The Year the Mirror Cracked

Every year a small team of researchers at UCLA does something the entertainment industry will not do for itself. They count. They count who got to be the hero, who got to direct, who got to write, and who got to appear on screen at all, and then they hold those counts up against the people sitting in the dark watching. The result is the closest thing American film has to an honest mirror.

In 2025, the mirror showed a culture pulling away from its own reflection.

This is not a story about quotas or controversy, the tired frame that gets bolted onto any conversation about who appears on screen. It is a story about a far stranger and more revealing phenomenon: an art form actively retreating from the audience that sustains it, in a year when that audience had never been clearer about what it wanted. To understand 2025 in American film, you have to hold two facts in your head at once, and the friction between them tells you almost everything about the cultural moment we are living through.

Fact one: audiences rewarded films that looked like the actual country. Fact two: the industry made fewer of them.

I have spent years writing about cultural fragmentation, the slow tearing of the connective tissue that once let a society see itself as a whole. My Enemies of You series is, at bottom, an extended argument about what happens when the institutions that shape a culture lose contact with the people inside it. The 2025 film data is one of the cleanest illustrations of that thesis I have encountered, a real-time snapshot of an institution drifting away from its audience and calling the drift caution. So let me treat these numbers the way they deserve to be treated, not as a diversity scorecard, but as cultural evidence.

How Reliable Is the UCLA Report?

First, the instrument itself, because its quality is the reason any of this can be said with confidence. UCLA’s Entertainment and Media Research Initiative has produced the Hollywood Diversity Report for thirteen years. The film findings come in two parts: a theatrical report, which this year examined 109 of the top English-language releases at the global box office, and a streaming report, which examined 89 streaming original films on the major platforms. Both cover the films of 2025. Both were authored by Dr. Ana-Christina Ramón, Dr. Michael Tran, Jade Abston, Nico Garcia, and Dr. Darnell Hunt.

The decisive design choice is that the researchers do not take the industry’s word for anything. They build their data independently, from production credits, subscription and box office databases, and audience measurement firms, instead of from numbers the studios hand them. This is why the report functions as a mirror and not a press release. It reflects what is there, including the parts the subject would rather not see.

What it reflected in 2025 was a reversal.

What Changed in 2025?

Consider the lead role, the character a story is built around, the face on the poster, the person the audience is asked to follow for two hours. In the year’s top theatrical films, the share of leads played by people of color fell to 23.1 percent, sliding down from a peak it had reached in 2023. Against a national population that is now 45.2 percent people of color, that means the heroes of America’s biggest films were drawn from the country’s white population at nearly double the rate their numbers would suggest. White actors, whose share of the population is shrinking, took 76.9 percent of theatrical leads, up from the year before.

Streaming, which had been the more adventurous medium, snapped back hard. People of color had peaked as streaming film leads in 2024. In 2025 their share dropped to 36 percent, while white actors reversed a two-year decline and climbed back to 64 percent of leads. Among Native performers the number was not merely low. It was zero. Not one Native lead in the entire streaming film slate for the year. An entire people, written out of the center of the story completely.

Women experienced the same undertow. In theatrical film they had come within striking distance of parity as leads in 2024, holding 47.6 percent of lead roles, the closest American film had come to simply reflecting the gender split of the country. By 2025 that had collapsed to 37 percent. Four years of apparent gains, gone in one. Behind the camera the erosion cut deeper. Women directed a mere 10.1 percent of the top theatrical films, the lowest figure since 2018, down from 15.4 percent a year earlier. One of the report’s authors, the doctoral candidate Jade Abston, put it with unusual bluntness for an academic: it was as if the progress women had experienced simply disappeared.

The word the researchers kept returning to was volatility. Not progress, not backlash exactly, but a violent year-to-year swinging that never settles into anything durable. The industry lurches toward reflecting its audience when the money is flowing and flinches back the moment the climate turns, leaving behind a jagged line that goes nowhere over time. Two-thirds of the top theatrical films in 2025 featured majority-male casts. White men alone accounted for half of all lead actors. The needle did not merely fail to advance. It was yanked backward.

The audience the industry stopped trusting

Here is where the cultural story turns genuinely strange, because the retreat was not a response to audience indifference. The audience was not asking for this. The audience, by every measure UCLA tracks, wanted the opposite.

The report’s most durable finding, repeated across more than a decade of data, is that films reflecting the real diversity of the country are not a commercial gamble. They are a commercial strength. In 2025 the films with casts between 41 and 50 percent people of color, the band that sits almost exactly on the nation’s actual demographic makeup, earned the highest median global box office, the highest median domestic box office, opened in the widest release, and posted the best average opening-weekend rank. The films that looked most like America made the most money in America and beyond it.

The genre data drove a stake through the old assumption. Horror, the single most common type of film made in 2025 and a genre beloved by Black and Latino audiences, delivered the highest median return on investment of anything in theaters. The genres that skewed whitest in their audience, biography, documentary, drama, brought up the rear with the lowest median receipts. The most profitable territory in film was the territory serving the most diverse crowd.

Streaming told the same truth from the other direction. People of color were overrepresented as viewers, beyond their share of the population, for nine of the top ten streaming films of the year. Women were the majority of the audience for six of the top ten. And the single biggest streaming event of 2025, an animated film that lodged itself in Netflix’s global top ten for fifty-two consecutive weeks and topped the household ratings across every demographic group the researchers measured, was directed by a woman of color and centered on young women of color. The audience did not merely tolerate the diverse hit. The audience made it the defining phenomenon of the year.

So we arrive at the contradiction in its full absurdity. The people watching were more diverse than the stories being greenlit. The audience had moved ahead of the institution that exists to serve it. And instead of follow the audience toward the choices the data said were both more representative and more profitable, the industry retreated to the familiar.

Dr. Ramón named the mechanism without flinching. “In times of contraction, diversity is often the first thing sacrificed,” she said. “Economic uncertainty becomes an excuse.” That is the cultural diagnosis in a single line. The retreat was not driven by what audiences wanted or by what the numbers rewarded. It was driven by fear, and fear reaches for the familiar. When an institution gets scared, it does not optimize. It regresses toward whatever felt safe in the past, even when the past is demonstrably less profitable than the future the audience is begging for.

Who gets the budget is who gets to matter

The deepest cut in the report is not about who appears on screen. It is about who gets the resources to make something that can command a screen in the first place. And this is where the cultural stakes become unmistakable, because a story made for 15 million dollars and a story made for 150 million dollars are not competing in the same arena. One whispers. The other gets to shout.

The streaming data exposed the hierarchy with brutal clarity. Every single streaming film directed by a white woman in 2025 came in under a 20 million dollar budget. All of them. Directors of color clustered under the same ceiling for the most part. White men were the ones entrusted with the spectacles, far likelier to helm the films budgeted above 50 million dollars. Of the eight streaming films made that year with budgets of 100 million dollars or more, exactly two were directed by people of color: Guillermo del Toro, for his Frankenstein, and Maggie Kang, for KPop Demon Hunters. Kang was the only woman of color anywhere near that budget tier.

Think about what that means culturally. It is one thing to decide who gets to direct. It is another, quieter, and more consequential thing to decide who gets to direct at scale, who gets the marketing budget and the production resources that determine whether a film can become an event or is doomed to remain a footnote. You can hand out directing jobs to a diverse roster and still keep total control over the culture by controlling which of those films gets the firepower to enter the cultural conversation. Inclusion at the level of hiring, exclusion at the level of resources. The credit without the clout.

This is the form of gatekeeping that does not register in a headline about representation, and it is the form that matters most. Because the films that get the budgets get the billboards, the press tours, the opening-weekend saturation, the cultural oxygen. The under-resourced film, however good, however true to an underserved audience, is sent into the world without the means to find that audience, and then its modest performance is quietly entered into evidence as proof that films like it do not work. The verdict is rigged at the budget meeting, long before the audience ever gets a vote.

The theatrical numbers complicate this, and honesty requires saying so. In theaters, leads of color in 2025 were the most likely to anchor the biggest-budget films, the 100 million dollar and up productions, and the least likely to be stuck with the smallest budgets, an inversion of the streaming pattern. When Hollywood committed to a diverse-led theatrical tentpole, it sometimes committed real money. The picture is not a single clean line of villainy. But the streaming data, where the future of the industry increasingly lives, shows the resource hierarchy reasserting itself, and that is the direction the medium is heading.

The fragmentation thesis, on screen

This is the point where the film data stops being a film story and becomes a cultural one, and where it connects to the larger argument I have been making across the Enemies of You books.

The thesis of that series, in brief, is that American culture is fragmenting because its central institutions have lost the ability, or the will, to see and serve the whole of the people they claim to represent. The institutions retreat into the familiar, the comfortable, the safe, and in doing so they abandon the actual, diverse, complicated audience to find its stories elsewhere. The result is a culture that no longer has a shared center, because the entities that used to build that center have stopped trying.

The 2025 film data is that thesis rendered in numbers. Look at what it describes. An audience that is ahead of its institutions, more diverse, more open, more eager for stories that reflect the real world than the people deciding which stories get made. An institution that, when frightened, retreats to a narrower vision of who matters and whose stories are worth resources. And a widening gap between the two that the institution papers over with the language of caution and risk.

This is fragmentation in motion. When the biggest theatrical hit of the year is a Chinese animated film that earns 99 percent of its 2.2 billion dollars outside North America, when a Japanese anime sequel cracks the global top ten, when audiences reach past the traditional Hollywood center for stories that center did not make, what you are watching is an audience outgrowing its institutions. The center is not holding because the people who run the center stopped serving the whole. The audience did what audiences do when they are not served. It went elsewhere. The Hollywood landscape, as the report dryly notes, is changing, and it is changing because the gap between the makers and the watched finally grew too wide to ignore.

A healthy culture has institutions that reflect it back to itself, that let a society of enormous variety nonetheless recognize a shared story. What the 2025 data documents is the opposite: institutions that reflect an ever-narrower slice while the society itself grows broader, until the reflection and the reality no longer match and the audience stops recognizing itself in the mirror at all. That is what fragmentation looks like from the inside. Not a dramatic rupture, but a slow, year-by-year drifting apart of the culture and the institutions that are supposed to hold it together.

The films that prove it could be otherwise

The report is not a counsel of despair, and neither is the cultural moment, because scattered through the data are the films that show what happens when an institution does trust its audience.

KPop Demon Hunters is the obvious one, and it deserves its moment. Directed by Maggie Kang, centered on young women of color, it became one of the most successful streaming films in Netflix history and topped the ratings across every single demographic group the researchers measured. Every group. Not a niche film that overperformed within its niche, but a film made from a specific, authentic, non-default perspective that turned out to have universal reach. The researchers noted, pointedly, that the studio left money on the table by not giving it a proper theatrical run first. The institution underestimated its own hit. The audience did not. Given a direct path to the film, audiences across every demographic line made it the phenomenon of the year.

The argument is in the title. The specific, authentically made story did not narrow the audience. It expanded it. The film succeeded not despite being made from a particular perspective but because of it, because audiences are hungry for vantage points that the default machinery rarely offers, and because authenticity at the origin point is what makes a story land. The most reliable way to reach everyone, it turns out, is not to make something for no one in particular. It is to make something true for someone in particular, with the craft to let everyone else in.

Ryan Coogler’s Sinners served the same lesson in theaters, becoming the most Oscar-nominated film in history and a model the report holds up for an industry trying to draw audiences back. The streaming data reinforced it structurally: films directed and written by people of color, and by women of color especially, were overwhelmingly the ones with the most diverse and gender-balanced casts. The authentic vantage point at the top of the production shaped a more reflective work all the way down. Who tells the story determines whose story gets told, and the stories that reflected the real audience were the ones that audience showed up for.

These films are not exceptions to a depressing rule. They are evidence of a possibility the institution keeps declining to pursue at scale. They show that the gap between the audience and the industry is a choice, not a fate. The audience is right there, ready, proven. The only variable is whether the people with the resources will trust them.

Which Absence Is Nobody Discussing?

Tucked into both reports is a category that gets less attention than race or gender and reveals something just as stark about whose stories the culture considers worth telling. The researchers have, for four years now, tracked the disability status of actors, and the 2025 numbers are a quiet indictment.

Adults with disabilities make up at least 26 percent of the United States population, more than a quarter of the country. In the top theatrical films of 2025, performers with a known disability held 9.3 percent of lead roles and 8.7 percent of total acting roles. In streaming, the leads number was higher at 14.6 percent but total actors sat at just 6.5 percent. To reach proportionate representation in theatrical leads, the share would need to roughly triple.

But the number that stopped me was a smaller one. Among the lead actors with a known disability in theatrical film, the report notes there were once again no lead actors with visible disabilities in the top releases of the year. The disabilities represented were overwhelmingly ones the camera does not see. The bodies that move through the world differently, that a quarter of American families live with daily, were essentially absent from the center of the country’s biggest films.

This is the part of the fragmentation story that gets the least airtime, precisely because the audience it concerns is so rarely imagined as an audience at all. An entire dimension of human experience, common, universal in the sense that disability touches nearly every family eventually, simply does not appear in the mirror. Not contested, not debated, just quietly left out, year after year, in a way so consistent that the researchers can note its recurrence with weary precision.

For a culture that claims to tell the human story, the persistent erasure of how a quarter of humans live is not a minor omission. It is a tell. It reveals how narrow the industry’s working definition of the human story has become, how much of the actual range of human existence it has decided is not bankable, not relatable, not worth the screen. The mirror does not merely under-reflect race and gender. It has a whole population it does not show at all.

Authenticity at the origin point

There is a deeper artistic principle threaded through all of this data, and it deserves its own paragraph, because it is the thing that separates the films that worked from the films that merely existed.

The report tracks writers separately, and the writing numbers are among the most lopsided in the entire study. People of color held 20 percent of theatrical writing credits and 21.3 percent in streaming. Women held 26.6 percent in theaters and 37.1 percent in streaming. The people who originate the story, who set its perspective before anyone is cast or any frame is shot, were among the least representative of the audience in the whole industry.

This matters more than the casting numbers, in a way, because writing is upstream of everything. A story’s perspective, its sense of whose experience is central and whose is incidental, is fixed on the page. You can cast a film diversely and still tell its story from a narrow vantage point, because the vantage point was decided in the script. And the data bears this out: films written by people of color, and by women of color especially, were dramatically more likely to feature diverse, gender-balanced casts. In the streaming data, nearly all of the films written by people of color had casts that were more than 30 percent people of color. Who writes the story shapes who lives inside it.

What this points to is a truth about storytelling that the industry keeps relearning and forgetting. The power of a story comes from the specificity of its origin, not from the breadth of its target. KPop Demon Hunters did not top every demographic group because it was engineered to please every demographic group. It topped them because it came from somewhere specific and true, and specificity, rendered with craft, is what audiences respond to. The committee-smoothed product designed to offend no one compels no one. The authentic vision aimed at someone in particular turns out to reach everyone.

This is the artistic heart of the fragmentation argument, and it is why the retreat documented in these reports is an aesthetic loss as much as a social one. When the origin point of the culture’s stories narrows, the stories themselves flatten. They lose the texture that comes from being made by people who have lived the experience on the screen. They become competent and forgettable, technically proficient reflections of an increasingly narrow vantage point, and audiences feel the hollowness even when they cannot name it. The diversity the report measures is not a moral garnish on the art. It is a precondition for the art being good, for the stories carrying the specific human truth that makes a story worth telling at all.

An industry that narrows who gets to originate its stories is not merely being unfair. It is making worse art, and the audience, which can always tell, drifts toward the places where the stories still feel true. That drift is the fragmentation. It is what happens when the culture’s official storytellers stop telling the whole truth and the audience goes looking for it elsewhere.

What Is the Strongest Argument Against This Reading?

A cultural critic who only indicts is not doing the job, so let me give the other reading its due.

The volatility the report documents cuts both ways. A year of retreat can be followed by a year of advance, and the same instability that produced 2025’s rollback has, in other years, produced surprising leaps forward. The pattern is not a steady slide into darkness. It is a churning, and churning can break in hopeful directions as easily as bleak ones. Reading 2025 as the beginning of an irreversible decline would be as much a distortion as the industry’s own happy talk in the good years.

The institutions also face genuine pressures that the audience does not see and the data cannot fully capture. The theatrical business is being reshaped by forces, the collapse of the mid-budget film, the dominance of franchises, the rise of non-English global cinema, the migration of attention to streaming, that have nothing to do with diversity and everything to do with a medium in structural upheaval. Some of the retreat the report measures is the wreckage of an industry in crisis flailing for whatever feels safe, not a deliberate turning away from any particular audience. Fear is a real explanation, and frightened institutions make bad decisions across every dimension at once.

And there is the simple fact that any single film can fail for reasons no dataset predicts, which means the executive choosing the familiar option is not purely a coward. He is a person whose career can be ended by one expensive miss, making a defensible bet that the known quantity is less likely to humiliate him than the unknown one. The aggregate data says that bet is wrong on average. But no executive lives in the aggregate. They live in the specific, terrifying particular of the one decision in front of them, and the retreat to safety is what frightened people in high-stakes positions have always done.

None of this excuses the outcome. But it complicates the easy story, and the easy story is always worth resisting. The industry is not a cartoon of villains turning away from a virtuous audience. It is a frightened, struggling institution making the oldest mistake there is, mistaking the familiar for the safe, in a year when the data was screaming that the two had come apart.

The mirror, and us

What stays with me, after sitting with both reports, is the durability of the audience against the volatility of the institution. Year after year, in good times and bad, through every swing of the industry’s nerve, audiences keep showing up for stories that reflect the real world. The preference does not flicker. It is the most stable finding in thirteen years of data. What swings wildly is not what people want. It is whether the people in charge are willing to give it to them.

That is, finally, a hopeful fact, even buried inside a report full of discouraging numbers. The culture is not the problem. The audience is not fragmenting. The audience is, if anything, more unified in its appetite for honest, varied, reflective storytelling than the institutions give it credit for. The fragmentation is happening at the level of the institutions, which keep losing contact with a public that is right there, ready, and clear about what it wants.

The mirror cracked in 2025. But a cracked mirror still shows you something true, if you are willing to look. What it showed was an industry flinching away from its own audience, and an audience that kept faith anyway, kept buying the tickets, kept driving the ratings, kept making the authentic film the year’s defining hit even when the institution did everything but bet on it.

The audience has been holding up its end of the bargain the whole time. The only question that matters, the question the 2026 report poses to everyone who makes culture for a living, is whether the people with the resources will finally decide to trust the people they have spent a year backing away from. The data says they should. The hits say they should. The audience has been saying it for thirteen years.

All that is left is for someone with the budget to believe it.

This essay analyzes findings from the UCLA Hollywood Diversity Report 2026, Parts 1 and 2, published by the Entertainment and Media Research Initiative at UCLA’s Institute for Research on Labor and Employment. All data and statistics are the work of its authors. The cultural interpretation is my own.

A companion piece examining the same research through the lens of publishing and the business of telling your own story appears on The Writing King. For a book-length treatment of the cultural fragmentation this report illustrates, see my Enemies of You series.

Suggested citation: Ramón, A.-C., Tran, M., Abston, J., Garcia, N., & Hunt, D. (2026). Hollywood Diversity Report 2026, Part 1: Theatrical Film and Part 2: Streaming Film. UCLA Entertainment and Media Research Initiative, Institute for Research on Labor and Employment. https://socialsciences.ucla.edu/hollywood-diversity-report-2026/

Frequently Asked Questions