Chapter 25: Race, Ethnicity, and the Politics of Desire — Who Is Desirable and Why?

"We cannot separate what we find beautiful from what we have been taught to see as beautiful." — Toni Morrison

Opening: The Preprint That Caused a Firestorm

It was a Wednesday morning in March when Dr. Adaeze Okafor received the email that would define her year. A right-wing website had picked up a preprint — a paper she and Dr. Carlos Reyes had posted three days earlier, a supplement to their Global Attraction Project analyzing racial preference patterns they had observed in a dating app behavioral dataset. The paper had been posted to a preprint server with careful caveats and a detailed methodological appendix. It was intended for academic peer review.

By Wednesday morning, the paper had been stripped of its caveats. Its most provocative numbers — cross-racial match rates organized into a hierarchy — had been extracted into an infographic circulating on white nationalist Telegram channels. The framing: Science proves racial preferences are natural and that certain groups are simply more desirable. By noon, Okafor was receiving threatening emails. By that evening, she had been mentioned in fourteen social media posts from accounts with Nazi imagery in their profiles.

Reyes called her at 10 p.m. "We need to talk about whether we can still publish this."

This scenario — fictional here but closely modeled on real events that have confronted researchers including Christian Rudder, Elizabeth Bruch, and Jennifer Lundquist — is where this chapter begins, because it is the most honest entry point into its subject matter. The data on racial preferences in dating is real. The racial hierarchies it reveals are real. And those hierarchies are immediately, predictably weaponizable by people who want to use "science" to naturalize white supremacy.

The question before Okafor and Reyes is also the question before this chapter: How do we study something uncomfortable, report it honestly, and refuse to surrender our findings to those who would use them as ammunition?

The answer — the chapter argues — is that we do it anyway, carefully, with structural analysis as the central frame. The hierarchy exists. It is not natural. It was built, and understanding how it was built is the only way to dismantle it.

25.1 What the Data Actually Shows: Racial Preferences in Online Dating

Let's begin with what the data says before we say anything else.

Christian Rudder, co-founder of OkCupid, published an analysis of his platform's messaging and rating data in 2014 in his book Dataclysm. What he found was stark: across every racial and gender combination, White users received the highest ratings from all groups. Black women and Asian men received systematically lower ratings. Users overwhelmingly favored their own racial group, but these in-group preferences were not symmetric — White users' in-group preference was stronger and more exclusionary than that of most other groups. And critically, patterns were not random noise: they were consistent, robust, and tracked along recognizable lines.

The data showed something else, too, that is frequently omitted from popular summaries: the hierarchy was present not only in explicit ratings but in messaging behavior. White users sent messages at similar rates to other White users and to users of most other racial groups — but responded to cross-racial messages at rates that followed the hierarchy. Black users sent messages at high rates but received responses at lower rates than sending rates would predict. The gap between sending behavior and response behavior is analytically important: it means racial hierarchy is not only about who users approach (which might reflect self-selection) but about who users choose to engage when approached (which reflects evaluative hierarchy more directly).

Subsequent research confirmed and complicated these findings. Cynthia Feliciano, Belinda Robnett, and Golnaz Komaie (2009) analyzed Yahoo! Personals profiles and found that racial exclusions were explicit and widespread — a majority of White men and women stated racial preferences in their profiles, with Black partners most commonly excluded. Josue Ortega and Philipp Hergovich's (2018) network analysis of dating app data found that interracial marriages had increased significantly since the rise of online dating, but this masked continued intra-racial preferential matching at the behavioral level. Elizabeth Bruch and M.E.J. Newman (2018), using a large proprietary dating site dataset, found that aspirational messaging — contacting users rated more desirable than oneself — was nearly universal across the platform, and that the racial structure of these aspirational patterns closely tracked the documented hierarchy: users of lower-ranked racial groups sent aspirational messages upward in the hierarchy at higher rates, while users of higher-ranked groups sent cross-racial messages less often and responded to upward-aspiring cross-racial messages at lower rates.

This body of evidence is not from a single dataset or a single methodological tradition. It converges from behavioral analysis, explicit stated preferences, network analysis, and message response patterns. The convergence strengthens the conclusion that what we are observing is a structural phenomenon, not a measurement artifact.

The Swipe Right Dataset — our synthetic 50,000-profile dataset modeled on patterns from this literature — shows a familiar pattern when we analyze racial matching rates. The Python analysis in this chapter's code/racial_preference_analysis.py generates a racial matching matrix and calculates an in-group preference ratio (IPR) for each racial group: how much more likely is a user to match within their racial group than we would expect by chance if preferences were randomly distributed?

📊 Research Spotlight: In-Group Preference Ratios from the Swipe Right Dataset

When we run this analysis on our synthetic data, we find IPR values ranging from 1.3 to 2.8 — meaning users are between 30% and 180% more likely to match within their racial group than chance would predict. But crucially, these ratios are not symmetric: cross-racial match rates also follow the documented hierarchy, with White profiles receiving higher cross-racial match rates from all groups, and Black profiles — particularly Black women's profiles — receiving lower cross-racial match rates from all non-Black groups.

These are not findings about individual hearts. They are structural patterns. They tell us about the aggregate shape of desire in a racially structured society. Crucially, the fact that they are aggregate patterns does not mean they are determinative of any individual experience — but it does mean that any individual experience of racial hierarchy in dating is not idiosyncratic. It is the expression of a structure that is being felt by millions of people simultaneously.

25.2 The Racial Hierarchy Problem: Not All Groups Are Equally Desired

Here is where we must be unflinching — and precise about what "unflinching" means.

The literature does not show that every racial group has equal levels of desirability in the aggregate dating market. It shows a hierarchy — one that maps remarkably closely onto historical hierarchies of racial value constructed during European colonialism. At the top of cross-racial desirability scores, consistently, are White people. Closest to them, in most U.S. datasets, are East Asians (with important gender asymmetries we will discuss). Black men receive moderate cross-racial desirability from some groups; Black women receive the lowest cross-racial desirability ratings in most datasets. Latino and Latina individuals occupy intermediate and internally differentiated positions. South Asians, MENA (Middle Eastern and North African) individuals, and Indigenous Americans appear inconsistently in data due to sample size limitations.

⚠️ Critical Caveat: Dataset Limitations Are Structural, Not Incidental

Before proceeding, we must name a methodological problem embedded in all of this data: the datasets themselves reflect the composition of the platforms and the categories used to classify race. OkCupid's user base skews young, educated, and urban. Dating apps' race categories are user-selected and vary by platform. When a user selects "Asian," we do not know whether that means East Asian, South Asian, Southeast Asian, or Pacific Islander — groups with very different social positions. Lumping them together produces artifactual conclusions. The hierarchy we observe is real within these datasets, but the datasets themselves are artifacts of a racially structured world.

A second methodological note: measuring "desirability" through platform interactions captures only one dimension of a complex social phenomenon. Dating app behavior reflects not only genuine preference but also platform-induced behavior (what the interface rewards), safety concerns (particularly for women and LGBTQ+ users who may filter for perceived safety rather than pure preference), and strategic behavior (sending messages to users one believes will respond rather than users one actually prefers). All of these complicate the interpretation of aggregate match-rate data.

This caveat is not a reason to dismiss the data. It is a reason to interpret it structurally rather than as a neutral map of "natural" preferences, and to hold the specific numbers loosely while taking the overall pattern seriously.

💡 Key Insight: The Hierarchy Maps Onto History

If the hierarchy of desirability in 2024 U.S. dating apps closely resembles the racial hierarchy constructed by eighteenth- and nineteenth-century European colonialism — with Whiteness at the top and Blackness most excluded — this is not coincidence. It is the product of centuries of propaganda, legal enforcement, and cultural production designed to make Whiteness appear natural, universal, and beautiful and to make Blackness appear threatening, animalistic, and undesirable. The dating app does not generate the hierarchy. It reflects and perpetuates it.

25.3 The Colonial Inheritance: How History Shapes Contemporary Desire

The idea that "I'm just not attracted to X" is a purely personal, spontaneous, apolitical preference is a comforting fiction. Desire does not originate in a cultural vacuum. It is formed through:

Explicit historical propaganda. During chattel slavery, enslaved Black people were systematically portrayed in proslavery media as sexually dangerous and simultaneously as animalistic — a contradiction designed to justify both the terror of slavery and the policing of interracial intimacy. After abolition, this propaganda intensified through the "Lost Cause" mythology, eugenics movements, and anti-miscegenation laws (which remained on the books in seventeen U.S. states until Loving v. Virginia in 1967 — just one generation ago).

Media representation as ideological training. Hollywood cinema has historically positioned White femininity as the apex of beauty and desirability, creating what scholars call the default White problem. When beauty is visualized — in rom-coms, in advertisements, in romance novel covers — the default protagonist is White, thin, and conventionally feminine. Deviation from this default requires explicit marking and often othering. Non-White characters in romantic roles are frequently exoticized, hypersexualized, or desexualized depending on their race and gender. These representations are not innocent: they are millions of hours of cultural training in what and who is desirable.

Legal enforcement of racial separation. Anti-miscegenation laws did not simply prohibit interracial marriage — they enforced the social perception that interracial desire was abnormal, dangerous, or criminal. When legal systems treat a desire as aberrant, they shape the phenomenology of desire itself. The generation that grew up under these laws transmitted, consciously and unconsciously, the attitudes those laws instantiated.

🔵 Ethical Lens: Fanon on Colonial Desire

Frantz Fanon's Black Skin, White Masks (1952) offers a psychoanalytically inflected account of how colonialism structures desire in both the colonized and the colonizer. For the colonized subject, Fanon argued, desire has been deformed by the internalization of the colonizer's value system — including its racial hierarchy of beauty. The colonized person who desires Whiteness is not simply expressing a preference; they are navigating a world in which their own image has been systematically devalued and in which proximity to Whiteness has historically offered social survival. This is not false consciousness but rather a rational response to an irrational structure. Understanding desire means understanding the structure that produces it.

Intersectional compounding. Kimberlé Crenshaw's framework of intersectionality (1989, 1991) reminds us that race does not operate alone. The desirability hierarchy looks different for Black women than for Black men; different for Asian women than for Asian men; different for wealthy Black professionals than for lower-income Black people; different for light-skinned versus dark-skinned individuals within any racial group. Race compounds with gender, class, colorism, and sexuality to produce not one hierarchy but a web of interlocking positions.

25.4 Colorism Within Racial Groups: The Hierarchy Within the Hierarchy

Racial hierarchy is not only an inter-group phenomenon. Within racial groups, skin tone itself is a source of differential desirability — a phenomenon called colorism, a term coined by Alice Walker in 1983.

Research consistently finds that lighter-skinned individuals within Black, Latino, South Asian, and East Asian communities receive higher desirability ratings, better economic outcomes, and more favorable treatment in dating contexts. This is not a coincidence or a mysterious quirk of human perception. It is the direct inheritance of colonialism's association of Whiteness with civilization, intelligence, and beauty — an association that was enforced with legal sanction, economic benefit, and violence.

Margaret Hunter's research on colorism in the United States documents that lighter-skinned Black and Latina women are rated as more attractive by both White and non-White evaluators, receive more dating interest, and are more likely to marry men of higher socioeconomic status. This "beauty premium" for lighter skin operates even controlling for other physical features — it is skin tone itself that carries the cultural encoding.

Within South Asian communities, skin-whitening products represent a multi-billion dollar industry. Within Black American communities, the historical distinction between house slaves (often lighter-skinned, favored because of White ancestry) and field slaves has left a legacy of intragroup colorism that continues to shape social dynamics, including romantic ones. Within East Asian communities, paleness has been coded as beautiful for centuries — but this has been intensified and given new global reach through Western colonial contact.

📊 Research Spotlight: The "Skin Tone Premium" in Dating Data

A 2012 analysis by psychologists at the University of Georgia found that lighter skin tone predicted higher attractiveness ratings and more romantic interest from evaluators, controlling for facial features. Hersch (2008) found that among Latinos and Blacks, lighter skin was associated with a statistically significant earnings premium — meaning colorism operates not only in romantic contexts but across domains of social life, reinforcing that it is structural, not perceptual.

In the Swipe Right Dataset, users who self-report as "light" or "medium" skin within traditionally darker-skinned racial groups show measurably higher match rates than users who self-report as "dark" — a synthetic pattern deliberately modeled on these documented effects. The code in this chapter visualizes these patterns.

25.5 "Just a Preference": The Philosophical Challenge

The most common defense of racialized dating preferences is: "It's just a preference. I can't control who I'm attracted to. Calling it racist is policing my sexuality."

This argument deserves serious engagement rather than dismissal, because it contains a kernel of legitimate concern. Desire is not fully voluntary. We did not choose our attractions through deliberate decision-making. Telling people they must be attracted to categories they are not currently attracted to is both unrealistic and uncomfortably coercive.

But the "just a preference" defense fails on several grounds:

First, the empirical critique. If racial preferences were simply the outcome of spontaneous individual variation, we would expect them to be randomly distributed across racial groups. They are not. They are patterned, consistent, and correlated with historical racial hierarchies. The non-randomness demands a structural explanation that "individual preference" cannot provide.

Second, the philosophical critique. Philosopher Robin Zheng (2016) draws a crucial distinction between being racist and having preferences that are the product of racist socialization. The latter does not make the individual a moral monster. But it does mean that the "preference" carries a social history and has social effects regardless of individual intent. Saying "I never date Black people" or "I'm not attracted to Asians" is not a neutral statement about interior experience — it is a behavioral policy with racially disparate impacts that also communicates something to the people excluded. The person who never dates Black people may not be consciously racist, but their dating behavior contributes to the aggregate pattern that tells Black people they are less desirable.

Third, the mutability critique. If racial preferences were genuinely fixed, we would not observe them changing across generations, across media environments, and in response to cross-racial friendships and contact. Research on the "contact hypothesis" (Allport, 1954; Pettigrew & Tropp, 2006) consistently finds that meaningful cross-racial contact reduces racial prejudice — including racialized aesthetic evaluations. The fact that preferences can shift under social conditions is evidence that they are learned, not hardwired.

⚖️ Debate Point: The Limits of "Racial Justice in Desire" as a Framework

Some scholars and activists argue that individuals have an ethical obligation to examine and challenge their racialized preferences — to do the work of noticing when they exclude entire racial groups and to ask why. This "racial justice in desire" framework (drawing on work by Mimi Schippers, Shantel Buggs, and others) is compelling in its analysis but raises a genuine tension: can we really require people to become attracted to specific partners? And does framing desire as a site of political obligation risk sexualizing the project of anti-racism in ways that create new coercions?

There is no clean answer here. The framework's value is not in demanding specific desires but in insisting that our desires are not separate from politics — that the work of racial justice extends into intimate life, not as a mandate for whom to love but as an invitation to notice the architecture that shapes what we find beautiful.

25.6 Fetishization vs. Preference: Can Desire Be Racist?

There is an important distinction between racial exclusion and racial fetishization — though both are forms of racialized desire.

Racial exclusion means never or rarely being attracted to members of a particular racial group. It operates as a boundary: "I don't date X."

Racial fetishization means being attracted to members of a racial group because of racial stereotypes — stereotypes that reduce individuals to their racial identity and project upon them essentialized characteristics the fetishizer finds appealing. It operates as an equation: "I love X people because they are all so [stereotype]."

Fetishization feels, to the person experiencing it, like intense attraction or even positive interest. But its psychological structure is profoundly dehumanizing: the fetishized person is not being seen as an individual but as a representative of a racial type. Their individuality is irrelevant; they are desired as an avatar of racial meaning.

The most extensively documented case in the U.S. context involves the fetishization of Asian women — the subject of Case Study 25.2 — where a cluster of stereotypes ("lotus flower," "submissive," "sexually exotic") has been so thoroughly cultivated through decades of media, military occupation, and colonial fantasy that many White men who prefer Asian women have no conscious awareness that their attraction is organized around stereotypes rather than individual perception.

💡 Key Insight: Fetishization Structures the Relationship from the Start

When someone enters a relationship animated primarily by racial fetish, the relationship is structured as a performance of the fetish rather than as an encounter between two people. Partners of fetishizers often report feeling surveilled for conformity to racial type, experiencing rejection when they deviate from the stereotype, and having their personality and biography treated as peripheral to their racial identity. The fetish is, in Fanon's terms, a colonial fantasy projected onto a body.

This is distinct from, say, having an aesthetic preference for darker skin tones or a cultural affinity developed through genuine cross-cultural experience and relationship. The difference is in the structure of the perception: does the person see an individual who happens to be of a certain racial background, or do they see a racial type who happens to have an individual attached?

🔴 Myth Busted: "Racial Fetishism Is Just Having a Strong Type"

Having a "type" in the conventional sense means finding certain physical features consistently attractive — height, hair color, body shape. These preferences, while also socially constructed to a degree, do not carry the same historical weight or dehumanizing logic as racial fetishism. Racial fetishism is not about physical features; it is about the social meanings that have been attached to racial identity through centuries of colonial fantasy. Reducing racial fetishism to "having a type" obscures the structural difference between personal aesthetics and racial ideology.

25.7 The Okafor-Reyes Dilemma: When Data Gets Weaponized

Return to that Wednesday morning.

Okafor and Reyes spent three days in urgent conversation before deciding what to do. The preprint could not be un-posted — it was already circulating. But the peer-reviewed version would be their opportunity to control the framing.

Reyes initially wanted to simply add a stronger caveats section. Okafor pushed further: the paper itself needed to be restructured so that the structural analysis was not buried in discussion but was the organizing frame for the entire presentation of data. "If we lead with the matrix of match rates," she argued, "the numbers become the story. We need the historical analysis to be the story, and the numbers to be evidence of structural racism — not evidence of natural hierarchy."

This disagreement reflects a genuine methodological and ethical tension in social science. Presenting data "neutrally" — as mere numbers — is not actually neutral. Numbers, like all texts, are read in context. When the context is a white nationalist information environment eager for data that confirms racial hierarchy, "neutral" presentation becomes functionally supportive of that hierarchy. Scientific framing is never innocent of its social context.

They reached a compromise: the paper would be restructured, with the historical and structural analysis placed first and the quantitative data explicitly framed as evidence of structural racism's effects. They would also add a "misuse and misinterpretation" section explicitly addressing how the data should not be used and why common misreadings were methodologically invalid.

🔵 Ethical Lens: Researcher Responsibility for Data's Social Life

The Okafor-Reyes dilemma raises a question social scientists have been grappling with since the Nazi eugenicists: What responsibility do researchers bear for how their work is used? The answer is contested. Some argue for a strong version of value-neutrality: publish the data, and society will sort out its meaning. Others argue that in a world of information asymmetry and bad-faith actors, researchers must take responsibility for the framing and communication of their findings, not only their methods. Most contemporary research ethics frameworks have moved toward the latter view — requiring researchers to anticipate misuse and build structural guards against it, without suppressing findings entirely.

25.8 Sam's Experience: Being Biracial in a Racialized Desire Market

Sam Nakamura-Bright has a complicated relationship with his own desirability.

His mother is a Japanese immigrant who came to Portland for graduate school; his father is a Black American who grew up in Chicago. Sam himself reads differently in different contexts: to some people, he presents as Asian; to others, as Black; to many, as ambiguous. He has been told he is "exotic," "intriguing," "hard to place." He has also been underprioritized in dating contexts by people who, it seemed clear, were filtering him out — though whether because he was too Black or too Asian or too mixed was always unclear.

The biracial experience in a racialized desire market is one of being simultaneously overdetermined (assigned racial meaning before you can introduce yourself) and underdetermined (not legibly fitting any of the boxes the market uses). Research on multiracial individuals in dating contexts finds that biracial people face unique challenges: they are often perceived as racially ambiguous in ways that increase approach anxiety in potential partners (who are uncertain how to categorize them); they may be fetishized for their "exotic" appearance while simultaneously being excluded by both of their home racial communities; and they often experience their own desire as navigating an impossible set of expectations.

For Sam, the particular intersection of Japanese and Black heritage is significant. Asian men and Black women are, as we have documented, the most disadvantaged by the aggregate desirability hierarchy — yet the intersection of these two identities in a single body does not produce a simple combination of disadvantages. It produces something more complex: a body that confounds categorical expectations and that consequently does not fit cleanly into anyone's pre-mapped desirability framework.

Sam has thought about this a lot. He knows, abstractly, that the way people respond to him in dating contexts is not about him. But the felt experience of being categorized, fetishized, or dismissed — of having your individual selfhood persistently overridden by others' racial projections — does not feel abstract. It is intimate, recurring, and specific to the architecture of desire in a racist society.

25.9 How Dating App Design Encodes Racial Bias

The racial hierarchy in dating apps is not only a reflection of social attitudes. It is actively produced and amplified by the platforms themselves — through deliberate design choices, algorithmic systems, and the structures of engagement they incentivize.

Racial filtering features. Most major dating apps have at some point offered explicit race filtering — the ability to see only profiles of users of specified racial backgrounds. OkCupid removed its race filter in 2020, citing concerns about racial discrimination. Hinge and other platforms have faced ongoing criticism for maintaining features that enable racial exclusion. The filter makes racial exclusion frictionless and explicit — and normalizes it as a legitimate dating preference indistinguishable from filtering by height or age.

Algorithmic amplification. Dating app algorithms optimize for engagement: they surface profiles that are rated highly and that generate response behavior. If users from certain racial groups receive systematically lower ratings — as the literature documents — the algorithm learns to surface those profiles less often, reducing their visibility, which reduces their ratings further, which reduces their visibility further. This is a feedback loop that encodes and amplifies racial bias without any individual programmer intending racial discrimination. The algorithm is not racist in intent; it is racist in effect.

⚠️ Critical Caveat: "Race-Blind" Algorithms Are Not Neutral

When platforms claim their algorithms are "race-blind" — that they do not explicitly use race as an input — they are making a technically true but functionally misleading claim. Algorithms trained on user behavior learn from racially structured behavior. A race-blind algorithm that learns from ten million swipes in a racist society will reproduce the racial preferences embedded in those swipes. The absence of an explicit race variable does not produce racial neutrality; it produces racial bias laundered through the language of meritocracy.

The "ELO-style" desirability score. Several platforms, including Tinder in its earlier design, used an ELO-style ranking system — borrowed from chess rating — to rank users by desirability based on swipe behavior. This made racial hierarchy not merely a user-level preference but a platform-level architecture: your desirability score determined who saw your profile. Users in lower desirability tiers — which tracked racial identity through the mechanisms described above — were shown fewer profiles and had their own profiles shown less. The design did not merely reflect the hierarchy; it institutionalized it.

Visual default settings. The way profiles are displayed — image-first, with identity text secondary — encodes a visual bias toward conventional attractiveness standards that are themselves racially encoded. Platforms that prioritize or expand bio text or shared interests over photos create modestly more equitable matching environments, though the overall effect is limited.

25.10 Interracial Relationships: What People Say vs. What They Do

There is a persistent gap between Americans' stated attitudes about interracial relationships and their actual behavior.

Gallup polling has tracked approval of Black-White marriages since 1958. In 1958, 4% of Americans approved. By 2021, 94% approved. Over this same period, interracial marriage rates have risen — but the rise has been slower, and the composition is not uniform. In 2019, about 19% of newlyweds in the U.S. were married to a spouse of a different race or ethnicity — up from 3% in 1967. But this aggregate obscures significant variation: interracial marriage rates are highest for Asian Americans (particularly women), intermediate for Latino Americans, and lower (though rising) for Black Americans. And approval rates differ by question type: approving of interracial marriage in general is easier than reporting openness to personally marrying across racial lines.

This gap between stated attitudes and behavioral reality is not hypocrisy so much as a structural effect: individuals can sincerely endorse interracial relationships at an abstract level while having racialized aesthetic responses that shape their actual desires. The social desirability of anti-racism coexists with racial preferences that have not been subjected to the same critical scrutiny.

The internet has accelerated one form of cross-racial contact — exposure to diverse populations through social media, streaming, and dating apps — while creating new mechanisms for racial filtering and segregation. The net effect on interracial relationship rates appears to be modest: interracial marriage has increased in the app era, but the hierarchy of desirability within apps has not noticeably flattened.

25.11 The "Default White" Problem in Media and Desire

When a romance novel does not specify the protagonists' race, readers typically imagine them as White. When a romantic hero in a Hollywood film is not explicitly marked as non-White, they are White by default. This phenomenon — the "default White" — is not merely a representational problem. It is a structure that positions Whiteness as universal and all other racial identities as particular, marked, and in need of explanation.

The consequences for desire are concrete. Attraction is partly learned through exposure, fantasy, and narrative identification. When the dominant cultural archive of romantic desirability is White, non-White individuals grow up with fewer cultural templates for imagining themselves as the hero or heroine of a love story. They also grow up with fewer templates for imagining non-White people as the generic object of romantic aspiration.

This is not uniform — non-White communities create their own media, their own beauty ideals, their own romantic narratives — but these exist in tension with the dominant culture's imagery and in a market that consistently underfunds and underplatforms them. The recent surge in romance novels and streaming series featuring non-White protagonists (driven in part by social media activism and platform diversification) represents a genuine shift — but a shift beginning from a baseline of pervasive under-representation.

🔗 Connections: Chapter 35 — Media and Romantic Scripts

Chapter 35 examines how media shapes romantic expectations in more depth, including the "manic pixie dream girl" trope, the rom-com formula's assumptions about class and gender, and the research on how viewing habits correlate with relationship expectations. The racial dimension of media desirability documented here connects directly to those broader arguments about narrative scripts and desire formation.

25.12 Activist and Theoretical Responses: Frameworks for Change

If racial hierarchy in desire is structural rather than inevitable, what do researchers, activists, and theorists propose?

Exposing and naming structural patterns. The first act is what this chapter has attempted: documenting the hierarchy, naming its origins in colonial history and media representation, and refusing the "just a preference" evasion. Naming the structure makes it visible, and visible structures can be challenged.

Media diversity as a long-term intervention. Research on parasocial relationships — the sense of connection people develop with media characters — suggests that exposure to desirable, complex, romantic non-White characters in mainstream media shifts aesthetic norms over time. This is a slow process requiring sustained cultural change rather than individual-level behavior.

Platform accountability. Dating platforms can and have made design choices that reduce racial discrimination: removing race filters, auditing algorithmic outputs for racial disparate impact, centering biographical information over photos, and experimenting with designs that reduce early-stage filtering by race. These changes face commercial resistance because they may reduce engagement metrics in the short term, but they represent genuine policy levers.

The "critical desire" framework. Scholars including Celine Parreñas Shimizu (on Asian American women and fetishism), Mireille Miller-Young (on Black women and sexuality), and others in critical race and sexuality studies argue for what we might call critical desire — the practice of noticing our own racialized desires, tracing their origins, and refusing to treat them as fixed or as beyond ethical examination. This is not about guilt or self-policing but about curiosity: Why do I find this beautiful? What has shaped that response? This framework aligns with the broader arc of this textbook's argument that desire is not destiny but history made intimate.

⚖️ Debate Point: Individual Ethics vs. Structural Change

A persistent tension in this literature is between individual-level ethics (each person should examine their desires) and structural analysis (this requires systemic change, not individual introspection). These are not mutually exclusive, but they have different political implications. Overemphasizing individual ethics risks placing the burden of racial justice on the intimate choices of individuals — a burden that falls disproportionately on those most harmed by the hierarchy. Overemphasizing structural change risks treating individuals as passive products of systems with no capacity for growth or examination. The most productive frame, this chapter suggests, holds both: structural change is necessary and individual examination is worthwhile, and neither is sufficient alone.

25.13 Undifferentiated vs. Differentiated Racial Preferences: A Crucial Distinction

One of the more nuanced contributions to this field involves distinguishing between two structurally different types of racialized preference. Undifferentiated racial preference operates as a blanket policy: "I don't date Black people" or "I only date White people." The racial category itself functions as the sole filtering criterion. Differentiated racial preference is more complex: it involves finding certain people of a given racial background attractive while not finding others so, based on a combination of physical features, personality, and shared cultural background — a combination in which racial identity is one factor among many rather than the whole story.

The distinction matters because these two patterns reflect quite different psychological structures and have quite different implications.

Undifferentiated racial preference — the blanket policy — most closely resembles the aggregate patterns documented in OkCupid and Yahoo! Personals data, where significant numbers of users stated explicit racial exclusions ("no Black men," "White women only"). This pattern is hard to explain without reference to racial stereotyping: if you have never dated a person of a given racial background and are unwilling to consider it regardless of individual characteristics, you are effectively treating racial identity as a disqualifying characteristic — applying a group-level judgment to individuals.

Differentiated preferences are more ambiguous. The person who has dated across racial lines, who finds some members of every racial group attractive and others not, who can identify specific individuals of every racial background as desirable, is operating in a different register. Their attraction may still reflect racialized aesthetic norms (the particular features they find beautiful may be coded racially in the broader culture), but they are not categorically excluding entire groups from consideration.

⚖️ Debate Point: Does the Distinction Excuse the Pattern?

Some scholars argue that the undifferentiated/differentiated distinction lets individuals off the hook too easily. Even a person whose preferences are differentiated rather than categorical is still operating within a desirability hierarchy shaped by racial history, and their aggregate choices contribute to the documented patterns. Robin Zheng (2016) is careful to note that the question is not only about individual culpability but about social effects — and the social effects of differentiated preferences are not categorically different from those of undifferentiated ones. Both contribute, in aggregate, to the hierarchy. The distinction is ethically relevant but not exculpatory.

25.14 Race, Gender, and the Double Bind of Intersectional Desire

The racial hierarchy of desire does not operate uniformly across gender lines. One of the most consistent and important findings in this literature is that racial disadvantage and racial advantage are both gendered — that the same racial background can produce very different positions in the desirability hierarchy depending on whether you are a man or a woman, and that these differences often involve a cruel inversion.

The Asian gender asymmetry is perhaps the most documented example. In most U.S. dating datasets, Asian women receive higher cross-racial desirability ratings than would be predicted by their proportion of the population, while Asian men receive lower cross-racial desirability ratings. Both effects are attributable to the same underlying phenomenon: the racialized stereotype of Asian femininity as submissive and sexually available (which increases desire for Asian women among some non-Asian men) and Asian masculinity as emasculated, technically competent but sexually passive (which decreases desire for Asian men among non-Asian women). One racial stereotype is generating opposite outcomes for Asian men and Asian women — it is the same architecture of racist imagery, with gender determining whether you end up at a structural advantage or disadvantage in the market.

The Black gender asymmetry operates differently but comparably. Black men receive moderate cross-racial desirability — in some datasets, particularly among White women, rates that exceed what simple demographic probability would predict. Black women receive the lowest cross-racial desirability of any group in most U.S. studies. Again, racial stereotyping is the underlying driver: the hypersexualization of Black masculinity (a racist trope dating to proslavery propaganda) generates a particular form of racialized desire toward Black men, while the systematic desexualization and devaluation of Black women's bodies — rooted in the same propaganda tradition but operating through a different archetype — produces exclusion rather than hypersexual fetishization. Both are forms of dehumanization operating through opposite desirability outcomes.

💡 Key Insight: Hypersexualization Is Not the Same as Desirability

It is important not to confuse being hypersexualized with being genuinely desired. The hypersexualization of Black men in American culture generates a specific, fetishized form of interest that bears all the marks of fetishization described in Section 25.6: it is based on racial stereotype rather than individual encounter, it produces its own distinctive form of dehumanization, and it is not the same as being seen as a full human being capable of romantic love and intimate partnership. Black men who navigate this form of racialized desire frequently report the same invisibility described by Asian-American women in fetishizing relationships — being desired as a racial archetype rather than as a person.

The Latino/a asymmetry involves its own internal complexity, which colorism intensifies. Lighter-skinned Latinos and Latinas tend to receive significantly higher cross-racial desirability ratings than darker-skinned ones — and the gender dynamics within this are shaped by intersecting stereotypes of Latina hypersexuality, Latino machismo, and "passionate" Latin culture that affect men and women differently. The internal colorism of Latino desirability patterns is both a consequence of colonial racial hierarchy and a perpetuation of it.

Intersectionality is not a theoretical luxury in this analysis. It is the minimum analytic requirement for making sense of the data.

25.15 Class, Race, and the Compounding of Desire

The racial hierarchy of desire does not operate in isolation from socioeconomic status. Research consistently finds that class resources — income, education, social capital, embodied markers of professional status — can partially offset racial disadvantage in the desirability market, while their absence can amplify it.

This is documented most clearly for Black Americans, where studies find that Black men and women with higher educational attainment and professional status receive meaningfully higher cross-racial desirability ratings than comparable individuals with fewer class resources. The partial offset is real — but it is incomplete, it is not symmetrical across all racial groups, and it comes at its own cost: the expectation that Black individuals must achieve significantly higher class status to compete for the same desirability outcomes as lower-class White individuals is itself a form of racial taxation.

Sociologist Eduardo Bonilla-Silva has argued that the United States is developing a more complex racial order in which middle-class non-White individuals are increasingly incorporated into a "honorary White" status — gaining proximity to Whiteness's social privileges through class achievement, which includes desirability. If this analysis is correct, class resources do not simply add to racial status; they partially substitute for it in some social domains. This is not racial equality; it is racial hierarchy with a permeable upper boundary that admits some non-White people under specific conditions.

🔵 Ethical Lens: The Racial-Class Compound in Sam Nakamura-Bright's Experience

Sam's situation illustrates the compound of race and class in desirability dynamics. As a computer science major with strong professional prospects, he has class resources that some of his peers lack. But these do not translate straightforwardly into desirability in his experience. His biracial identity — Japanese and Black — does not fit any of the simple scripts that class resources partially offset racial disadvantage for. He is not neatly assignable to the "honorary White" proximity that middle-class Asian identity can sometimes provide, because his Black heritage complicates that assignment. Nor does the hypersexualized racial script for Black men apply cleanly, because his Asian heritage complicates that script too. His experience in racialized desire markets is one of persistent categorical illegibility — and class resources cannot resolve that illegibility, because it is not fundamentally about class.

25.16 Global Dimensions: Racial Hierarchy Is Not Only American

The racial hierarchy of desire documented in U.S. dating data is not exclusively a product of American racial history, though American history is its most important context for readers of this textbook. Analogous hierarchies have been documented across the globe, shaped by local colonial histories and media industries.

In India and South Asia, colorism has deep historical roots predating British colonialism but was intensified and systematized by British rule, which associated fair skin with the colonial class and darker skin with agricultural labor and low caste. Contemporary matrimonial advertising in India routinely specifies skin tone preferences — almost always favoring fairer skin — and the skin-whitening product industry generates over $500 million annually in India alone.

In Latin America, the concept of blanqueamiento (whitening) — the post-colonial ideological project of racially "improving" the population through lighter-skinned marriage partners — explicitly linked desirability to racial proximity to Whiteness. The aesthetic preferences this ideology generated persist across Latin American cultures, with lighter skin, European features, and straight hair consistently associated with higher desirability and social value.

In East Asia, pale skin has been associated with beauty for centuries — predating Western contact — but was given new global reach and pharmaceutical infrastructure by Western colonial presence and American media dominance in the post-WWII period. Korean cosmetics and Japanese media export beauty standards rooted in relative paleness that then circulate globally through the Korean Wave (hallyu) and anime.

📊 Research Spotlight: The Okafor-Reyes Global Attraction Project

Adaeze Okafor had insisted, from the project's earliest design phases, that the Global Attraction Project include explicit modules on colorism within each of the twelve study countries. Her methodological argument was that any cross-cultural attractiveness data that did not measure within-country color preference would simply reproduce the Western Black/White binary rather than capturing the full complexity of racial desirability. The colorism module — measuring within-group preferences for skin tone using digitally manipulated photographic stimuli presented to local participants — is among the study's most carefully designed components. Early data from the Brazil, India, and South Africa sites already suggest that within-country colorism effects are at least as large as between-country racial effects — a finding that, if it holds through peer review, will have significant implications for how we model the relationship between racial hierarchy and desire globally.

25.17 The Lived Experience: Jordan, Nadia, and Sam in Conversation

After their SOCL 312 seminar covered a week of readings on racial hierarchy in dating — including Rudder's OkCupid analysis, Feliciano et al.'s Yahoo! Personals study, and excerpts from Zheng's philosophical work — Jordan, Sam, and Nadia found themselves in a long, charged conversation over late-night pizza.

Jordan went first, in the characteristically analytical mode they defaulted to when a topic hit close to home: "The thing that gets me is the precision of the hierarchy. It's not random. It's not 'people just happen to prefer their own group.' The cross-racial numbers map exactly onto what you'd predict if you'd designed a system from scratch to validate colonial racial value. That's not a coincidence. That's a document."

Nadia was quieter than usual. She'd grown up in a mostly Arab and South Asian community in Dearborn and had been reading the chapter through the lens of her own dating history. "The MENA numbers are bad but they're not the worst," she said finally. "And I think about how I navigate being Arab in dating contexts — not just the discrimination but the fetishization. The 'exotic' comments. The guys who want a 'mysterious Middle Eastern woman.' I'm Lebanese-American. I grew up in Michigan. There's nothing mysterious about my life."

Sam had been listening more than talking. He pulled at the label on his water bottle. "The biracial stuff resonates in a specific way," he said. "Because I don't get the clear narrative on either side. I'm not simply Asian-disadvantaged or Black-disadvantaged. I'm just — illegible. People don't know what to do with me. And I think that illegibility is its own form of what the chapter is describing. The system is built on legible racial categories. If you don't fit a category, the system doesn't have a script for you."

"That's the thing about intersectionality as a lived experience rather than an analytical tool," Jordan said. "In theory, multiple marginalized identities intersect to produce a distinctive position. In practice, it often just means you're a problem that the system wasn't designed to handle."

They sat with that for a while. There wasn't much to add.

This conversation is not unusual among students who encounter the data in this chapter for the first time. The patterns described are not abstractions for the people they describe. They are the texture of dating app notifications, the experience of being categorized before you've introduced yourself, the weight of a hierarchy you did not choose and did not build.

25.18 The Swipe Right Dataset: Running the Analysis

The Python analysis in code/racial_preference_analysis.py gives students hands-on access to the patterns described in this chapter. The script generates a synthetic 50,000-profile dataset modeled on patterns documented in the literature, calculates cross-racial match rates, and produces two visualizations:

1. The Racial Matching Matrix (Heatmap): Rows represent the race of the person swiping; columns represent the race of the profile being swiped. Cell values show the match rate (percentage of swipes resulting in matches) for each racial combination. Structural inequality shows up as asymmetry in the matrix — not just high in-group matching but differential cross-racial match rates that favor White profiles.

2. In-Group Preference Ratio Bar Chart: For each racial group, the ratio of within-group match rate to the expected match rate if preferences were random. This visualizes not just that in-group preference exists (it does for all groups) but that it varies in intensity and that cross-racial patterns differ by group.

The script includes prominent comments on ethical interpretation: what the data can and cannot show, how to resist naturalizing interpretations, and how to use the analysis to illuminate structural patterns rather than rank individuals.

🧪 Methodology Note: Synthetic Data and Structural Fidelity

The Swipe Right Dataset is synthetic — not real user data — which raises the question of what it can demonstrate. Synthetic data modeled on documented patterns from peer-reviewed literature can illuminate the structure of patterns without requiring us to use actual user data (which carries privacy and ethics issues). But this also means students should treat the specific numbers in their analysis as illustrative rather than as empirical findings. The patterns are realistic; the precise magnitudes are calibrated for analytical clarity rather than as claims about real platforms.

25.19 Teaching This Chapter: A Note on Classroom Dynamics

A brief word to instructors — and to students who may find themselves processing this chapter in community with others.

This chapter presents material that is simultaneously empirical and deeply personal. Students of color in your classroom — and particularly Black women, Asian men, and biracial students — may have direct, ongoing experience of the hierarchy described in these pages. The data does not describe an abstraction for them; it quantifies their lived experience of being systematically underprioritized or fetishized in dating spaces. That experience deserves recognition in the classroom, and the pedagogical question is not only "are students learning the content?" but "are students being asked to perform their own devaluation as an intellectual exercise?"

Several approaches can mitigate this concern without sanitizing the content:

Frame the data as evidence of structure, not identity. The hierarchy describes what a racist society has done to desire — not what individuals are worth. This distinction should be explicit and repeated. Students who are disadvantaged by the hierarchy should not be invited to reflect on why they are less desirable; they should be invited to analyze the structure that produced that outcome.

Use the Okafor-Reyes controversy as a model. Okafor and Reyes's dilemma about how to publish uncomfortable data parallels the classroom's pedagogical challenge: how to present findings that could be experienced as harmful or weaponized without suppressing what is real and important. Walking through their decision-making process with students is itself a methodological and ethical exercise.

The personal reflection exercise (Exercise 25.4) should be handled with care. Consider making it grade-protected (complete/incomplete) and allowing students to complete it privately rather than sharing in a group discussion. The goal is honest introspection, not public disclosure of attractions or racial attitudes.

The discomfort this chapter produces in many readers is not incidental — it is evidence that the chapter is doing what it needs to do. But discomfort should illuminate, not wound. The aim is structural analysis that unsettles unexamined assumptions, not the humiliation or re-traumatization of students who are already living inside the structure.

25.20 Conclusion: Desire Is Not Destiny

The data in this chapter is uncomfortable. It should be. A world in which desire is racially structured — in which centuries of colonial propaganda, legal enforcement, and media indoctrination have produced a hierarchy of desirability that maps onto a hierarchy of racial value — is a world that requires discomfort to understand honestly.

But discomfort is not the end of the story. The racial hierarchy of desire is historical, which means it is contingent. It was built; it can, over generations and with structural change, be rebuilt differently. The same research that documents the hierarchy also documents its malleability: across generations, media environments, contact experiences, and institutional contexts, the hierarchy shifts. It has not shifted fast enough, or far enough. But it moves.

The evidence for malleability is worth dwelling on because it is often obscured by the weight of the hierarchy itself. Gallup's long-running data on interracial relationship approval shows movement from 4% approval of Black-White marriages in 1958 to 94% approval in 2021 — a transformation within living memory that most people in 1958 would have found unimaginable. Representation data from mainstream media shows genuine (if still incomplete) diversification in romantic leads over the past decade. Research on contact theory consistently finds that meaningful cross-racial relationships — friendships, coworkers, neighbors — shift aesthetic and romantic evaluations over time. Rudder himself documented a modest but real decrease in the magnitude of racial preference effects on OkCupid between 2008 and 2014. These shifts are insufficient. They do not undo the hierarchy. But they refute the fatalist position that the hierarchy is natural or immovable.

What is required — and what this chapter has tried to model — is a particular combination of intellectual dispositions: empirical honesty (the hierarchy is real and the data is uncomfortable), structural analysis (it was built, not discovered), historical grounding (it has a genealogy that traces to specific systems of colonial violence), and openness to agency (individuals can notice and examine their own racialized desires; systems can be designed differently; culture can be made differently). None of these dispositions alone is sufficient. Together, they constitute the only intellectually defensible way to engage with this subject.

Okafor and Reyes ultimately published their paper. The peer-reviewed version led with a thirty-page historical analysis of racial hierarchy's colonial construction. The quantitative data was framed explicitly as evidence of structural racism's persistence in digital contexts. They published a plain-language summary alongside the academic paper and worked with a science communication team to pre-emptively brief journalists. The white nationalist sites ignored the reframing — they had their infographic already. But the academic response was substantive, the citation record built appropriately, and the data entered the peer-reviewed literature with its structural frame intact.

It was not a perfect outcome. Okafor received threats for six more months. The infographic continued to circulate. The hierarchy it purported to "prove" continued to operate in millions of daily swipes. But the paper existed in the record as what it was — evidence of a structure that harms people and that requires not individual shame but collective change. And naming a structure is the first step toward undoing it.

That is the only honest conclusion about race and desire this chapter can offer: that what we find beautiful is never merely personal, never merely natural, and never beyond examination. Our desires are historical. They were made by the world we live in. The world can be made differently, and so, over time, can the shape of desire.

The question is whether we are willing to do that work — individually, structurally, and together. The answer this textbook does not provide is a simple one. The answer it insists upon is that the question is the right one to be asking.

Summary

This chapter has presented a rigorous examination of racial hierarchy in dating and desire, arguing that:

• Documented racial preference patterns in online dating data consistently show a hierarchy of desirability that maps onto the colonial racial hierarchy, with White profiles most broadly desired and Black profiles — particularly Black women's — least so across racial lines.
• These patterns are not natural but are the product of colonial history, legal enforcement of racial separation, media representation of Whiteness as the default of beauty, and algorithmic systems that encode and amplify racial bias.
• Colorism within racial groups compounds inter-group hierarchy, with lighter-skinned individuals consistently receiving higher desirability ratings.
• Fetishization is a distinctive form of racialized desire that dehumanizes through stereotyping, distinct from (though often confused with) racial exclusion.
• The "just a preference" defense fails empirically (preferences are patterned and correlated with historical hierarchy), philosophically (preferences have social effects regardless of intent), and empirically (preferences can shift under social conditions).
• Dating app design encodes racial bias through race filters, engagement-optimizing algorithms, and visual-first interfaces that amplify existing preferences.
• The critical desire framework offers an alternative to both shame and dismissal: curiosity about the origins of our attractions as an ethical practice, alongside structural change as the necessary context for genuine transformation.

Next: Chapter 26 examines how socioeconomic status and class intersect with desire — who gets to want, and who gets to be wanted, across lines of economic inequality.

Key Terms

Colorism — differential treatment and valuation based on skin tone within racial groups, favoring lighter skin.

Default White — the cultural tendency to imagine unspecified characters and ideals as White, positioning Whiteness as universal and other identities as marked.

Fetishization (racial) — attraction to members of a racial group based on stereotypes that reduce individuals to racial archetypes rather than seeing them as individuals.

In-Group Preference Ratio (IPR) — a statistical measure of how much more likely users are to match with members of their own racial group than would be expected by chance.

Intersectionality — Kimberlé Crenshaw's framework for understanding how multiple social identities (race, gender, class, sexuality) compound to create overlapping structures of disadvantage and privilege.

Racial hierarchy (in dating) — the empirically documented pattern in which members of certain racial groups receive systematically higher cross-racial desirability ratings, with White individuals typically at the apex.

Racial preference — behavioral patterns of preferring same-race or excluding other-race dating partners, which research shows are patterned and correlated with historical racial hierarchy rather than randomly distributed.