Written by Deb Verhoeven
Data Analysis and Visualisation by Pete Jones
In this second post in a 4-part blog series, the Kinomatics team continues its deep look at a recent, contentious tally of Australia’s Top 50 Films (of all time), as an example of what we call “Top Listing”. Top Listing is when an organisation or commentator creates and publishes an ordered, delimited ranking of phenomena (films, books, hamburger joints, shades of blue and so on). Typically this takes the form of media reports promoting an incessant stream of annual Top Tens but it could also include specialised lists of “the world’s most livable cities” or “best performing universities”. Top Listing well precedes the internet but has burgeoned in an algorithmic era that relies on low-hanging clickbait to drive engagement.
PART TWO: KINOSSEURS OR RANK HAPPY DUDES?
In the previous post in this series we looked in detail at the group of people who voted for the Top 50 Australian movies (of all time) and how their shared film preferences inadvertently created an imagined coterie of “kinosseurs”.
The slightly buried point in that post, if I can momentarily yoke the fabulous expert of rankings, Wendy Espeland, to this specific topic, is that Top Listing creates new neighbours.
Wendy mostly meant this at the level of what is ranked. So for the purpose of this discussion – Top Listing sweeps up films that would not otherwise, in any mapping of Australian cinema, be located close-by and brings them to the same side of the curb. In what world would we find ourselves discussing Samson and Delilah (#1) in the same breath as Gallipoli (#2)?
But this idea of the unexpected proximity produced by Top Listing could just as well be about the voters. For example, I never for a moment would have predicted that the person whose film choices aligned most closely with mine (a white, gender apathetic, recovering academic) would be Warwick Thornton (as far as I am aware, none of the above). Warwick and I also happened to be among the few outliers in the voting. Unlike us, the majority of voters, in all their variety, broadly concurred on their film choices.
Perhaps then, since distinctions of taste were barely evident between most of the voters, are there other ways to distinguish between the different Top Listers?
In a conversation in one of my social media posts about the Top 50 list, Australian media academic Rachel Wilson asked if we could look at whether films directed by men were particularly voted for by men. There are a few ways to look at this and before I reveal the data I want to talk a bit about how we approached gender for the purpose of this analysis.
First a personal admission. I am – and I think always have been – gender apathetic. I don’t really care about what gender people think I am. Since I was a child people have addressed me using various gendered salutations. Gender is not an important part of my identity and I am fundamentally indifferent to the idea of gender categories in general (and binary ones in particular). Despite my personal apathy I am sensitive that for many people I love and respect, gender provides an important and defining sense of their identity.
I am also a feminist and I am painfully (and personally) well aware that gender serves a function within patriarchy to establish and enforce social hierarchies typically through a binary form of classification. A substantial part of my academic career and film industry advocacy was devoted to figuring out ways to redress these gendered systems of domination, always mindful of how they are interleaved with white supremacy, heteronormativity, ableism, classism.
Which brings us to this analysis. I want to answer Rachel’s question. I want to show the way voters with he/him pronouns – which, in this case, is the single aspect of their gender that we can measure/observe – have a tendency to vote for films directed by people who use he/him pronouns. In doing this I am not for a moment trying to suggest that gender is binary, that pronouns always indicate gender identity (language options for this kind of work are not the friend of nuance), and I want to acknowledge that the lived experience of gender is complex and for many people confounding. I want to expose the way the system works at the same time as interrogating its terms. Tricky.
I am proceeding then with all the caveats above and am open to feedback about how our Kinomatics analysis can better articulate the ways the data shows voter bias and the way one gender in particular reinforces its domination systemically.
How we ascribe gender in our analysis
For this analysis we use the pronouns provided by voters in their biographical note or professional web pages and by filmmakers on their IMDb and/or Wikipedia pages, in reputable media interviews or personal websites. In almost all cases these were he/him or she/her. There were a couple of exceptions including me and Soda Jerk who previously have been gendered by the media but more recently have avoided gendered language in their published presence. On this basis – and because we were primarily interested in the behaviours of people using he/him pronouns we have produced visualizations with two categories – he/him (who we call men) and she/her/they (who we call women & gender minorities – WGM). None of this is intended to suggest that individuals personally identify as “men” or “women and gender minorities” – we are deliberately and possibly erroneously drawing a proxy category from pronouns.
One other thing to note: There’s a film directed by both a man and a woman. So it counts as a film directed by a WGM filmmaker but not as a film that isn’t directed by a man. This makes it look like the data doesn’t always add up (but it all makes more sense once you are aware of this characteristic).
What we found: Drawing the (Plimsoll) Line for SS Patriarchy
At the “headcount” level we can see that films directed by men (he/him) overshadow the Top Listings. (Figure 1). This is especially true for the films that made the Top 50: whilst two thirds of nominated films were directed by men that increases to 80 % of films in the Top 50.
Table 1 shows that the proportion of male directors in the data (rather than their films) is lower. This is not necessarily an equity improvement. The reason this number looks marginally more equitable is because many male directors have more than one film in the Top 50 (8 men have multiple entries compared to 2 women and gender minorities). Peter Weir has two out of his three Top 50 films in the Top 3 alone. I will talk more about this in the next post but for here the thing to note is that in terms of proportions of individual directors themselves (rather than films), of all the directors in the Top 50, 76.9% are men (30 out of 39; counting Soda Jerk as 2 directors). So select male directors are even more concentrated in terms of their influence than the statistics might at first suggest.
We can measure these stats against the prevalence of male directors and male-directed movies in the Australian production industry more generally. The currently published version (1.1.0) of the Kinomatics Australian Film Production Dataset (KAFPD) shows that since 1975 just a tad under 77% of Australian films were directed by someone with a man’s name and around 70% of all the credited directors in the dataset use names commonly attributed to men (see Table 1).
| All nominations | Top 50 only | Industry Generally (KAFPD) | |
| Male Directors† | 65.1% | 76.9% | 71.4% |
| Films Directed by Men† | 68.3% | 80% | 76.9% |
Table 1: Predominance of men among directors in the nominated films, benchmarked against the wider industry.
† It is important to note that gender was approached differently to produce these industry wide figures. For each credited director in the KAFPD, we cross-referenced their given name against an official dataset of baby name registrations. When at least 95% of the babies with that name are assigned male at birth, we tag that name as a man’s name, and vice versa for babies assigned female at birth. All other names are tagged as ambiguous names. Then, every director in the KAFPD is tagged as either having a man’s name, or a woman’s name, or an ambiguous name. It should be noted that the number of directors with a man’s name is lower than the true number of male directors in the sense that the method is both conservative (95% is a high threshold) and can’t account for nicknames such as Dave or Bob that are rarely the names that a baby is registered with.
There are two things I want to explore here. The first notable observation is the increased gender skew from the nominations to the final chosen Top 50.
The top 50 is significantly more dominated by men’s films (80%) than the overall nominations list is (68.3%). As Pete pointed out to me – of the 54 films that were nominated but do NOT appear in the top 50, only 31 (57.4%) are directed by he/him directors. So the periphery has much more space for films that aren’t by men, but when it comes to consensus it’s a dudefest.
The second thing to consider is the way the Top 50 compares to the industry more broadly. In the KAFPD – which describes Australian film productions and crews since 1975 – approximately 77% of films are directed by people with names that are typically assigned to men (which is probably a slight undercount given we don’t count ambiguous names that might apply to different genders). This percentage, 77%+, is uncannily close to the Top 50 at 80%.
For some people, seeing this lack of significant statistical variation in the domination of male-direced movies might suggest the Top 50 is “representative” of the Men-WGM gender breakdown in the industry as a whole. You can hear the pushback already: “Great! Nothing to see here – the meritocracy is neutral and the Top 50 simply reflects the status quo (however inequitable that might be).”
But if we see Top Listing as a constitutive part of the industry – rather than apart from it – then perhaps a different reaction is called for: “Woah – look how almost exactly the Top 50 reiterates and naturalizes the overwhelming domination of one gender – namely men.”
I have in the past talked about the way headcount gender statistics in a variety of film industries seem to land at a similar level despite the very different industrial and cultural settings for these statistics. This led me to speculate about how these weirdly consistent numbers could lead to premature conclusions about a universal numerical “comfort zone” in which a statistically consonant level of male domination, usually somewhere between 70 and 80%, becomes a safety-net for a kind of tolerable amount of systemic sexism; a Plimsoll Line for SS Patriarchy if you will.
Underneath all this preclusive equilibrium, underneath the apparently agreeable surface of aggregate statistics lies a brooding ocean of hidden rips and countercurrents. Let’s take a dive into how voters floated their votes.
Who Top Listed Men
Back to Rachel’s question then, specifically who voted only or mostly for male directors?
In Figures 2 and 3, voters are ranked by how many films they selected that were directed by men and then how many they voted for directed by WGM. Two men and two WGM voters did not select a single film directed by a WGM filmmaker (Figure 3).
The vast majority of voters proposed Top Ten lists that exceeded the proportion of male directors in the industry (approximately 77%). Out of 24 voters, 15 people voted for 8 or more films by men in their Top Ten (see Figure 2). Four of those voters could not find a single film helmed by a woman or gender minority director to vote for and seven voters gave a tokenistic nod to one such director only (Figure 3).
Gillian Armstrong stands alone as the only voter who gave all her votes to WGM directed films (remembering we can’t quite work out which film she meant by Black Dress so consequently we are only counting 9 of her nominations). Adrian Danks and I both nominated 6 films directed by women and gender minorities. What is interesting about Adrian’s nominations is that the films he selected bifurcate exactly on both gender and genre lines. All the films he selected by women are short experimental or essay films and often autobiographical works (In This Life’s Body, A Song of Air, The Illustrated Auschwitz, Memories and Dreams, Night Cries, The Silences). All the films he selected that were directed by men are feature length (Walkabout, Ten Canoes, Newsfront, Sunday Too Far Away) and are characterised by heroic or originary narratives and almost all have received critical accolades such as Best Feature Film awards.
To answer Rachel’s question about whether there is a tendency for men to support male directors we need to look at the distribution visualisations (Figures 4 and 5) which show relative proportions (i.e. tendencies within a group) rather than an exact number of votes.
If you look at the distributions you can see how voters with he/him pronouns (the red line) tended to vote for films directed by people using he/him pronouns and tended to overlook films directed by people who use she/her or do not specify a specific binary gender.
In Figure 4 we can see a higher proportion of he/him voters created Top Ten lists dominated by directors who use he/him than voters using she/her/they. Similarly, at the other end of the distribution, not one he/him voter produced a Top Ten list that excluded films directed by men (where n=0). Another way to analyse this is to say, male voters had a tendency to disregard films directed by women and gender minorities (Figure 5) and if typically when they did include them the effort was tokenistic. Take your pick – it’s not a good look either way.
Conclusion: Floating your Listing Vote
In the first post in this series, we developed a line of thinking around the way Top Listing encourages speculative aspirations: by voting and reacting to other people’s votes we are asking, what kind of person do I want to be? But I would argue, when we are Top Listing, we are also wondering what kind of industry do I want to see? This is the “perfect circle” to run away with Manohla Dargis’s pithy abstraction about film rankings – the veiled way Top Listing actively adds more spin to the cultural hierarchies of which it is a part.
Top Listing is a labour that makes hierarchies look effortless and domination look natural.
Forming a Top Ten is not just a response to what has already been made but imagines the kind of films we want to see. And those films, in this particular Top 50, are overwhelmingly male-directed ones.
To be very clear. I am not suggesting that anyone in the Top Listing process voted with a misogynist motive.
Instead, I am suggesting something much more foundational and for my mind, much more worrying: namely, no motive was necessary.
The bloc voting for men’s films by men (and some women) is not the result of a personal lapse in critical taste, neither is it a limitation in the Top Listing process, nor a failure of the film industry itself. This is the system working as it was designed to. A ranking infrastructure (including Top Listing, annual film awards, and film funding) that switches men’s directorial opportunities into individual signature success will, given enough repetition, simply continue to advance the cohort it already and repeatedly rewards: men.