Abstract
Despite the increasing body of literature surrounding online dating preferences, there remains a paucity of research that analyzes whether skin color influences the dating selection process. To fill this empirical gap, the present study uses data collected from 2,024 Asian dating profiles, including the skin tone of the daters, to assess the impact that skin color variation may have on the inter- and intraracial dating preferences of heterosexual males and females as well as gay males and lesbians. This research also examines whether skin tone has a pronounced effect on the relationship between sexual orientation and the willingness to date Asians, African Americans, and Latino/Latinas. The current findings suggest that darker skinned Asians are more likely to state a preference to date African Americans and Latino/Latinas compared to their lighter skinned counterparts; however, they are less willing to date another Asian. The results also document significant interaction effects between sexual orientation and skin color differences. Our findings are discussed in relation to the racial hierarchy of preference and privilege that are inherently linked to the longstanding concept of colorism.
The proliferation of online dating websites over the past couple of decades has seemingly created an enormous virtual romantic shopping mall for singles seeking companionship. With just a click of a mouse, romance “shoppers” can easily select or filter out people based on various characteristics or resources such as age, income, and race/ethnicity. By overtly advertising one’s attributes, the romantic exchange between potential partners becomes less of a “guessing game.” Online daters can now pick exactly who they want to date. The deployment of race/ethnicity by online daters as both a personal resource and a desired characteristic in a romantic partner has provided researchers with a unique opportunity to examine the racial dating preferences (biases) of Internet daters.
For the most part, existing studies tend to examine whether sociodemographic characteristics, such as age, sexual orientation, education, and region of residence, impact inter- and intraracial dating preferences (McGrath, Tsunokai, Schultz, Kavanagh, & Tarrence, 2016; Robnett & Feliciano, 2011; Tarrence, 2017; Tsunokai, McGrath, & Kavanagh, 2014). Data from these types of studies are routinely gleaned from the personal ads of randomly selected users who have completed an online dating profile consisting of a mixture of both open- and closed-ended type questions. Despite the growing body of literature on this topic, there remains a dearth of research that analyzes whether skin color influences the dating selection process. The omission of skin tone as a relevant predictor variable may be due in large part to the fact that most dating websites do not solicit from their users this specific piece of descriptive information. Nevertheless, research has consistently shown that skin tone impacts social outcomes such as one’s perception of attractiveness (Hill, 2002; Jones, 2013; Stephens & Fernandez, 2012), racial identity formation and self-esteem (Coard, Breland, & Raskin, 2001; Cunningham, 1997), and within-group prejudice and discrimination (Chavez-Duenas, Adames, & Organista, 2014; Rondilla & Spickard, 2007). By not collecting and analyzing these types of data, questions surrounding the influence of skin tone on mate selection remain either understudied or not studied at all.
The current study uses data collected from a nationwide sample of Asian Americans (N = 2,024) who were seeking dating partners via Match.com in 2013 and 2016. Besides utilizing information obtained from their online personal ads, profile pictures were assessed in relation to 36 varying skin tones or hues via Von Luschan’s chromatic scale. This research is one of the first studies to examine whether skin tone has any measurable impact on racial/ethnic dating preferences among Asians. Moreover, the present analysis investigates whether skin tone moderates the relationship between sexual orientation and online dating preferences while controlling for relevant individual characteristics such as age, education, political orientation, and region of residence. Our results are discussed in relation to the racial hierarchy of preference and privilege that are inherently linked to colorism.
The skin tone continuum
Asian Americans represent one of the fastest growing racial/ethnic groups in the U.S. According to a recent PEW Research Center report, the Asian American population grew by 72% from 2000 to 2015, from 11.9 million to 20.4 million (López, Ruiz, & Patten, 2017). Besides the rich cultural diversity that exists among Asians, there are also visible skin tone differences, from very light hues of white (e.g., Japanese, Koreans, and Chinese) to darker shades of brown (e.g., Indians, Pakistanis, Bangladeshis, and Sri Lankans). Considerable skin tone differences are similarly visible within single countries. For instance, tan complexions are more common in North India, while deep brown skin tones are more prominent in South India. In China, residents of Harbin typically have pale skin tones in comparison with the tan complexions of Hainan. As with race, skin tone has played a critical role in shaping the lives of people of color, both in terms of how they view themselves and in terms of how others perceive and treat them (Jones, 2013; Rondilla & Spickard, 2007). Although the body of literature on this topic is robust for African Americans and Latino/Latinas (Gullickson, 2005; Kaufman & Wiese, 2012; Quiros & Dawson, 2013; Stephens & Fernandez, 2012; Wilder & Cain, 2011; Wilkinson & Earle, 2012), it remains underdeveloped for Asians Americans. There are currently only a handful of studies that have examined the history and influence that colorism—otherwise known as a form of prejudice or discrimination that is based on the reified social meanings attached to skin tone (Hunter, 2008)—has had within the Asian American community. For example, in one of the most comprehensive studies to date, Rondilla and Spickard (2007) draw from in-depth interviews and survey data to document how Asian Americans have adopted and promoted tenets of colorism, which has in turn led to such negative outcomes as the internalization of racism (e.g., bleaching their skin to meet the idealized white standard of beauty or desirability), or the perpetuation of prejudice and discrimination against individuals who fall on the darker end of the color continuum (e.g., using negative stereotypes such as “less attractive” to describe dark-skinned Asians).
The origin of colorism within Asian communities is somewhat different from that of other groups of color. Some scholars point to European colonization (Karnow, 1989; Rafael, 2000) as one of the roots of skin tone bias, while other academics highlight notions of class-based distinctions (Rondilla & Spikard, 2007). Regarding the former, lighter skin tones and Anglo facial features were often emphasized as the standard of desirability in Asian countries (e.g., India, Philippines, and Vietnam) that were colonized by the English, French, and Spanish (Hunter, 2008). Asians who were closer to this aesthetic ideal were able to secure resources and opportunities that were not readily available to people who were darker. The valorization of this colonial beauty standard served to perpetuate a system of oppression.
For other Asian groups (i.e., Asians who lived in countries that were not colonized), “Whiteness” was not the point of origin for colorism (Rondilla & Spickard, 2007). Rather, as some scholars have argued, colorism was a class-based notion where skin tone reflected one’s position or socioeconomic status (King-O’Rianin, 2006). In this context, “Whiteness” refers not only to the possession of a white skin tone but may also be defined as a social construct that reifies the racial hierarchy and racial biases. “Whiteness” represents a position of privilege and power and often interacts with other systems such as masculinity, cisnormativity, high socioeconomic status, and heteronormativity. For example, dark-skinned Asians were often viewed as manual laborers or peasants who were relegated to working outside. Their skin tone and class position were frequently associated with poverty and “backwardness” (Hunter, 2008; Rondilla & Spikard, 2007). On the other hand, Asians who possessed a lighter complexion were often perceived as members of the upper or elite classes. This color hierarchy which delineated class distinctions was also supported by various skin lightening rituals. For example, it was customary practice up to the mid-19th century for upper class Japanese men and women to use white-lead powder makeup to appear as white as possible (Glenn, 2009). This ritual seemingly reinforced the class imperative that being lighter meant having more wealth and resources. Although many Asians favor lighter skin, this sentiment, as some scholars have argued, should not necessarily be interpreted as Asians wanting to be white (Rondilla & Spikard, 2007). Regardless, Asians living in the U.S. are consistently exposed to a culture that elevates “Whiteness.” From the faces that regularly appear on television or in the movies, to the people who sit in positions of power, “Whiteness” is often treated as the norm by which other groups are consistently judged (Larson, 2005). As such, readers are advised to interpret the findings of the present study beyond the scope of visible skin tone and to consider the ways in which “Whiteness” and racial dating preferences may interact with other oppressive systems.
The impact of colorism on dating preferences
As noted earlier, in one of the only studies to have explored the broad topic of colorism among Asians, Rondilla and Spickard (2007) interviewed 99 Asian American men and women to inquire about various issues surrounding skin tone discrimination, including the importance of skin color when it comes to choosing a suitable mate. Many of their interviewees expressed a desire to date someone who was lighter than them. As stated by one of their respondents (a 26-year-old Japanese man): “My personal preference would tend to be on the lighter side. I wouldn’t ever be with a dark-skinned person…I guess that image of the princess with light skin” (Rondilla & Spickard, 2007, p. 61). In a related vein, some female respondents reported that their family discouraged them from dating individuals who had darker complexions. As relayed by a 23-year-old Chinese-Cambodian woman: “I used to go out with this dark guy and my dad said he was ugly. And that our kids would come out ugly” (Rondilla & Spickard, 2007, p. 51). Although their results highlight a bias toward lighter skin, the researchers also documented an interesting contradiction between their respondents’ beliefs and actions regarding skin tone preference. During the interview process, 20 male respondents were shown pictures of three Asian women with skin tones ranging from light to dark. The men were then asked which woman they would most like to date or marry. Respondents ranked the medium-toned woman as their first choice, followed by the dark-skinned woman, and the Asian woman with the lightest skin tone was surprisingly selected last. Although their research produced some interesting findings, Rondilla and Spickard (2007) did not formulate any grounded theory concerning the processes or dynamics surrounding colorism and dating preferences among Asians. Instead, however, their work serves as a springboard for future studies. To date, no subsequent research has continued this line of investigation. Moreover, there are still no studies available that examine whether an Asian’s skin tone impacts his or her dating preferences.
Despite this lack of research, relevant studies that focus on African Americans and Latino/Latinas may offer some valuable insights concerning the causal mechanisms surrounding the relationship between skin tone and dating beliefs. For instance, several existing studies have used some variations of social perception theory to help shed light on this topic (Stephens & Fernandez, 2012; Stephens & Thomas, 2012). This framework posits that individuals form impressions of and make inferences about other people by drawing upon information from influential sources such as family members, friends, and the media (McArthur & Baron, 1983). Moreover, when creating the generalized “other,” individuals also develop impressions about themselves that may impact their own attitudes and behaviors (Jacobs, 2006). For example, in a study that investigated the influence of skin tone on dating beliefs among 28 African American college women, Stephens and Thomas (2012) found that most respondents viewed darker skinned women in a more negative light compared to their fairer skin counterparts. These women also believed that African American men ascribed similar undesirable characteristics (e.g., unattractive, loud, and unintelligent) to dark women. In terms of their own impressions about themselves, approximately 90% of the sample indicated that if they were three shades lighter, African American men would be more interested in them. Conversely, they mentioned that if they were three shades darker, African American men would view them as being unattractive and would most likely avoid them. Consequently, this perception regarding African American men’s dating preferences may prompt more African American women to increase their pool of potential romantic partners by considering men from other racial and ethnic groups (Child, 2005; Yancey, 2002).
Like African Americans, Asians with darker complexions may also be more willing to cross the color line when selecting a mate. However, until tested, this prediction remains speculative in nature. As of now, the current body of literature on Asian dating preferences suggests that Asians (of all skin tones) are less willing to date other non-White groups compared to their African American and Latino/Latina counterparts (Tsunokai et al., 2014). For example, in a recent study that examined the online dating preferences of 1,270 Asian Americans, Tsunokai, McGrath, and Kavanagh (2014) found that heterosexual women and gay men were more likely to want to date a white person than were straight Asian males. On the other hand, these two groups were also less willing to state a preference to date African Americans and Latinos compared to the same reference group. Although their results document that Asian women and gay men, when viewed as a homogenous group in terms of skin color, may be displaying tenets of colorism by preferring groups with lighter skin, it is unclear whether the results would change when considering individual skin color variation. For instance, within the gay community, most mainstream gay publications have deemed “Whiteness” as being the norm; conversely, Asians and other non-White gay males are frequently marginalized by either being ignored or consistently characterized as undesirable dating partners (Han, 2007, 2008). Consequently, it should come as no surprise that Phua and Kaufman (2003) found that 31% of their gay male respondents indicated that they only wanted to date white men; approximately 8% stated a preference to date another Asian, and none of the Asian gay male respondents were willing to date another non-White minority. While their results highlight some of the consequences of colorism, it is problematic to assume that there is a universal impact that colorism has on Asian gay males, regardless of their own skin color. For instance, it is plausible that gay males with darker complexions may not necessarily subscribe to the same types of dating beliefs as their lighter skinned counterparts. As members of two marginalized groups, being Asian and having a dark complexion, these individuals may be more open to dating people of all types in order to maximize the size of their dating pool.
Based on our general discussion concerning colorism, we pose the hypothesis that compared to their lighter skinned counterparts, Asians with darker complexions will be more willing to state a preference to date members from all racial and ethnic groups, including other Asians. As an additional topic of interest, we explore the role of skin tone in the context of sexual orientation and willingness to date African American, Latino/Latina, and Asian. While Tsunokai et al. (2014) found existent differences in the racial dating preferences of Asian heterosexual females and gay Asian males, this research did not consider the potential function of skin color. Thus, we hope to extend this research by investigating the influence of skin tone on the dating selection processes of gay and lesbian Asians. We hypothesize that skin tone will have a pronounced effect on the dating preferences of Asians with stigmatized sexual orientations (i.e., gays and lesbians). More specifically, darker skinned gay and lesbian Asians—who may experience increased marginality through the tenets of colorism and homophobia—will exhibit a greater willingness to date both intra- and interracially when compared to their light-skinned Asian heterosexual male counterparts. As light-skinned Asian heterosexual males experience a more “elevated” place in society by holding a greater number of non-stigmatized identities (sexual orientation and gender), they may exhibit more selective dating patterns than gay and lesbian Asians. Using a sample of light-skinned Asian heterosexual males as our reference group (along with our inclusion of Asian heterosexual females), we are thus able to interpret our results as they relate to one’s skin tone, sexual orientation, and gender more critically. As mentioned earlier, unlike previous studies that have routinely treated Asians as being a homogenous group in terms of skin color, this research is one of the first of its kind to consider the role that skin tone plays on Asian online dating preferences.
Method
The data analyzed in this study were collected by students in two undergraduate capstone courses in 2013 and 2016. Approximately 25 students were extensively trained on how to collect data from the largest online dating website, Match.com. Information was gleaned from personal advertisements of Asian users who had completed a dating profile comprising a combination of both open- and closed-ended questions. This “electronic dossier,” which is free and available to all users who have Internet access, highlights descriptive information such as the dater’s physical features, personality traits, interests and hobbies, socioeconomic status, and the desired characteristics that the person is seeking in his or her dating partner(s). Concerning the latter, Match.com members are prompted to indicate (by checking the appropriate box or boxes) which racial and ethnic group(s) they prefer to date. Nine categories are provided to the user: Asian, Black/African descent, East Indian, Latino/Hispanic, Middle Eastern, Native American, Pacific Islander, white/Caucasian, and other.
To obtain a diverse sample of respondents (ages 18 and over), student coders targeted profiles from Asian men and women (both gay and straight) living within 50 miles of 16 major U.S. cities located in the West, Midwest, South, and Northeast. The cities within each of the four regions included: (1) West: Los Angeles, CA, San Francisco, CA, Seattle, WA, and Portland, OR; (2) Midwest: Chicago, IL, Columbus, OH, Minneapolis, MN, and Indianapolis, IN; (3) South: Atlanta, GA, New Orleans, LA, Charlotte, SC, Jacksonville, FL; and (4) Northeast: Boston, MA, New York City, NY, Philadelphia, PA, and Newark, NJ. Using systematic sampling (i.e., every seventh ad, with the first case being selected randomly), each coder sampled roughly 130 ads within their assigned city and sexual orientation. The population of Match.com users was also narrowed by excluding (via the “photo only” function) all profiles that did not include a photograph of the dater. Using a standardized coding sheet, coders recorded responses from eight closed-ended questions that measured relevant sociodemographic variables such as age, sexual orientation, political orientation, and education level. A total of 2,024 dating profiles were coded and entered into the Statistical Package for Social Sciences database. The effective sample size is based on the number of cases for which complete information was available. Table 1 presents summary statistics for variables included in the present study.
Descriptive statistics of Match.com respondents.
Note. N = 2,020 Asian Match.com dating profiles.
Variables and measures
The present study examines four dependent variables. The variables assess the degree to which Asian individuals were specifically willing to date whites, African Americans, Latino/Latinas, and Asian Americans. Each variable was dichotomized, assigning the value of “1” to respondents who indicated that they were willing to date individuals of a specific racial/ethnic background, and a value of “0” to those who were not.
Concerning the independent variable, capstone students used Felix von Luschan’s Chromatic Scale (1–36 different skin tones) to code the skin tone of Match.com daters. Profile pictures were assessed in relation to the 36 varying colors on the scale. When a profile had more than one color photograph, the coder selected the picture that best represented the skin tone for that particular user (e.g., natural lighting). An intercoder reliability analysis (Cohen’s κ) was performed on roughly 5% of the sample in order to determine how consistent the coders were in regard to rating skin tone. Although skin tone was originally measured at an interval level, for the present κ test, the 36 skin tones were collapsed into three ordinal categories: light skin tone (1–22), moderate skin tone (23–28), and dark skin tone (29–36). The statistical analysis produced a Cohen’s κ of .79. The numerical ranges correspond to Fitzpatrick’s Photo-typing Scale which consists of six categories of human skin pigmentation (Fitzpatrick, 1975). Based on his descriptions, some categories were combined (e.g., pale white, white, cream white), resulting in three skin color categories. Each skin tone number from the von Luschan scale was assigned into the appropriate skin tone grouping. Based on guidelines put forth by Altman (1999), and adapted from Landis and Koch (1977), a κ of .79 represents a substantial strength of agreement. It is important to note, however, for ease of statistical interpretation (especially with interaction effects), skin tone was left in its continuous form when performing regression analyses.
Relevant control variables were incorporated into the analysis to determine their net effect on the focal relationship. Sexual orientation was transformed into three dummy variables, “straight female,” “gay male,” and “lesbians,” with straight males serving as the reference group. Level of education was captured by five dummy variables, “some college” (which included respondents who had earned an associate degree), “BA or BS,” and “Graduate or professional” and “unknown level of education” (i.e., respondents who selected the response category, “I’ll tell you later”). Respondents who had only received a high school diploma served as the reference group. Political orientation was recoded from a seven-category variable (ultraconservative, conservative, moderate, liberal, very liberal, nonconformist, and some other viewpoint) into four dummy variables, “conservative,” “liberal,” “other” (the last two categories), and “unknown political orientation” (selected the category, “I’ll tell you later”), with moderates employed as the reference group. Religious affiliation was delimited into two separate variables. Daters who identified themselves as being Catholic, Protestant, Jewish, Buddhist/Taoist, Hindu, and Muslim/Islam were collapsed into the variable, “organized religion.” A second religion dummy variable was created which consisted of those Asian daters who did not report their religious orientation. Individuals, who identified as Agnostic, Atheist, and/or spiritual, served as the reference category (“no organized religion”). Finally, region of residence was introduced into the analysis as a series of dummy variables, “South,” “West,” “Northeast,” and “Midwest,” which served as the comparison category. Three interaction terms that measure the product of the respondent’s skin tone and sexual orientation (heterosexual female, gay male, and lesbian) are introduced to test our hypothesized conditional relationship (with Skin Tone × Heterosexual Males serving as the reference category).
Statistical estimation
Due to the binary nature of the outcome variable (willingness to date someone of a different race/ethnicity), logistic regression models were fitted to the data. For each covariate, the parameter estimate signifies the log of the odds of falling into category 1 of the response variable (coded as 1) versus not (coded as 0), adjusted for the effects of other variables in the model. For ease of interpretation, the likelihood of reporting racial or ethnic dating preferences is represented in the form of odds ratios calculated by exponentiating the logistic regression coefficients. Odds ratios greater than one are indicative of respondents having a greater likelihood of showing a racial/ethnic preference for potential partners compared to the reference group; conversely, an odds ratio less than one indicates a lower likelihood. Global model fit is assessed via the likelihood ratio statistic.
Results
We present findings by placing them within the context of the two hypotheses specified in the literature review. It was predicted in hypothesis 1 that compared to their lighter skinned counterparts, Asians with darker complexions would be more willing to state a preference to date members from all racial and ethnic groups, including other Asians. Relevant results testing the abovementioned hypothesis are shown in Table 2. The first model seeks to investigate the association of Asian skin tone and willingness to date African Americans. As may be seen in the table, the darker the skin tone the greater the log of the odds of willingness to date African Americans (Adjusted Odds Ratio (AOR) = 1.040, 95% CI = 1.017–1.063), a finding that provides evidence in support of hypothesis 1.
Multivariate logistic regression results of the association of skin tone on Asian willingness to date other racial/ethnic groups: Main effects only.
Note. AOR = Adjusted Odds Ratio
Next, results on Asian willingness to date Latinos/Latinas are shown in Model 2 (Table 2). As may be observed, findings on skin tone are remarkably similar to those noted with regard to willingness to date African Americans. A metric increment in skin tone increased the log-odds of willingness to date Latinos/Latinas by .033 adjusted for the effects of other variables in the equation.
The third model (Table 2) shows results of the association of Asian skin tone and Asian willingness to date members of their own race. Asian skin tone was found to be inconsequential with regard to Asian willingness to date members of their race. While hypothesis 1 was supported in data analysis with regard to African Americans and Latinos/Latina, no evidence was found to sustain the hypothesis for Asians.
Other notable findings in Table 2 are worth comment. The older the Asian sample member the less inclined their willingness to date Latino/Latinas, a finding consistent with that regarding African Americans and other Asians as well. Rather interesting findings emerged with reference to Sexual Identification. Heterosexual female Asians were less willing to date Latino/Latinas than heterosexual male Asians (AOR = 0.753, 95% CI = 0.570–0.995). The log-odds of willingness to date Latino/Latinas were more elevated among gay male Asians than among heterosexual male Asians (AOR = 1.427, 95% CI = 1.093–1.864). Likewise, the difference between lesbian Asians and heterosexual males (the reference group) was .640 in the log-odds.
The next set of findings are about the association of skin tone with Asian willingness to date various groups. The only difference between models presented in Table 3 and those in Table 2 is that statistical interactions (effect modifiers) are included in Table 3. These interaction terms test whether sexual orientation moderates (modifies) the relationship between skin tone and willingness to date African Americans and Latino/Latinas. Specifically, it was anticipated in hypothesis 2 that skin tone has a pronounced effect on the relationship between sexual orientation and willingness to date African American, Latino/Latina, and Asian. Relevant findings about hypothesis 2 are shown in Table 3.
Multivariate logistic regression results of the association of skin tone on Asian willingness to date other racial/ethnic groups: Main and interaction effects of sexual identification.
Note. AOR = Adjusted Odds Ratio
In Model 4 (willingness to date African Americans), no interaction term was significant, suggesting that the main effects only model was a better one than that containing interactions. Given the perils of interpreting nonsignificant results, we move next to Model 5, as hypothesis 2 received no support with regard to willingness to date African Americans. In Model 5 (willingness to date Latino/Latinas), there is evidence to suggest that the main effects model alone (presented earlier in Table 2) better fit the data than the one with interactions. Hypothesis 2 received partial support regarding Asian willingness to date Latino/Latinas as one interaction term reached statistical significance: Skin Tone × Heterosexual Female. In view of the positive sign, findings mean that the effect of every unit increment of skin tone is enhanced by .074 in the log-odds if the subject is a female. That is, compared to a man of average skin tone, women of lighter complexion were less likely to date Latino, but women of darker complexion were more likely to date Latino (Gordon, 2015; Menard, 2010; Osborne, 2015). In Model 6 on Table 3, the only skin tone by sexual orientation that reached statistical significance was that of Skin Tone × Gay. Given that the product term reached statistical significance, it warrants interpretation. As skin tone increases by one unit in its metric, the difference in the log-odds of willingness to date Asians is accentuated by .086 for gays relative to the reference group (heterosexual male). Asian willingness to date whites was also explored but due to space limitations, findings are available in the supplementary file.
Discussion
Asians represent one of the fastest growing and most diverse racial groups in the U.S. Besides their differing histories, cultures, and languages, they are also well represented on the skin color continuum, from very light hues of white to shades of dark brown. Like other racial minorities, Asian Americans are also susceptible to the long-standing ideology of colorism—a hegemonic notion that elevates “Whiteness,” which in turn can negatively impact how people of color form and maintain their social and personal relationships. Despite its influence, little is known regarding how one’s skin tone influences the dating selection process among Asians. Accordingly, the present study investigated whether skin tone had a measurable impact on inter and intraracial dating preferences. This research also assessed whether skin color magnified the relationship between sexual orientation and the willingness to date members from various racial/ethnic groups. The present examined two hypotheses. Based on our discussion concerning colorism, we first predicted that there would be a positive relationship between skin tone and the willingness to date African Americans, Latino/Latinas, and other Asians. Second, we also expected that darker skinned Asians (especially gay males and lesbians) would demonstrate different dating preferences compared to heterosexual Asian males who are lighter skinned. Results of the data analysis showed that skin tone was a significant predictor of Asians being willing to date African Americans and Latino/Latinas. In general, the darker the skin tone of participants in Match.com, the greater their willingness to date African Americans and Latino/Latinas; however, the results also documented a negative relationship concerning skin tone and willingness to date another Asian.
As one of the first studies to utilize skin tone as a predictor variable, our results challenge previous studies that have treated Asians as a homogenous group when investigating online dating preferences. In these types of studies (Robnett & Feliciano, 2011; Tsunokai, Kposowa, & Adams, 2009), Asians, like their white counterparts, were often less willing to state a preference to date Latino/Latinas and African Americans. For example, Robnett and Feliciano (2011) found that 94% of Asians did not express any desire to date someone who is African American. In another study, Tsunokai, Kposowa, and Adams (2009) documented that Asians and whites were 31% and 24% less likely to state a preference to date an African American compared to their Latino/Latina counterparts, respectively. Such findings have added to an ongoing debate as to whether Asians—a group that Bonilla-Silva (2013) has called “honorary whites”—have adopted similar patterns of racial exclusion to that of whites. By utilizing skin tone, however, we are now in a better position to develop a more complete understanding about the differences that exist among Asians when it comes to their dating preferences. Contrary to past studies, the present analysis documented that darker skinned Asians expressed a greater willingness to date African Americans and Latino/Latinas compared to their lighter counterparts. A plausible explanation for this finding is that Asians with relatively dark skin may experience some form of prejudice due to their own color and are thus more understanding of the plight of out groups that have the same or darker skin tones. It is possible that some Asians actively resist the internalization of negative stereotypes associated with colorism by demonstrating a greater willingness to date people of all races and colors; they may not want to perpetuate the same types of prejudice and discrimination that are often leveled against them due to their darker complexion.
There also exists the possibility that some dark-skinned Asians are more amenable to dating African Americans and Latino/Latinas since they often have diminished options to date whites and light-skinned Asians. Past studies have consistently documented that Asian men are the least desired and most excluded when it comes to dating preferences (Feliciano, Robnett, & Komaie, 2009; Fisman, Iyengar, Kamenica, & Simonson, 2008). For example, a speed dating experiment found that members from other races (white women in particular) expressed the strongest preferences against Asian males due in large part to their perceived unattractiveness (Fisman et al., 2008). Robnett and Feliciano (2011) also reported that Asian women were more likely to exclude Asian men (40%) over white men (11%) as potential dates. This rejection may be due in some part to the perpetual demasculization of Asian men by the media. Asian men are routinely depicted as asexual or effeminate; they are essentially stripped away of any sexual desirability (Eng, 2001). Perhaps due in part to their effeminate portrayals by the media, Chan, Wilkins, and Kaiser (2011) found that Asian men who looked more “prototypically Asian” were perceived as less masculine and less attractive by white participants when compared to Asian men who did not possess as many stereotypically Asian physical traits, such as dark hair or full cheeks. These unflattering characterizations, in combination with the stereotypes surrounding darker skin tones, may prompt some Asians to select members from other non-White groups who have similar or darker complexions, with the belief that these individuals may be less inclined to reject them as potential dating partners. It has been documented that members from marginalized communities often seek out one another, expressing greater positivity and perceived similarity with each other (Craig & Richeson, 2012).
Although Asians with darker complexions expressed a willingness to date African Americans and Latino/Latinas, our analysis produced an unanticipated result regarding the impact of skin tone on one’s willingness to date another Asian. In our sample, darker skinned Asians were less willing to state a preference to date someone of the same race. Although we had anticipated that dark-skinned Asians would be more amenable to date members from all groups, it is plausible that some Asians with darker complexions may be hesitant to romantically reach out to other Asians in fear that they will simply be ignored or rejected. For example, Rondilla and Spickard (2007) found that although Asian men preferred light skin in the abstract, they picked women who had medium to dark complexions when asked who they viewed as a more serious candidate for a life partner.
When it comes to dating desirability within the gay community, gay Asians, like their male heterosexual counterparts, are often sexually marginalized by the media; they are routinely depicted as perpetual foreigners who are nonmuscular, passive, and sexually submissive—qualities are that are not desired within the gay community (Han, 2007, 2008). On the opposite end of the spectrum, young and attractive white males, whose faces regularly adorn gay magazine covers and Internet sites, are often seen as the desired standard or norm. Besides being gay, Asian Americans in LGBT communities are often subjected to additional forms of marginalization that are rooted in race and skin color. Consequently, this multiple marginality may have a negative impact on their dating preferences or perceived desirability. For example, past studies have continually documented a heightened degree of resistance among gay males, including other Asians, to date an Asian (Tsunokai et al., 2014). Han (2008) documented that many non-Asians (whites in particular) have openly expressed their aversion toward Asians as potential romantic partners using such phrases as, “no fats, femmes, or Asians” as tag lines in gay personal classifieds. As one of the first studies to examine whether skin tone has a pronounced effect on the relationship between sexual orientation and dating preferences, the present analysis revealed that gay Asian males with darker complexions were more likely to state a willingness to date another Asian. This result may be driven in part by the smaller pool of potential daters that are available to Asian men who are darker. At the same time, the willingness to date a fellow Asian may be linked to a conscious effort by dark-skinned Asians to fight against the hegemonic notions of colorism. However, it is important to note that skin tone did not magnify the effect of sexual orientation and one’s willingness to date another Asian among Asian lesbians. In other words, Asian lesbians with darker complexions, who are similarly subjected to multiple forms of marginalization based on sexual orientation and skin tone, were not more likely to state a willingness to date another Asian. This difference between gay Asian males and Asian lesbians may be attributed to gender. While Asian men are commonly viewed as unattractive dating partners (Feliciano et al., 2009; Fisman et al., 2008), Asian women may be perceived as more desirable due to their feminine physical traits (Stephen, Salter, Tan, Tan, & Stevenson, 2018). Regardless of one’s own complexion, Asian lesbians may recognize their somewhat elevated position in the dating market and feel that they are able to demonstrate a higher level of selectiveness when seeking potential female partners. However, as studies such as Stephen, Salter, Tan, Tan, and Stevenson (2018) do not consider the role of skin tone on perceptions of desirability, this is merely speculation. Further investigation into the topic of skin tone preferences among Asian lesbians of darker complexions is needed before drawing definitive conclusions.
A similar dating pattern was found regarding straight Asian women and their preference to date Latino men; the present results revealed that Asian women with darker skin were still more unlikely to date Latino/Latina compared to Asian men with darker skin. The logistic regression coefficient of the interaction term was .074. This means that the effect of every unit increment of skin tone is enhanced by .074 in the log-odds if the subject is a female. That is, compared to a man of average skin tone, women of lighter complexion were less likely to date Latino, but women of darker complexion were more likely to date Latino. This finding contradicts previous studies that document Asian women as being less willing to date Latinos and African Americans compared to their male counterparts (Tsunokai et al., 2014). Although Asian women are often sought out by men of all races, darker skinned Asian women, as discussed, may be viewed as less attractive or desirable by men (Rondilla & Spikard, 2007). Since colorism within the Latino/Latina community is also pervasive, darker skinned Asian women may be more willing to reach out to individuals who are also perceived to be devalued or negatively stereotyped based on their skin tone.
Although this study provides unique insights about the impact of skin tone on dating preferences among Asian Americans, there are some limitations that need to be acknowledged. First, the data utilized do not constitute a nationally representative sample of individuals seeking romantic partnerships. Our data exclusively targeted members from one online dating community, Match.com. Although this website has consistently been identified as the largest online dating site, it still competes against hundreds of other dating websites and apps. Consequently, our sample may not be representative of online daters across other sites. Moreover, given that Match.com is an online dating platform, the present findings may not accurately reflect the attitudes and behaviors of those who do not rely on dating websites when seeking romantic partners.
In addition to the limitations associated with generalizability, sensitive topics surrounding race relations, especially interracial romantic relationships, may lead respondents to provide socially desirable answers to such questions as, “Are you willing to date someone from a different racial or ethnic group?” While there is the possibility that individual dating preferences may not mirror their actual dating behaviors, it is important to note that the data garnered from Match.com are not designed for research, but rather to help daters find a suitable partner. Users may be more willing to express their true preferences within their profiles to facilitate the process of finding a perfect match. Yancey (2009) has noted that most users of online websites do not know that they are being studied, thus the likelihood of potential social desirability effects tends to be lessened. Nevertheless, readers should interpret the present findings judiciously and recognize that the findings are based on intentions, not actual behavior.
Another limitation is that there are no questions on Match.com that directly evaluate respondents’ attitudes vis-à-vis skin tone; it is impossible to know whether respondents subscribe to positive or negative stereotypes surrounding lighter or darker complexions. Consequently, the causal mechanisms outlined in the present study—specifically, the impact of skin color—cannot be empirically measured, but rather, serve to situate the findings and provide a narrative that may help us to better understand dating preferences among Asian Americans.
Despite these potential limitations, our results suggest that skin tone has an impact on who Asians prefer to date. Future research should examine the social psychological factors and social processes that colorism may play when seeking potential romantic partnerships. Ultimately, further research on this topic will shed light on the shifting racial hierarchy and the influences of colorism on individuals’ racial dating preferences.
Supplemental material
Supplementary_Materials_2019 - The color continuum: Skin tone and online dating preferences among Asian Americans
Supplementary_Materials_2019 for The color continuum: Skin tone and online dating preferences among Asian Americans by Glenn T. Tsunokai, Augustine J. Kposowa, Ellen Carroll and Miriam Karamoko in Journal of Social and Personal Relationships
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Open research statement
As part of IARR’s encouragement of open research practices, the author(s) have provided the following information: This research was not pre-registered. The data used in the research are available. Information was gleaned from personal advertisements of Asian users who had completed a dating profile comprising a combination of both open and closed-ended questions. This “electronic dossier,” which is free and available to all users who have Internet access, highlights descriptive information such as the dater’s physical features, personality traits, interests and hobbies, socioeconomic status, as well as the desired characteristics that the person is seeking in his or her dating partner(s). To obtain a diverse sample of respondents (ages 18 and over), 25 trained coders targeted profiles from Asian men and women (both gay and straight) living within 50 miles of 16 major U.S. cities located in the West, Midwest, South, and Northeast in 2013 and 2016. Using systematic sampling (i.e., every seventh ad, with the first case being selected randomly), each coder sampled roughly 130 ads within their assigned city and sexual orientation. The population of 
users were also narrowed by excluding (via the “photo only” function) all profiles that did not include a photograph of the dater. A total of 2,024 dating profiles were coded and entered into the Statistical Package for Social Sciences database. The data can be obtained at: Match.com or by emailing:
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
