Stages Within Stages: The Minority Report, Year Two


Ever since 2009, I’ve conducted an ongoing study of diversity in Star Wars fiction—first (and still) at the Jedi Council Forums, then here at Eleven-ThirtyEight. Over time, I developed a means of “diversity scoring” various stories based on the demographics of their casts, and began looking for trends and precedents in the franchise, for good or ill. One huge thing I’ve learned from this process is that it’s very, very hard to quantify diversity in a useful way; people inclined to argue with me will often yell “Quotas! One of everything!”, which is an easy logical leap to make but hardly a solution. Not all roles and stories are created equal, so simple math is at best a limited measure of work’s value.

This became especially clear to me a year ago, when scoring the first several works of the new Star Wars canon. While at first glance these stories had established a number of remarkable things like an all-female stormtrooper unit, a black main character for a middle-grade series, and several LGBTQ characters in a single book, these bold steps weren’t showing up in the scoring—if anything, the raw figures were slightly worse than they had been in my studies of Legends material. While the average score has ticked down a little over the last year—from 67 to 60—that feeling has mostly held up.


If you read my last Minority Report, let me go over some of the differences in this chart:

  • First of all, I’ve switched from listing items in release order to chronological order. While I initially hoped to see steady improvement as the real-life years went by and the franchise got more progressive, the fact is that a great deal of canon content thus far has focused on the Imperial era, and the rise of what the Expanded Universe called “High Human Culture”. While this has been tamped down a lot in canon, it’s hard to deny the near-universal lack of (and prejudice toward) aliens among original trilogy Imperials, and it’s understandable for material covering the Empire at its height to reflect that—thus, a much higher batch of scores in the prequel era follows by a lot of dips between Episodes III and VI, followed by some recovery in the sequel era. Despite the prequels’ depiction of a galaxy falling apart, I think it’s fair to see those first few scores in the high eighties (which translates to roughly eight in nine characters being something other than a straight, white human man) as a reasonable “ideal” for the post-OT period to eventually return to.
  • Now that the number of new canon books has exceeded the number of films, I’ve chosen to demarcate the visual media (films and TV, in dark blue) from the publishing (light blue), and give average scores for each. Given the differences in their respective audiences, I plan to weigh each of the two averages equally from now on no matter how many more books are eventually published—meaning that the average score for the overall Star Wars canon is currently (70+60)/2 = 65. For the record, that’s just a teensy bit lower than the EU average of 67.
  • You may also notice that Servants of the Empire is no longer listed—in addition to the reservations I discussed last year about holding that series to this standard, the simple truth is that I don’t usually read the middle-grade books, and rather than open up a whole new category I decided to limit myself (at least for now) to the “adult” and “young adult” novels.
  • Another quirk comes with scoring the two TV series—it could be argued that the purest way to analyze them would be doing a separate score for each episode, the same way each film is its own Episode, but in terms of workload and presentation that’s just beyond the limits of this operation right now, so instead I’ve treated each entire series as one giant cast of characters (Rebels‘ score only accounts for the first two seasons, naturally). It should be understood that this probably has some positive effect on their scores, as it counts Obi-Wan and Anakin, for example, only once each despite their huge prominence in The Clone Wars overall. This effect is certainly less prominent with Rebels, as its entire main cast, and even many of its more prominent Imperials, are either people of color or aliens. Someday I may go to the trouble of a true “average episode” score for the two series, but today is not that day.
  • Many of the scores above have changed slightly; this is due to my leaning much harder on Wookieepedia now than I did last year. Their early coverage of canon novels wasn’t quite thorough enough for my purposes, but that—combined with the fact that most of the material has been out for several months—is no longer the case. I still referred to original text in many cases, but when the wook lists a minor character as having pale skin (or more likely, doesn’t mention their skin tone at all) I’ve chosen to trust the community’s fact-checking over my own.
  • Likewise, the films are now scored not by the actual end credits cast lists but by the slightly more thorough “Cast” listings on their wook pages. While the “named character” standard works differently in a visual media where there’s less need to say, for example, Red Leader’s actual name, I chose to use these lists rather than their full “Appearance” lists, which are both comprehensive to the point of insanity and kind of a mess post-reboot (fun example: “Brea Tonnika” is canon due to a cameo in TCW, but “Senni Tonnika” is still technically Legends despite the fact that she’s standing right there).

With all those caveats and variations, I prefer to use these scores not as critiques of any individual work but as indicators of broad trends against which expectations can be measured—Tarkin and The Empire Strikes Back, for example, have especially low scores even by the standards of their era and the stories they’re telling. Likewise, both Twilight Company and Heir to the Jedi take place around the same time and focus almost exclusively on Rebel and Rebel-adjacent characters, but TC’s giant cast and tendency to toss off tertiary character names without any context kept it from scoring anywhere near as well. Could many of those tertiary characters have been women and POCs? Absolutely, but history suggests that an ambiguous character might as well be another white guy.

aftermath-lifedebtWith all that said, though, where are we now? Things are definitely improving. While not spelled out in the above, I did count LGBTQ characters toward a story’s score, and I’m proud to report that we’ve gone from about eight canon characters a year ago to over a dozen now, including the first nonbinary human character in the form of Life Debt‘s Eleodie Maracavanya (who, incidentally, does not yet have a Wookieepedia page—someone get on that).

Like last year, this year’s improvement in the diversity realm is less of quantity than of quality. The Force Awakens‘s diversity score was 62—better than A New Hope and ESB, but nowhere near the prequels and not even quite as high as Return of the Jedi or the EU. But much like Servants of the Empire before it, its two main characters were a woman and a black man, which in a Star Wars film, the first in a decade, is as game-changing as it gets. Rey and Finn were joined by a Latino fighter pilot, a female Resistance general, and a female stormtrooper commander in secondary roles. Many of the film’s WHMs (white human men) were tertiary First Order officers with one line, much like the Imperials of the OT, except this time even the FO had a handful of women. TFA, while short of ideal, is the perfect distillation of the new canon’s approach to diversity—the demographics are roughly the same, but the balance of power has shifted drastically.

The stages of inclusion

Last week, the Guardian published an essay by Rogue One actor Riz Ahmed, excerpted from The Good Immigrant, discussing his experiences as a young actor with a look that mostly gets one cast as “terrorist”. Early in the piece he breaks down his prospective roles into a series of three ascending “stages”—roughly, Stereotype, Stereotype Subverter, and Regular Guy (aka “the Promised Land”). I’ve noticed a similar sequence myself both in and out of Star Wars, and this struck me as a good opportunity to delineate the bad roles from the good ones from the great ones—now that we’re seeing enough of the latter to warrant the distinction.

To be abundantly clear, I am a lanky, blue-eyed white guy—and straight, besides. I will never know the personal offense that comes with stereotyping because the stereotype our culture has assigned to me is “normal”. What I am, though, is a keen observer; and I’ve been observing this for a long time. This is what I’ve seen.

Type I – zero inclusion. This is the realm of blackface; not only are the roles caricatures, but a real actor of color (for example) doesn’t even get the job. Also included here are stories with no women or people of color at all—though debatably, even that’s still preferable to the former. It’s important to note that this is the only type that I see as inherently irredeemable; standards evolve, and even a token is better than a guy in blackface.

Type II – tokens and stereotypes. This, alas, is where Star Wars enters the playing field. While Leia was an excellent and boundary-busting female character for her time, she was still basically the only one. Same goes with Lando—giving Han a black friend was a big deal in 1980, but he was the Black Character just as much as Leia was the Woman; see ESB’s diversity score of 35, the lowest of the films by far. Meanwhile, in more terrestrial stories, this is the domain of infamous stock characters like the Sassy Black Friend of the Magical Negro—or worse, stories in which the only black character is a drug dealer or a murderer, and is pursued by the virtuous white protagonists.

Type III – this is where I would put Ahmed’s stage two—stereotype busters. Roles like Danny Glover’s veteran cop in Lethal Weapon or, for a different kind of example, Sigourney Weaver as Ellen Ripley. Ripley, like Leia a couple years before her, was an early example of a female character designed specifically to counteract expectations regarding women in sci-fi movies, but unlike Star Wars, Alien at least managed to get another woman onto the crew of the Nostromo. It also had a token black guy, demonstrating that these types are fully intersectional—you can smash boundaries in one area and still fail in another.

Type IV – normal is the point. Star Wars as a whole is representative of this category even while many of its individual characters might fall into the earlier types; a better example might be the original Star Trek, which portrayed a diverse cast of characters encompassing multiple nationalities and races (and multiple women, a decade before Leia), with the very specific intention of saying “Look how totally normal this is! It’s nothing to worry about!” More recently, this is series like The Cosby Show and Will and Grace, which make a big deal out of their diversity by not making a big deal out of it—by demonstrating to the masses that black and gay people, respectively, can have the same silly sitcom troubles any white family might have. Notably, this is where you can finally start to see non-WHMs in real numbers; if anything, they’re focused on disproportionately to underscore the story’s messaging.

bodhiType V – Riz Ahmed’s Promised Land. These are stories where the people of color are just people, the women are just people, and the LGBTQ people are just people. They have the same joys and troubles, skills and limitations, any other character might have, and they’re present in enough numbers that one can be a shitheel (or just plain ill-conceived) without casting aspersions on their entire demographic. This is certainly Star Wars’ ideal world; stories in the Galaxy Far, Far Away aren’t bound by any real-world history so it’s harder to teach a lesson about tolerance in quite the same way that Trek did, but that allows them to skip Type IV and go straight to someone like Bodhi Rook, a renegade Imperial pilot who just happens to not be white. Or someone like Rey—who is exactly who she is and no one in her universe would expect her to be anything less solely because she’s female. Finn, likewise, is judged by Captain Phasma solely on his loyalty and combat effectiveness and his skin doesn’t factor into it at all—because why would it, in this universe?

In a way, Star Wars has it easy here; there’s no real-world baggage whatsoever to prevent Rey from being a Jedi or Finn from being a stormtrooper or Eleodie from being a badass Super Star Destroyer pirate captain; all the writers have to do is bother to do it. And now that they are doing it, they just have to do it more.


And that brings me to my final topic—as I mentioned last year, I’ve begun focusing not just on WHMs but on gender specifically. In addition to scoring all the latest canon material and re-scoring the earlier ones, I came up with a second score on a much simpler metric: how many characters aren’t men?

Unlike diversity scores, which can never really have a perfect number (not that the occasional story with no WHMs would be the end of the world), scoring gender parity is pretty goddamned simple: are half of the characters female? No? Then you probably messed up. Are there any nonbinary characters whatsoever? No? Even worse. Not only that, but one of the biggest flaws of diversity scoring—the highly variable percentage of ambiguous characters whom I treat as WHMs but could technically be something else—is significantly reduced with a simple gender metric; a two-sentence throwaway character is much, much more likely to have their gender referenced than their skin color or even species. It doesn’t always happen, but it’s pretty damn close, enough that I’m much more willing to point to a story with a low score and cry foul. So with that in mind, which stories shall I point to?


Alas, pretty much all of them—the average Star Wars story is 78% male. The highest proportion of women in a book is Heir to the Jedi‘s 28%, and the highest proportion in a film or TV series is Attack of the Clones‘ 37%. In the entire canon, only AotC and The Phantom Menace manage to hit or exceed one-third, and most are less than one-quarter. Tarkin and The Empire Strikes Back, already the lowest two diversity scores, consist of an absolutely dismal 4% and 7% female characters respectively. ESB, sadly, is no surprise by now (and to be honest, it could be argued that some of the no-name women we see walking around Cloud City deserve to be counted as much as the Tonnika sisters do), but Tarkin really has no excuse. So disappointing is that score that just now while writing this paragraph I had to go back and recount it just to be absolutely sure, and yep—in a cast of forty-one, there are three named female characters: Anora Fair, Hask Taff, and Q’anah.

But it would be easier to forgive the occasional weak spot if there were a stronger track record overall, and I’m sad to report that there isn’t. While it’s technically possible for a book or film to get a perfect 50 (hell, it’s technically possible to go higher, but I won’t hold my breath), note that the graph above doesn’t even go above 40; it had no reason to. I’m coming to see this as a very simple pass/fail scenario—if your story has more than a handful of characters, and it’s not set in a men’s prison or prep school, either half of those characters should be female or you’re doing it wrong. This is why people are forever arguing about Padmé, or Black Widow in the Marvel films—when there’s only one woman of real significance, she has no choice but to stand in for all women, and no one character can be everything to everyone. Imagine how young boys would have felt coming out of the prequels if the only character they had to identify with was Anakin!


This is where I can’t help but be worried about Rogue One. TFA, I’ll admit, gave no hints that it would be the first film to feature female stormtroopers and “Imperial” officers, so seeing them in the final product was a great surprise (though note that it only scored a 28—beating the film/TV average but only tying with the books), but seeing shots like the above in the trailer, and another featuring several male Imperial officers on the Death Star, makes me very, very wary. We know Jyn won’t be the only woman in the film; there’s Mon Mothma, of course, and one other female Rebel is seen at Yavin Base in the original teaser. But as much as this cast really seems to be taking huge strides to open up the racial demographics of the Rebellion, in terms of gender, the film looks lodged very solidly into Type III. As it happens, Felicity Jones has even discussed Ellen Ripley as a role model for her portrayal of Jyn; that’s awesome as far as Jyn herself goes, and I have little doubt I’ll love the character as much as people loved Ripley. But 1979 was a long time ago. C’mon, Star Wars—that’s not enough.

Update: a small math error in my spreadsheet caused me to overstate the two average book scores somewhat. That’s been corrected in both the graphs and the article.


  1. Eric Harzer says:

    I guess we differ in the way we categorize characters, because I would not call Lando a token. I’d slot him in III or IV, if only because of his character’s backstory, current situation and arc within the story. I’m just thinking about Episode V here though. Lando goes above being the black friend to Han; he’s the baron-administrator of an entire city and he makes decisions for his city to keep them safe. Aspersions could be cast on the black character for betraying the heroes, but there’s a very real reason he made that decision. And to top it all off, his altruism wins the day and he helps the heroes and becomes a hero. Lando is just as much of a person or character as any one else, he has hard decisions to make and has actual development. I’d say his arc and role in the story qualify him as much more than a token.

    I guess I’m saying I’d argue that more important than representation en masse is the breadth of characterization present in the story. If Donnie Yen is just a blind wuxia master in Rogue One, is that any different from a stereotype of the Asian kung fu master? If he’s not majorly characterized otherwise (and it looks like he might be, his relationship with Baze, trust in the Force etc.) he would be more of a negative stereotype to me than good representation.

    This isn’t a science and it’s very subjective so it’s hard to quantify easily. I’d rather have a well fleshed out character on screen than a sea of nameless POC who have no impact on the story or exhibit a character.

    • Mike Cooper Mike Cooper says:

      I’ll concede that he straddles a couple types depending on how you read him, but I’d say II and IV more than III given that the impetus for his inclusion, as I understand it, was to demonstrate the normality of his and Han’s friendship more than the normality of his being a “responsible leader”. That you can cast aspersions on him re: ESB, though, is what makes him a token—it doesn’t matter whether there are valid story reasons for his betrayal because there’s no other example of a black character making different decisions; he has to stand in for his whole race. That being the case, to call him a token isn’t a criticism of the character at all—whom I agree is well written—but a criticism of ESB.

      Regarding RO, we’ll have to wait and see, but I would say that there’s less pressure on Chirrut to not be stereotypical because he’s not the only Asian character in the thing—a viewer is less likely to come away thinking “all Asians are nimble martial artists” when the Asian guy standing next to Chirrut is running around with a cannon on his shoulder. Whether two POC with middling characterization are truly better than one with solid characterization is definitely subjective though, so the details of the final film will be really important.

      • Eric Harzer says:

        If Lando were a streetwise, jive-talking criminal I could easily see him being a token black character. Making a traitorous turn would certainly be a tokenistic aspersion on his representation of his race, but the fact that he turns back at the end throws the token label out the window for me.

        Also, the ‘responsible leader’ angle is more than a throwaway detail. Han and Leia are only there because Lando, more than being a friend, is able to provide a safe haven from the Empire. It also reflects Lando’s character here because while Han was staying with his smuggler ways (even though he’s working with the Rebels he is being drawn back into that world to settle old debts), Lando is seen as having gone into being a legitimate businessman. This puts Lando into an impossible decision; prioritize an old friend, or the people he’s in charge of’s safety. In the eyes of our heroes he’s betraying them, but Lando is trying to do the best for his people. When he realizes the deal he’s made is a bad one, not only for his friends but for his people, he promptly changes sides.

        It might be easy to read it as just “Lando’s a traitor” but there’s enough in the text of the movie to see that he’s a lot more than simply a token black friend.

      • Mike Cooper Mike Cooper says:

        To be clear, I’m a big fan of Lando. While I chose to categorize them together because they’re both “first steps” toward ideal inclusivity, I don’t think being a token is the same as being a stereotype—and I certainly don’t think it means the character can’t be good. Leia was the token female for most of the OT. Leia’s also awesome. Both can be true.

  2. Palizinha says:

    The Rebels female problem annoys me so much. The show is great and I really love it, but there are few female characters and two of them are dead now while Ahsoka is Ahsoka, and of the new characters seen in the trailer only one is female. It’s probably Rebels’ main problem, Sabine’s apparent rising role or not.

    Give me at least Stormtroopers voiced by women, what even is the issue of doing that.

    • Mike Cooper Mike Cooper says:

      Yeah, the longer that show goes on the stranger it is—though it shows how much these things really are up to the individual creators regardless of what the official line is.

  3. Palizinha says:

    On an unrelated note, as far as your scoring goes, if say, a book comes out and it makes Obi-Wan canonly bisexual. Would that only affect works with him since that book, or older works like the OT would also have Obi-Wan’s position in the scoring change? Because I imagine the people who keep up with books would be aware of it and it would probably become common knowledge due to social media, but it still wouldn’t be clear from the movies/TCW, and wouldn’t be the original intent in these works.

    And yeah, the thing about Chirrut is that the character by himself may read like a stereotype, but there are other two Asian main characters in that movie, both of which seem very different from him. So it’s less Every Asian Does Martial Arts than it would have been otherwise.

    • Mike Cooper Mike Cooper says:

      That’s an interesting question—one that I couldn’t help but think about over the last year when people brought up the possibility of Diego Luna playing Biggs. I guess I’d have to count it retroactively because the scores are a way of saying “here’s where we are right now” more than of judging a story’s initial merits (or intent).

      Here’s a better example than Obi-Wan: Poe Dameron. I honestly doubt he was conceived as gay when TFA was made, but if they went ahead and validated the fan speculation in Episode VIII I would definitely see TFA from then on as including a gay character even though it wasn’t officially known at the time it came out.

  4. SF Eggbert says:

    Eleodie wasn’t a human

%d bloggers like this: