Media representations can influence people’s attitudes and perceptions regarding sexuality, relationships, and gender roles. Hence, to understand both the prevalence of gender disparities and the perpetuation of stereotypes in media content, media analyses must address two core dimensions of representation: the visibility of women (the extent to which they are represented) and the portrayal of women (how they are represented). Across various forms of audiovisual media content, such as in the television program or in advertising, women continue to be underrepresented (Beckers et al., 2024; Jürgens et al., 2022; Lind & Meltzer, 2020; Matthes et al., 2016); however, when they do appear, they are often relegated to stereotypical gender roles (Collins, 2011). Such portrayals have persisted for decades, despite the expanded roles of and opportunities for women in society since the 1970s, when the first few studies analyzing the media representation of women and men were conducted (Gerbner & Signorielli, 1979). Music videos are no exception to this trend, consistently underrepresenting women (see below) while depicting them in stereotypical gender roles (Wallis, 2011). Such videos tend to objectify women and portray them in submissive roles; in contrast, men are depicted as dominant and, occasionally, (sexually) aggressive (Andsager & Roe, 2003; Conrad et al., 2009; Sommers-Flanagan et al., 1993). In fact, the inclusion of sexual content in music videos has been recognized as an important marketing strategy within the music industry (Cummins, 2007).

Existing research has extensively explored the visibility and (problematic) representation of women in various popular music genres, including punk (Nguyen, 2012), hip-hop (Ward et al., 2013), rock (Hill et al., 2023), country (Andsager & Roe, 1999), and metal (Vasan, 2011). However, most of these studies focus on a single music genre. This carries the inherent risk of assigning blame to particular genres for perpetuating gender inequality and misogyny and further risks confounding counterdirectional trends of popularity with genre. To provide a comprehensive analysis, it is imperative to examine and compare different music genres to distinguish genre-specific depictions from those that permeate the entire music industry. An additional notable gap in the existing literature is the reliance on manual coding, which often results in relatively small sample sizes. This limitation hinders the ability to conduct meaningful comparisons over time, which are essential for three reasons. First, particularly with samples that only cover a relatively short amount of time, the analysis can be biased by artists who had hit songs during the sampling period. Such artists may not accurately represent the broader trends in gender representation and sexualization in music videos. Second, over the last four decades, the theme of gender and equality has persistently resonated in both society and the entertainment industry, driven by movements such as #MeToo. Investigating music videos over time enables us to analyze how these societal shifts are visually reflected. Third, the distribution of music videos has significantly evolved. While they were primarily distributed in a linear fashion through big TV stations such as MTV in the past, today, they are frequently viewed and shared on social media, online platforms, and streaming services (Beuscart et al., 2023; Edmond, 2014). This may have affected content adaptation, allowing not only TV stations but also the audience itself to shape the content of popular videos. For example, lower-budget videos or those featuring progressive gender images might gain eyeballs through viral sharing. Thus, this evolution could have affected the popularity of certain video types or genres as compared to others. Our study seeks to address these gaps by conducting a comprehensive automated quantitative analysis of music videos in various music genres spanning four decades from 1980-2020. This enables the consideration of both temporal changes and variations across genres.

In the digital age, music videos are easily available to a global audience, with popular artists serving as role models for young viewers. Music videos might shape notions of gender that have an impact on society far beyond artificial or artistic staging. Moreover, they could be crucial drivers of how (young) viewers perceive and engage with notions of masculinity and femininity, societal roles, relationships, and sexuality (Ward & Grower, 2020). Hence, the first section here draws on two key perspectives: social constructionism and the objectification theory to contextualize how music videos shape and reinforce gender roles. The subsequent sections analyze empirical data, focusing on the gender ratio in music videos and the sexualization of women across genres and over the decades.

This section first discusses the visibility of women in music videos and then their sexualized depiction. Both aspects are considered first in absolute terms and then by genre and over time. Finally, we analyze the influence of artist characteristics with regard to both outcomes. In terms of the mere visibility of women, they constituted 34% of the visible people in music videos in our sample. Their average share was substantially higher in dance (46%), R&B (43%), and pop (40%) music videos. In most other genres, the visibility of women was in line with the general average, with the notable exception of rock music videos, which showed the lowest proportion of women overall (27%). In terms of the visibility of women in music over time (see Figure 2), we found that women were less visible in music videos in the 1980s (approximately 20% or less). Moreover, the proportion of visible women in music videos rose continuously between 1980 and 2000, remaining at a fairly stable level of 35%–40% thereafter. However, at no time were men and women equally visible in music videos. This pattern is more or less consistent across all genres, with a proportionate decline in the number of visible women since 2010, particularly in the genres of R&B, rock, and rap, However, a continuous increase was evident in the dance genre in particular (see Figure 2); from the late 1990s to the early 2000s, and since 2015, more women than men were visible in the dance genre, thus making it the only genre in which women were over-proportionally visible in music videos in our dataset.

However, the greatest explanatory factor for the visibility of women in music videos is not due to differences between music genres but in the characteristics of the artists. By far, the gender of the artist in the music video explained the most variability in the visibility of women. A substantially higher proportion of women were visible in music videos by women artists over the entire time span of the study. Moreover, the (apparent) ethnicity of the artists appeared related to the visibility of women in music videos: the proportion of visible women was lower in Black artists’ videos, and decreased sharply since 2010 onward. Given the different dynamics both in terms of genres (i.e., the share of rock music videos declined substantially since the 1980s, while rap videos became much more prominent; see Figure A1 in the appendix) and artists (i.e., the rising prominence of Black artists in the 1990s) and the interrelations between artist demographics and their musical genres, the overall patterns are difficult to assess using only graphical or bivariate analyses. Put differently, it is difficult to say how much changes in the visibility of women are due to changes in the composition of the music charts in terms of artists or genres or change within certain genres or among certain groups of artists. Therefore, we estimated a logistic multilevel regression model, the results of which are displayed in Figure 3. The coefficients (average marginal effects) can be directly interpreted as differences in the probability of women being visible. The results reveal that artist gender was, by far, the strongest predictor of visibility, with an estimated difference of 0.30 between men and women artists. Figure 3 also reveals that women were significantly less (-0.07) visible in rap videos and significantly more visible in dance (+0.03) and pop (+0.05) videos. After accounting for artist gender and music genre, there were no significant effects of artist ethnicity on the visibility of women in music videos.

In a second step, we examined not only whether women are made more visible in music videos but also in what context this is done. Specifically, we examined the degree of sexualization through a sexual objectification score that encompassed (1) revealing or suggestive clothing, (2) sexually suggestive poses and facial expressions, and (3) sexually suggestive moves. Overall, approximately 10% of all coded video frames included a sexualized depiction of a woman. The sexual objectification of women was relatively low in the 1980s and beginning of 1990s. In particular, in the mid 1990s, there was a strong increase in women’s sexualization, which reached its peak around the year 2000. Since then, there has been a gradual decline, although the overall level of sexualization remains high, with slight fluctuations over time (see Figure 4). Further, genre-specific analysis reveals substantial variation in the degree of sexualization across genres, with dance, R&B, and rap music videos consistently exhibiting the highest levels of sexualization over time. In particular, dance music continues to show an increasing trend in the sexualized depictions of women, even in recent years, while this trend in rap and R&B peaked around 2000–2010 and has slightly declined since then. In contrast, country and rock music videos display the lowest levels of sexualization, with rock videos exhibiting a clear downward trend after 2010. Notably, the gap between pop and other highly sexualized genres has narrowed, thereby suggesting that pop music videos—which initially had lower sexualization levels—have become more aligned with trends in R&B and rap, particularly during the late 1990s and early 2000s. Further, women artists’ videos consistently exhibited higher levels of sexualization compared to those of men artists, with the trend peaking around the early 2000s before experiencing a slight decline but still remaining at a high level. Notably, the gap between men and women artists in terms of sexualization has persisted over time. While the (visible) ethnicity of the artists had little influence in the overall visibility of women, their sexualization score differed between Black and White artists—the abovementioned peak in sexualized depictions of women from 2000 to 2015 was more pronounced in Black artists’ videos, but has since then converged in terms of the overall average, thus suggesting a shift in industry-wide representations.

Again, the bivariate descriptions in Figure 4 cannot reveal the complex interrelations between genre and artist characteristics, which tend to be correlated and might follow different trajectories over time. As mentioned above, we estimated a logistic multilevel regression model, with the sexual objectification score as the outcome variable and genre as well as artist characteristics as predictor variables. In addition, we controlled for the proportion of visible women in the video, since, naturally, our sexualization score is related to the overall visibility of women in the video. The results displayed in Figure 5 reveal that, first, videos by women artists show significantly higher levels of sexualization compared to those by men artists, with a positive and statistically robust effect. Second, while Black artists barely differ from the baseline, the probability of women’s sexualization in videos by White artists is .025 lower than that of videos by artists of other races. Two genres exhibited markedly lower (rock and country) and one higher (dance) women sexualization scores. Overall, these differences were rather small—that is, in the range of .01 to .03 on the probability scale.

Finally, we examined the intersection of artist gender and race to better understand how these characteristics combine to shape both the visibility of women and their sexual objectification in music videos. Figure 6 presents the predicted probabilities of these two outcomes. Women are significantly more visible in music videos by women artists compared to those by men artists, regardless of race. However, an interaction effect emerges when examining sexual objectification as an outcome—for women artists, sexual objectification levels are high regardless of ethnicity. Non-White men artists’ videos feature higher levels of sexual objectification of women than those by White men artists. White men artists’ videos contain the lowest levels of sexual objectification. These findings suggest that the overall lower sexual objectification scores found in White artists’ videos are driven by White men artists.

In this study, we analyzed the representation of men and women in music videos as well as the sexual objectification of women in music videos. Using state-of-the-art computer vision models, we analyzed 9,799 videos for Billboard 100 songs from 1980 to 2020. Across all videos, men in music videos outnumbered women by a ratio of 2 to 1. This is in line with other studies (Aubrey & Frisby, 2011; Turner, 2011). Our analysis also revealed an increase in women’s visibility in music videos over time, which is also in line with existing research (Aubrey & Frisby, 2011; Seidman, 1992; Sommers-Flanagan et al., 1993). Despite this upward trend, equality in visibility between men and women was never achieved. Notably, the pop genre stood out with a significantly higher proportion of visible women compared to men, thus highlighting a deviation from the general trend. In contrast, genres such as R&B and country music exhibited the lowest visibility of women in music videos. Again, this is in line with existing research (Aubrey & Frisby, 2011). However, critically speaking, this also implies that the results of the studies from the 2010s have changed very little in terms of the visibility of women in music videos. With regard to the exploration of sexualization, we identified fluctuations over time, peaking around the 2000s. Dance, R&B, and rap genres were found to have the highest levels of sexualization (Aubrey & Frisby, 2011; Conrad et al., 2009), with the level of sexualization in the dance genre showing a sustained increase even in recent years. Thus, our results contribute to a crucial reassessment of longitudinal trends and indicate that sexualization does not follow a linear trajectory (in either direction). However, there appear to be more similarities than differences among music genres when it comes to the sexualization of women. This is in line with other comparative research on music videos (Aubrey & Frisby, 2011) and emphasizes that we are dealing with a sexualized industry rather than individual “problematic” genres in this industry. Music videos, as an integral component of the music industry, serve as reflections of larger trends and structures within the realm of popular music. The sexual objectification of women in music videos reflects broader systemic issues within the music industry and align with broader trends across media, such as the increased sexualization of women found in analyses of Rolling Stone magazine covers over time (Hatton & Trautner, 2011). Our findings further suggest that these disparities are not only gendered but also racialized in men artists’ videos. Hence, the gender disparities observed in music videos are rather likely not isolated occurrences, but are intricately linked to the prevailing imbalances within the industry itself. Research by Sergeant & Himonides (2023) reveals a stark gender asymmetry, with men performers significantly outnumbering women performers, not only in front of the camera but also in leadership roles within musical groups. This indicates that the gender bias extends beyond the visual representation in music videos and originates from the initial composition of performers. Moreover, the disparity may be compounded by behind-the-scenes factors, as highlighted by Andsager & Roe (1999), who found a predominance of men directors in their sample of music videos. Recognizing these systemic issues emphasizes the need for a comprehensive and industry-wide approach to address gender imbalances. Future research could include the gender of directors and artist managers in the analysis to broaden the picture painted in this study. While the greater representation of women in music videos can be viewed as a step toward inclusivity, the concomitant escalation in sexualization raises concerns regarding the perpetuation of objectification and stereotyping. It is worth noting that our study did not assess the sexualization of men in music videos, but existing research suggests a notable asymmetry, with men generally not experiencing the same degree of objectification (Andsager & Roe, 1999; Hatton & Trautner, 2011). Hence, visibility does not automatically translate into equality. Further, the most influential factor impacting the visibility of women in music videos was not associated with individual genres but rather with the artist’s characteristics. Women artists consistently exhibited higher visibility and, simultaneously, higher sexualization of women throughout the study period; in contrast, men artists, although not significantly correlated, tended to have lower visibility. However, our analysis of the interaction between the artist’s gender and race revealed that the lower sexualization scores in white artists’ videos are primarily driven by white men artists. This suggests that racialized gender norms play a role in shaping how women are portrayed in men artists’ music videos, while sexualized portrayals in women artists’ videos appear to be an industry-wide expectation rather than a racially differentiated phenomenon. These quantitative findings should not be interpreted as suggesting that White men artists are inherently more feminist or progressive. Rather, the observed patterns may reflect broader structural, genre-specific, or commercial influences, and not necessarily intentional choices by the artists themselves. Our results reflect broader cultural norms tied to “doing gender” (West & Zimmerman, 1987), according to which performances of femininity and masculinity are shaped by social expectations, including those perpetuated in music videos. Women artists’ greater visibility, coupled with heightened sexual objectification, exemplifies how women in the music industry are expected to embody and amplify traditional notions of femininity, often emphasizing physical appearance and sexual allure as part of their public image. This dynamic also ties into the objectification theory (Fredrickson & Roberts, 1997), which emphasizes how the frequent portrayal of women as objects of sexual desire not only reflects societal norms but also fosters self-objectification. In music videos, women artists often play an active role in their own sexualized portrayal. A few feminist scholars have argued that this can be interpreted by audiences as a form of empowerment, where women are reclaiming their agency and sexuality in the public eye (Frisby & Aubrey, 2012). However, others point out that these portrayals may simultaneously reinforce systemic norms of sexual objectification, as success within the music industry is often tied to conforming to such expectations (Linden, 2004; Butler, 1990). This reflects a tension between agency and structural pressures, thereby highlighting the blurred lines between celebrating women’s sexuality and reinforcing cultural norms that objectify them. Future research could build on these findings by exploring how audiences perceive sexualized portrayals of women artists, particularly in terms of empowerment and objectification. Experimental studies could investigate whether such portrayals challenge or perpetuate traditional gender roles.

Of course, our study does not come without limitations. First, this study primarily focuses on the visual elements of music videos, and it is important to recognize that these visual aspects may convey messages distinct from the musical content (Thompson et al., 2005) or lyrics. Although it appears attractive to simultaneously address both visual and textual messages, sexualization in lyrics represents a more complex research subject due to irony, ambiguity, references and even the intertextual use of patterns. Future research could explore the intricate interplay between visual and auditory elements to provide a more comprehensive understanding of the impact of this interplay on audience perceptions. Second, our study only examined the visibility of men and women, but it does not go beyond that. It aims at providing a comprehensive longitudinal overview; however, in the process of doing so, it ignores the nuanced aspects of display, such as nonverbal behaviors (e.g., body movement, gestures, facial expressions) (Wallis, 2011), differences between lead and supporting roles, including props or dancers, and whether they are portrayed as being subject to sexual violence (Hill et al., 2023; Sherman & Dominick, 1986; Sommers-Flanagan et al., 1993). A more nuanced examination of role distinctions could provide a deeper understanding of how gender dynamics manifest in different capacities within music videos. Further, although our operationalization of sexual objectification includes multiple dimensions—such as revealing clothing, sexually suggestive poses, and movements—it is not exhaustive. The subtle but significant aspects of sexual objectification, such as referential gaze and decorative roles, remain unaddressed. These elements could provide additional nuance, particularly in terms of understanding the broader cultural and societal implications of objectification. Given that existing literature suggests that men are not subject to the same degree of sexualization (King et al., 2006), this represents a noteworthy gap in the analysis. Future research should aim to address this disparity to comprehensively capture the spectrum of gender portrayals and sexualization in music videos. Additionally, our findings highlight that racialized gender norms play a crucial role in shaping the representation of women, particularly in men artists’ music videos. Future research should explore the underlying factors that drive these racialized gender patterns, including industry expectations, genre conventions, and audience reception.

Given the evidently sensitive nature of the subject of this paper, we conclude with a few considerations of the underlying ethical problems—both to transparently discuss our own approach and to appreciate the implicit decisions and their normative context. Automated analysis of biometric data, of faces in particular, involves highly sensitive personal information and enjoys special protection under existing legislation. Therefore, we wish to highlight that regardless of nonbinding ethical guidelines, researchers today are already bound by legal duties arising from norms, such as the General Data Protection Regulation (GDPR) of the EU (Jasserand, 2024), which were expanded further under the new AI Act (1, 2024). It is important to note that the regulation and its embedded ethical concerns primarily address the most sensitive aspect of facial recognition—that is, the identification of identities from facial patterns. Identity-agnostic biometrical data, such as the expression of emotions, are considered less sensitive and, thus, require weaker protections (Häuselmann et al., 2023). Under the new provisions of the AI Act, facial analysis remains widely permitted, given that it obeys two prohibitions (26b): AI may not be used to (1) create or expand facial recognition databases using (2) untargeted scraping. Our work with non-PII data and publicly available (broadcast) video material independently satisfies both criteria (even though one would be sufficient). Turning to a more comprehensive ethical perspective, we further considered the unintended implications of the methods and analysis itself. Both the theoretical framework and the empirical analysis should clearly convey our critical assessment and problematization of the prevalent “male gaze” (Oliver, 2017). Thus, it bears a certain irony that the LLM used for our composite sexualization score itself is partially trained on material that replicates this phenomenon. Although we neither alter (train) the neural network nor distribute its input material, there is an indirect exploitative element to the use of extant objectifying images. The use of readymade tools still appears to be an ethically superior choice in comparison to the use of student annotators for explicit imagery. Finally, it appears prudent to reiterate a significant limitation discussed above: Analyzing visual presentations always bears the risk of inviting simplified and superficial interpretations of responsibility. We must assume that the music industry, as a particularly power-driven environment, exerts an important influence on the production of music videos. Therefore, a measured (self-)objectification of women should not be taken as clear-cut evidence of the causal responsibility without taking into account the role of producers, directors, marketing specialists, etc. In this study, our broad empirical analysis revealed that objectification is prevalent and exhibits interesting variations across time, genre, and gender; however, unveiling its causes certainly requires significant future research.