lawrence Erlbaum
Associates, Inc.
Gosselt, J. F., Van Hoof, J. J., Gent, B. S., & Fox, J. P. (2015). Violent frames: Analyzing Internet Movie Database reviewers’ text descriptions of media violence and gender differences from 39 years of U.S. action, thriller, crime, and adventure movies.
International Journal of Communication, 9(1), 547–567.
Grabe, M. E. & Drew, D. G. (2007). Crime cultivation: Comparisons across media genres and channels. Journal of Broadcasting & Electronic Media, 51(1), 147–171.
https://doi.org/10.1080/08838150701308143
Gottfried, J. A., Vaala, S. E., Bleakley, A., Hennessy, M., & Jordan, A. (2013). Does the effect of exposure to TV sex on adolescent sexual behavior vary by genre?
Communication Research, 40(1), 73–95. https://doi.org/10.1177/0093650211415399
Hadida, A. L. (2009). Motion picture performance: A review and research agenda.
International Journal of Management Reviews, 11(3), 297–335.
Hickey, W. (2014, April 1). The dollar-and-cents case against Hollywood’s exclusion of women. FiveThirtyFive. Retrieved from: https://fivethirtyeight.com/features/the-dollar-and-cents-case-against-hollywoods-exclusion-of-women/
History of Ratings. (n.d.). Film Ratings. Retrieved from https://filmratings.com/History How we create the metascore magic. (n.d.). Metacritic. Retrieved from
https://www.metacritic.com/about-metascores
Hsu, G., Negro, G., & Perretti, F. (2012). Hybrids in Hollywood: A study of the production and performance of genre-spanning films. Industrial and Corporate Change, 21(6), 1427–1450. https://doi.org/10.1093/icc/dts011
Hunt, D., Ramón, A., & Tran, M. (2016, February). 2016 Hollywood diversity report: Business as usual? Los Angeles, CA: Ralph J. Bunche Center for African American Studies at UCLA
Lauzen, M. M. (2012). Where are the film directors (who happen to be women)? Quarterly
Review of Film and Video, 29(4), 310–319.
https://doi.org/10.1080/10509201003601167
Lauzen, M. M. (2018a). It’s a man’s (celluloid) world: Portrayals of female characters in the top 100 films of 2017. San Diego, CA: Center for the Study of Women in Television and Film.
Lauzen, M. M. (2018b). Boxed in 2017-18: Women on screen and behind the scenes in television. San Diego, CA: Center for the Study of Women in Television and Film. Lauzen, M. M. (2018c). Indie women: Behind-the-scenes employment of women in
independent film, 2017-18. San Diego, CA: Center for the Study of Women in Television and Film.
Lauzen, M. M., & Dozier, D. M. (2004). Evening the score in prime time: The relationship between behind-the-scenes women and on-screen portrayals in the 2002-2003 season.
Journal of Broadcasting & Electronic Media, 48(3), 484–500.
https://doi.org/10.1207/s15506878jobem4803_8
Lauzen, M. M., & Dozier, D. M. (2005). Maintaining the double standard: Portrayals of age and gender in popular films. Sex Roles, 52(7–8), 437–446.
https://doi.org/10.1007/s11199-005-3710-1
Lawrence, J. (2015, October 13). Jennifer Lawrence: "Why do I make less than my male co‑stars?". Lenny. Retrieved from https://www.lennyletter.com/story/jennifer-lawrence-why-do-i-make-less-than-my-male-costars
Lincoln, A. E., & Allen, M. P. (2004). Double jeopardy in Hollywood: Age and gender in the careers of film actors, 1926-1999. Sociological Forum, 19(4), 611–631.
https://doi.org/10.1007/s11206-004-0698-1
Lindner, A. M., Lindquist, M., & Arnold, J. (2015). Million dollar maybe? The effect of female presence in movies on box office returns. Sociological Inquiry, 85(3), 407– 428. https://doi.org/10.1111/soin.12081
Lindner, A. M., & Schulting, Z. (2017). How movies with a female presence fare with critics.
Socius: Sociological Research for a Dynamic World, 3, 1–6.
https://doi.org/10.1177/2378023117727636
Litman, B. R. (1983). Predicting success of theatrical movies: An empirical study. Journal of
Popular Culture, 16(4), 159–175.
McGregor Olney, A. (2013). Predicting film genres with implicit ideals. Frontiers in
Psychology, 3(JAN), 1–14. https://doi.org/10.3389/fpsyg.2012.00565
Olteanu, A. (2017, April 10). Whose ratings should you trust? IMDB, Rotten Tomatoes, Metacritic, or Fandango? A data scientist investigates. Medium. Retrieved from https://medium.freecodecamp.org/whose-reviews-should-you-trust-imdb-rotten-tomatoes-metacritic-or-fandango-7d1010c6cf19
O’Meara, J. (2016). What “The Bechdel Test” doesn’t tell us: examining women’s verbal and vocal (dis)empowerment in cinema. Feminist Media Studies, 16(6), 1120–1123. https://doi.org/10.1080/14680777.2016.1234239
Rule Two. (n.d.) Academy of Motion Pictures, Arts and Sciences (AMPAS). Retrieved from https://web.archive.org/web/20080905192937/http://www.oscars.org:80/81academya wards/rules/rule02.html
Schatz, T. (2009). New Hollywood, new millennium. In W. Buckland, Film theory and
contemporary Hollywood movies (9–46). New York: Routledge
Sharma, V. & Sender, H. (2014, January 2). Hollywood movies with strong female roles make more money. Vocativ. Retrieved from
https://www.vocativ.com/culture/celebrity/hollywood-movies-strong-female-roles-make-money/
Simonton, D. (2004). The “best actress” paradox: Outstanding feature films versus exceptional women’s performances. Sex Roles, 50(11–12), 781–794. https://doi.org/10.1023/B:SERS.0000029097.98802.2c
Smith, N. M. (2015a, October 13). Jennifer Lawrence expresses anger at Hollywood's gender pay gap. The Guardian. Retrieved from
https://www.theguardian.com/film/2015/oct/13/jennifer-lawrence-hollywood-gender-pay-gap
Smith, N. M. (2015b, October 6). Ashley Judd says she was sexually harassed by a film executive in the 90s. The Guardian. Retrieved from
https://www.theguardian.com/film/2015/oct/06/ashley-judd-sexually-harassed-film-executive
Smith, S. L., Choueiti, M., Pieper, K., Gillig, T., Lee, C., & DeLuca, K. (2015a). Inequality in 700 popular films: Examining portrayals of character gender, race, & LGBT status from 2007 to 2014. Los Angeles, CA: Media, Diversity: & Social Change Initiative. Smith, S. L., Pieper, K., & Choueiti, M. (2015b). Exploring the careers of female directors:
Phase III . Los Angeles, CA: University of Southern California, Annenberg School for Communication and Journalism.
Smith, S. L., Pieper, K. M., Granados, A., & Choueiti, M. (2010). Assessing gender-related portrayals in top-grossing g-rated films. Sex Roles, 62(11), 774–786.
https://doi.org/10.1007/s11199-009-9736-z
Stegner, B. (2018, February 21). IMDb vs. Rotten Tomatoes vs. Metacritic: Which is best?
Make Use Of. Retrieved from
https://www.makeuseof.com/tag/best-movie-ratings-sites/
Teti, E. (2013). The dark side of the movie. The difficult balance between risk and return.
Management Decision, 51(4), 730–741. https://doi.org/10.1108/00251741311326536
The Women Missing from the Silver Screen and the Technology Used to Find Them. (2017).
Google. Retrieved from
https://www.google.com/intl/en/about/main/gender-equality-films/
Toffoletti, K. (2014). Iranian women’s sports fandom: Gender, resistance, and identity in the football movie Offside. Journal of Sport and Social Issues, 38(1), 75-92.
