[Enjoy this stat-tastic guest post by Angela Jorgensen! – Ed.]
Turn on the television to any given prime-time TV show. Go on, go ahead, this article will still be here when you get back. Got it? Good.
Now, take a look at the couples you see depicted on the show you’re watching. What do they look like? Are they gay or straight? What is their race and ethnicity? Statistically speaking, odds are that the couple you see on prime-time network television is probably a heterosexual white couple. The majority of regularly occurring characters on television shows are white, and an even higher percentage of characters are straight. Of course, plenty of other sources have already commented on the lack (or perceived lack) of diversity in most entertainment mediums, so let’s move beyond that.
Just as a game, let’s think about your favorite same-sex couple from a current television show. Who do you think of? Santana/Brittany (Glee)? Jesus/Lafayette (True Blood)? Gil/Oscar (The Office)?
Wait a minute. Didn’t we just say that the majority of characters are white? Then why are so many same-sex couples interracial?
Well, before we jump to conclusions, it’s important that we’re not confusing anecdotal evidence for an actual trend. A handful of characters don’t necessarily represent a population sampling, so how to determine if there’s an actual trend in prime-time to depict same-sex characters as more likely to participate in an interracial relationship than to do so with heterosexual characters? How else than with a database?
I present to you the Gender and Race Character Breakdown Spreadsheet: an analysis of every relationship from 27 different influential television shows that aired during the past three years.
Now, before we get too deep into analysis, let’s start with a few definitions. First of all, what counts as a couple? “Couples” were defined as any two characters who kiss, go on a date, have sex, are married, or otherwise have a confirmed, cannon, on-screen romantic relationship.
I did not include implied attraction or relationships that were only hinted at. I also disregarded relationships that happened only in webisodes, books, or other expanded universe materials that were not part of the show itself. Only relationships that were consensual and real (i.e., no dream sequences, or cases where the characters were just pretending for some reason) counted. In addition, only couples who dated during the course of the show’s run counted, eliminating characters who dated as part of a back story (unless flashbacks to back story comprise a major part of the show’s mythology, such as in Lost.) For the ease of analysis, I only looked at relationships in which both characters occurred in more than one episode.
For purposes of this analysis, I counted Latino as a race rather than ethnicity. The racial breakdown included Asian (referring to East Asia), Black, Caucasian, Indian (including all ethnicities on the Indian sub-continent), Latino, Middle-Eastern, Multiracial, and Native American. If a character’s race was not specified in-show, I went by the actor/actress’s race.
So, now that we’ve got all of that out of the way, let’s dive into the really interesting part: the number crunching!
First of all, there are significantly fewer depictions of gay couples on television at all compared to the number of straight pairings. Out of the 405 relationships represented in this data set, only 12 pairings were among same-sex couples. Granted, with such a small number of gay pairings, it’s very easy for numbers to get skewed, but for the purposes of this post, the data set will be interpreted as representative of television homosexual depictions overall.
Same sex couples represent just under 3% of all relationships on television, which is consistent with GLADD’s 2011-2012 “Where We Are on TV” report, which reported that only 2.9% of all series regular characters on network television are gay. Many of these gay characters and same-sex relationships were concentrated in a small handful of shows. They were particularly represented in this data-set by the shows Glee and True Blood.
Of the 12 same-sex relationships depicted in this data set, 2/3 of those relationships were male/male pairings. In total, 50% of gay relationships in the data set (6 out of 12 relationships) were interracial. Among male/male gay couples, 37.5% of relationships were interracial, while interracial couples represented a whopping 75% of all lesbian pairings. In comparison, 14.5% of heterosexual couples were interracial.
In totality, 15.6% of all relationships in the data set were interracial. This is comparable to the rates in real life. In a recent study by the Pew Research Center, 14.6% of all new marriages in the US are interracial. Unfortunately, there is little data on interracial dating, and even less data on rates of interracial dating amongst same sex couples. As gay marriage is still illegal throughout much of the United States, same-sex couples from many states were not included in the Pew marriage figures.
So, the numbers strongly support the theory that same sex couples are far more likely to be depicted as being in an interracial relationship than straight couples. Why is this?
No matter how progressive society today may be, in many ways, people still make assumptions and use the same cultural coding as a shorthand for characterization. Gender and race both play a heavy role in our traditional readings of aggressive/submissive, active/passive, strong/weak, and other power dynamics that play out in romantic and other relationships.
When depicting traditional heterosexual couples, these power dynamics can be straight forward. Gender roles can be enforced or subverted; and the audience is used to seeing heterosexual couples that function with a variety of power dynamics so that these factors rarely get in the way of a character’s individual characterization.
Same sex couples, however, can’t typically be depicted using traditional sex roles. One member of the relationship must be portrayed as having many features of the opposite sex in order to fulfill traditional gender dynamics (i.e., a masculine man with an effeminate man; a feminine woman with a butch woman), or else both participants will be depicted as fulfilling more traditional gender roles, and power dynamics are drawn along non-gendered lines.
Like gender dynamics, race has culturally coded power dynamics of its own. While many of these power dynamics are based upon stereotype, they can still serve as short-hand for power dynamics that could otherwise be represented via gender politics. An audience member, no matter how PC, will hold certain assumptions about a black/white couple that are different from the assumptions about a Latino/white couple, or the far more commonly depicted white/white couple.
So, even though, say, Brittany and Santana from Glee both embody many traditionally female traits – at least to the degree that neither one could truly be described as “butch” or “masculine” in their relationship dynamic, Santana does fulfill an “aggressive, fiery Latina” archetype, allowing her to play a traditionally masculine role (assertive, aggressive, dynamic) without sacrificing her traditionally feminine traits (graceful, vulnerable, passionate.)
Of course, the power dynamic between Brittany and Santana is rather straight forward, so let’s take a look at a bit more complex relationship: that of True Blood‘s Jesus and Lafayette.
This relationship doesn’t have a clear assertive/passive breakdown, but instead, the couples have a give-and-take, depending on context. Latino Jesus embodies masculine traits as a mentor and leader, but also serves as a more feminine care-taker, and is the more sensitive member of the relationship. These traits are consistent with common archetypes of Latinos as depicted in American television: his leadership role is part of his status as a mystical (and exotic) wise person. His feminine traits are also common in depictions of Latinos, bringing a level of emotional passion and submission to this character.
Lafayette’s masculine traits are tied strongly to his stoicism and his aggressive risk-taking, while his feminine side is expressed through his submission to Jesus’s guidance, and his vulnerability as expressed when just the two of them are together. These traits: stoicism, aggression, and submission to a stronger male are all elements frequently associated with Black characters in American television.
And so, while these characters’ gender roles can be fluid, their depictions are not utterly alien to American audiences, who have grown used to these power dynamics not through gender roles, but racial roles.
In short, race and gender are similar in that they can both serve as culturally constructed short-hand for many of the assumptions that an audience holds about power dynamics and gendered roles within society. This can also account for why the instances of lesbian interracial relationship are so much more frequent than those of male/male gay relationships (again, 75% compared to 50%). Despite over a century of women’s rights activism in the United States, depictions of women in popular culture still are often more limited to archetypes than those of male depictions, leaving female/female relationships more prone to racial coding so as not to venture into risky depictions of women stepping outside traditional female roles. Although gay relationships are still taboo, showing men in non-traditional roles is still a less risky move than doing the same for women.
As with any medium, television is heavy on cultural coding and using gender and race as short-hand characterization. Same-sex couples are depicted as participants in interracial relationships far more often than straight couples are, in part because racial stereotype and expectations serve as coding for power dynamics that cannot be depicted through traditional gender roles, as is the case with heterosexual couples.
Angela Jorgensen originally hails from Iowa. She currently resides in Los Angeles and aspires to write for hour-long television dramas. She works with the Scriptwriters Network and is always up for meeting her fellow West Coast Overthinkers.
Perhaps there’s also an element of wanting to protect the apparently sacred white/straight/patriarchal heteronormative culture TV likes to present. They get to tick off both the ethnic and gay boxes in just one couple, leaving more room for white, straight roles.
That’s exactly what I was thinking. As long as we’re going to define these people as “other”, let’s get the other categories we need to cover out of the way. It would interesting to see how the numbers for racial underrepresentation compare.
I am sure I have read this from a producer or something somewhere. Damn fallible memory.
Kinda related: the anecodote in this post about film schools telling students the audience only wants to see white male leads: http://thehathorlegacy.com/why-film-schools-teach-screenwriters-not-to-pass-the-bechdel-test/
I’ll freely admit I have a bee in my bonnet about how non-white-straight-male(-youthful-attractive) roles are relegated to being supporting characters. So as much as I’d like to believe this was an isolated incident of one bad teacher, I have to conclude from the tv and movies I see getting released that it’s standard policy.
I don’t know why it took me so long to realize this but you’re comment shook me out of my haze. The real limiting factor here is that these are all AMERICAN shows. International readers, does this trend hold true in your tv shows and movies?
Very interesting. What happens if you do statistical tests on these numbers? I guess the lack of N means you might struggle to get results.
And the volatility of television programming and renewal rather negates the option of a time-series analysis, too, alas…
This comes with some large caveats:
In simple logistic regression, the odds ratio is 5.88. This can be interpreted as the odds of being same-sex interracial couple is 5.88 times the odds of being a male-female interracial couple. Confidence intervals are large (1.83 – 18.86) but significant. Adjusting for within-show correlation (for example, each Glee couple has a different probability of being interracial (or same-sex) than an NCIS couple, so these really aren’t independent observations) the odds ratio is 1.77 (95% Confidence Interval of 1.05 – 2.49).
It appears that statistically speaking that same sex couples are more likely to be interracial couples than male-female couples and that there is a meaningful amount of heterogeneity between shows.
Causality: Which way does causality go? Does same-sex dating cause increased chance of interracial dating or the other way around?
Selection of TV shows: As other readers mentioned, rather than an ad-hoc selection there needs to be an objective selection of shows. Perhaps include all the shows that pass some threshold of Nielson ratings?
Couple definition: in this analysis all couples are created equal. So, all of Jack’s flings on 30 Rock are each equivalent to Lily/Marshall on How I Met Your Mother. There is also likely an interesting case to be made about SES and age, as Gab mentions, but in this case all couples are considered equivalent.
Interracial definition: in this analysis all interracial combinations are created equal. A white-black combination is the same as a black-latino combination which is unlikely true in real life but this dataset is too small to really get at these differences.
I was interested in what you did with the April/Ben/Derek situation in Parks and Recreation, and leaving aside the fact you’ve got them all as separate couples (which I can understand for the purposes of this post), you’ve got Ben and Derek as MF in your database, and April as Caucasian? I may be mistaken, but doesn’t canonical Puerto Rican heritage classify her as multiracial?
Similarly, Blaine is not explicitly specified to be Caucasian in Glee canon, so by the logic of “going by the actor’s race” shouldn’t he also be categorized as multiracial? Darren Criss is very proud of his Hapa identity.
Further, Rachel’s dad Leroy on Glee is also Black.
But isn’t she adopted? Obviously she isn’t the genetic child of 2 dads.
She was produced by artificial insemination “mixing” her two fathers’ sperm together. However that’s not my point. My point was that Leroy is coded as white in the spreadsheet, just like Blaine and April from Parks & Rec. Leroy is canonically black (or at least multiracial?), April is canonically multiracial, and Blaine is NOT canonically white and played by a multiracial actor.
Thanks for pointing these out. I guess there might be some corrections to be made to the spreadsheet.
As I haven’t seen every episode of every show listed, I mostly went by what information on characters and actors I could find on Wikipedia or IMDB. Some actors didn’t have their races explicitly mentioned, and sometimes Google-ing didn’t turn up ready-to-find results, I did guess at some races. I guess I’m prone to my own tendencies toward ethnocentrism as well.
I enjoyed this quite a bit, and I’m curious as to if there are other factors that could be incorporated into the dataset to possibly tease out some other trends. For example, age and SES: Are these couples homogamous, too? Or do certain classes or ages tend to lend themselves toward a heterogamous relationship? For example, having a white person be the wealthier member in an interratial homosexual couple would also hold onto cultural expectations.
Like Tim said, due to the nature of the inquiry, since there are so few homosexual couples on TV, the amount of data is limited. But I like how it brings other questions to light, and makes at least me wonder just how far television has really come, and where society is in relation to that. Does society progress first, or does popular culture? In this case, I think that’s particularly difficult to discern, given how, as you pointed out, a lot of homosexual couples aren’t able to marry. But then that begs the question of whether marriage itself is necessary for people to be happy, have stable relationships, consider themselves a couple and all that kinda jazz, and even if monogamy is, too because what if… Rabbit hole- I love it! :)
Anyhoo, nicely done.
If you drop all TV shows that don’t show same-sex couples at all, how do the numbers look? My suspicion is that most shows don’t deal with same-sex or multi-race relationships at all, so shows that are willing to show same-sex relationships may also be more willing to show multi-racial couples, thus skewing the numbers.
This is a really good point.
Is there any way we can expand the data set. You chose 27 “influential” shows from the past 3 years…though I think a lot of them are pretty niche and Better Off Ted is a stretch. Are there any other shows that can be included under the guidelines?
It’s a really interesting analysis, but I agree with cat that I’d like to see the data set expanded. I’m not sure that I’d categorize all of these shows as “influential,” but I don’t know that the shows on the list really NEED to be “influential.” Perhaps they just need to rate on a certain scale of popularity, or have a run of at least X number of episodes, or something…
Also, in going through your data sheet, I noticed that Cam and Mitchell of Modern Family are incorrectly coded as MF (they should be MM).
I’m trying to think of other shows but I’m really tired right now so the only thing coming to mind is Smash. Christian Borle’s character had two boyfriends for at least a few episodes.
Game of Thrones has at least one among major characters.
I said this above (with a typo) but I think we should expand the list to non-American shows and see what happens.
Cam and Mitchell from Modern Family are marked MF on the spreadsheet. I only noticed because they were the same-sex couple that I first thought of.
You left Cam/Mitchell from Modern Family as a M/F relationship…
Nice work on that dataset! thats a lot of TV watching to do.I also like your conclusion, even if your dataset is still to small for confident correlation.
I think Penn has the key criticism here, that the shows with same sex couples are ALSO the shows with mix-race couples, so the comparison within these lefty-liberal shows is more useful for analyzing how race stands in for gender power dynamics.
How do you think it compares to the real-world population? Are there more interacial relationships among homosexual couples?
answered my own question
Boardwalk Empire did have a lesbian couple that featured prominently in the first season. Given how few data points there are I thought I would mention it.
This is an interesting post. I know this isn’t a really big deal, but since this is OTI, I might as well overthink your post.
I’m not a stats guy (I defer to JVO), but I don’t think your statement “the numbers strongly support the theory that same sex couples are far more likely to be depicted as being in an interracial relationship than straight couples” is supported by the data.
The “n” for same-sex couples is 13 (when switching Modern Family’s Mitchell and Cameron, who are the same race). That’s not much at all, especially compared to the sample size of 405 couples. Then, only six of those couples are interracial–you even use half of them in your examples in the third paragraph.
Your first paragraph in the “Data” section of your post acknowledges this limitation–but this limitation really should prevent any conclusions drawn from the data.
While we may be able to see some interesting anecdotal patterns in the data, and which are worthy of expanding to a larger sample of other shows, I don’t see 6 same-sex interracial couples as “strongly support[ing]” the thesis that same-sex couples are “far more likely” to be interracial.
But it’s an interesting idea–I’d like to see more data, especially stratified by the year the episode(s) aired.
I’m wondering if the inclusion of Mad Men skews the data significantly: the show’s couples account for 7% of the total number of couples, but the show’s explicitly retro setting seems to dictate that questions of interraciality and non-heteronormative couples won’t come up/are less appropriate in that context. Even so, there have been interracial couples depicted on the show: Paul Kinsey dated Sheila White (Caucasian/black) and Lane had an affair with Toni Charles (Caucasian/black).