Back to selection

The Data Says, “We Have a Problem”: Finding Solutions for Film’s Problems with Diversity

Oscar Class of '16. Photo: Image Group LA / ©A.M.P.A.S. Oscar Class of '16. Photo: Image Group LA / ©A.M.P.A.S.

Since the publication of “The Data Says, ‘We Have a Problem'” in our Winter print edition, the conversation around diversity and the movie business has become louder and even more urgent. As more and more studies are published detailing Hollywood’s biased hiring practices and hashtags like #oscarssowhite explode across social media, now, during the lead-up to the Academy Awards, is an apt time to unlock from our paywall this article by Esther Robinson. It cogently articulates the reasons why all of us must care about these issues before it then goes on to offer actual and practical solutions that we can agitate for and even, on an individual level, enact. — Editor

The cool thing about data is its concreteness. It can exist separate from what we feel or intuit, and it can tell us its own story through unambiguous measurement. As such, science and data can be useful tools for seeing our real world more clearly. And, when it comes to filmmaking, examining the data around its creation and distribution can help us make choices that better align with our individual and collective visions for its future.

American society is grappling with big changes. As we move toward a nation of plurality with no monolithic culture dominating (America’s children will reach this balance point by 2020), we as film and media makers need to imagine new ways to express our collective artistic output. And to do that, we need to look at the data, and what it has to say about diversity. And bias.

For example: Of 347 films in the DGA’s inaugural Feature Film Diversity Report, 82.4 percent were by white male directors.

Or: In the 2014-15 Independent Women Employment on Independent Films report by the Center for the Study of Women in Television and Film, men comprised 71 percent of directors working on documentaries screening at more than 20 high-profile film festivals in the United States

Taken together, these and other studies quite forcefully say: the film industry has a serious issue with race and gender bias.

Who we are in American cinema should reflect who we are as a nation — both in subject matter and in core creative positions — but as of this moment, that simply isn’t the case. This is true in narratives and docs alike. We can see it in the data of who gets funded, who screens at festivals and who gets picked up for commercial release, and in studies like the ones above and others coming out of the Media, Diversity & Social Change Initiative at the Annenberg School for Communication & Journalism at the University of Southern California, the Center for the Study of Women in Television & Film at San Diego State University, the Sundance Institute and the Ralph J. Bunche Center at the University of California, Los Angeles.

In terms of the hard numbers, what gets supported (funded, made, screened and promoted), often even within the nonprofit system (a system set up to support public good), is predominantly white and male. And sometimes you don’t even need data; you just need eyes. Our eyes can show us that in aggregate, the decision-makers are predominantly white, and they are, more often than not, male. Take a look at event photos from any industry magazine (even this one!) and you’ll get a snapshot of how we’ve failed to achieve diversity in this regard.

Choosing to reward a person or a film with funding, a festival slot or an award is a beautiful gift. It is also often a hidden process. The factors that influence what gets chosen in any adjudication process are nuanced and subjective, and individual context can be hard to parse out. A given person or panel’s green-light choices on any given day can seem mysterious, even to those of us that help build and run those systems. And while flagrant discrimination seems rare, it’s long been a puzzle to me how a film industry made up overwhelmingly of people with progressive values around race and gender has such a terrible track record on diversity.

Then I heard about a recent study on an old phenomenon called “In-Group Favoritism,” or “In-group Bias,” and a new understanding started to shake free. Social scientists studying in-group decision making have long observed a pervasive phenomenon: groups unconsciously privilege “their own kind” when deciding how to distribute resources.  And this is across the board. It’s something ingrained in human behavior and  studied for over 50 years.

But in 2014 a duo of social scientists, Tony Greenwald and Thomas Pettigrew, turned up a new wrinkle in this long-acknowledged phenomenon. In a theoretical review for American Psychologist, they examined 50 years of experiments and survey methods from published scientific research on discrimination. They found that discrimination occurred far more frequently in decisions aimed at helping people than those with the potential to actually harm someone.

According to Greenwald, “We can produce discrimination without having any intent to discriminate or any dislike for those who end up being disadvantaged by our behavior.” Bias, in other words, can be more about who we choose to help than any intent to do harm. And according to the report, this “positive” bias likely causes the majority of discrimination in the U.S.

And when we make group decisions — as is done on a panel, selection or nomination committee — with a group of people just like us, that bias is reinforced and amplified.

This has big implications: In an industry like film, where personal relationships and powerful gatekeepers govern access to resources in such a stark way, it represents a significant risk of biased outcomes, no matter how positive our intentions may be as individuals.

Think about it. When was the last time a friend recommended or hired you for a job?  When was the last time you didn’t have a personal reference for someone working for you? This is an industry built on trust; an industry sustained by connections; an industry that runs on people helping other people. And according to the data: we have a problem with whom we choose to help.

This problem is compounded by structural issues in the way the film industry functions: because the in-group dynamic influences decision-making gate-keepers at numerous points along the film-production arc, in-group bias has a cascade effect, a narrowing effect the further along a film gets from idea to screening. Think of it like a funnel. An early grant-giver can be diverse in its selection criteria, but a film needs more than a single grant. Private investors, festival programmers, distributors all come into the mix, and at each stage in-group bias is able to narrow the range of supported material. The result? Many, many diverse voices never make it to the screen. Women and people of color initiate work that never has the chance to be judged on its merits, while other, safer bets make it past the finish line.

The system needs to be fairer. People need to examine and correct the areas where unfairness creeps in. What’s more, this is our collective problem. It’s not a problem that will be solved by some external force but one that each of us needs to confront directly.

If you believe in a meritocracy, a place where the person who gets hired is the best person for the job, where the films that get made are the most exciting and engaging films, and the ones that best represent the story of our moment, then you owe it to yourself to make sure your internal bias isn’t actually subverting that ideal.

The thing is, making films is hard. Really damn hard. Most of us trying to make films are barely holding it together. We are under extreme pressure and are resource-deprived, and we end up working a lot of the time in a very self-protective space — a space where it’s hard to feel generous, or even think beyond survival needs.

In that space, helping people (or even choosing who and how we might help) becomes pressurized. And it’s in that space that we’re most vulnerable to our own unconscious bias. Overcoming that on an individual basis is a problem each of us needs to work at.

Many people are already actively engaged in helping to solve this problem, but to build that momentum into a movement for systemic change, we need a way for everyone to get involved in changing our industry’s approach to diversity. We need you.

Here are a few ways you can work toward that change:

KNOW YOUR BIAS. Does your review or hiring process contain peer-level diverse input?  Do you make decisions all on your own? Is your staff diverse? Your board? Answering “no” to any of these questions should raise a flag that your decision-making process is vulnerable to unconscious bias.

TEST YOUR BIAS. Nobody wants to admit to bias, however unconscious. But remember: we’ve already established that an industry full of progressive, well-meaning people is still somehow producing non-diverse outcomes. Maybe it’s time to take an objective look at ourselves, to see if hidden bias is as in-built for us as it is for everybody else on the planet? (See the Harvard link at

Design your decision-making process to double-check for bias.

To combat in-group bias, you need to create systems that counter that bias during the process. First, create a diverse network of peer colleagues that can challenge your assumptions and provide you with referrals. But don’t simply rely on others. You wouldn’t buy a camera based solely on your friends’ advice; you’d do your research and get some experts to weigh in.

Hire, cast and refer diversely.

We know our industry isn’t diverse enough. Guess what? You get a chance to change that with every casting decision, hire or referral you make. An important first step is to build diversity in who you employ (and who’s on your board, if you are an organization). View this as a priority: design a system that allows those viewpoints in. Trusted colleagues are a good first step, but remember: real change only happens when you step out of your comfort zone. The goal is to move yourself (and your industry) to a place where choosing diversity no longer feels risky.

Refuse to be on any panel that isn’t diverse.

Given the in-group bias problem described above, this should be obvious: panels that lack diversity in their makeup also lack diversity in their choices and perpetuate false stereotypes. But often, when it comes to panel selection, we hear that old chicken-and-egg argument: we’d love a diverse panel, but really, the only qualified prospects are, umm… White Guys.

Enter Greg Martin, mathematician, and Anand Prasad, statistician, who helpfully used math to prove that non-diverse panels aren’t just random accidents. They focused on gender in their own academic fields, but the method works for any field and any demographic if you know the basic percentages. Basically, they found that the chances of non-diverse panels ‘just happening’ were almost nil. In other words, arriving at an all-white-male panel by accident is statistically almost impossible.

So, armed with that knowledge, stick to your guns: only panels with diverse members can reflect our diverse views and make diverse choices.

Funders, festivals, awards and educators need to especially question systems built around first-time filmmakers. 

It seems obvious but it’s worth re-stating: if you want to diversify an industry, you need to support diversity from the very beginning. If the very first gatekeepers introduce bias into the mix, then every subsequent step of the process will be colored by that bias, all the way down the line.

Any system that supports first-time filmmakers, from an individual to a foundation to a festival, needs to be clear that if their slate is not diverse, they are part of the problem.

This is particularly important because first films are inherently risky, and the decision-makers who back a first-time filmmaker are often putting their professional reputation on the line. The impulse to mitigate that risk by making safe bets is understandable. But in-group bias strongly influences our ideas of what a safe bet is; often, it looks like what we know, and what we’ve already supported. In particular, the empathy needed to connect with a first-time filmmaker’s vision despite almost inevitable flaws is easier when that vision looks like our own, and the filmmaker feels like “one of us.”

First films are the most significant threshold in our industry, and one of the most susceptible to in-group bias. Getting a first film made requires any number of group decisions — an obstacle course where failure at any stage has devastating consequences. If we are not clear and deliberate about policing ourselves for bias in each of those decisions, we are failing ourselves and our field.

Get more data.

Data helps us understand more clearly. And in terms of diversity in our industry, we’ve still got a lot of understanding to do. The studies we’ve touched on here are a start, but  we still have a lot more data to get.

We need to make sure our cinematic identity represents our cultural identity as a nation. We need to understand who is telling the stories of our time, and how the financial and support resources to tell those stories are assigned. And we need to ensure that these stories are coming from a dynamic range of perspectives.

And we need even more and better data on racial and gender diversity related to who gets to make movies, who decides who makes movies and what subjects and stories get told. We need better data on funding levels that can be parsed for race and gender disparities. We need better data on ourselves and who we hire and support. We need this to ensure that our nation’s cultural expression supports not just familiar voices, but every voice.

But most of all, we need to commit ourselves to building a system that can truly support the amazing array of film talent we have in a way that is equitable, inclusive and awesome.

© 2016 Filmmaker Magazine
All Rights Reserved
A Publication of IPF