The Unfair Selection: A Study on Skin-Color Bias in Arranged Indian Marriages

Underlying the growing popularity of skin-lightening or fairness cosmetics in India is one of the most baseless biases experienced and practiced. Yet, the overriding importance of skin-color especially in context of marriage has been largely unaddressed. This exploratory study examined the influence of skin-color on preference for potential marriage partner. A 2 × 2 (gender × skin-color) between-group experimental design was used. Mothers (N = 108) of individuals of marriageable age group were presented with an option of five marital profiles containing education and work information only. The participants were shown profiles of either males or females depending on whether they had a son or a daughter. Once a profile was chosen, the participant was either shown a photograph of highly attractive fair girl/boy or a highly attractive dark girl/boy. The light-skinned and dark-skinned photograph was of the same person, except their skin tones were manipulated with the use of computer software. Participants were asked to rate how strongly would they recommend the girl/boy as potential bride/groom for their children. As expected, fair-skinned highly attractive people received higher ratings than dark-skinned highly attractive people. However, contrary to our expectations, ratings received for dark-skinned woman were not significantly lower than the ratings received for dark-skinned man. This study shows that the color of skin has the potential to even overpower traits such as general competency and physical attractiveness in both men and women.


Introduction
The fairness cosmetics industry is a huge business in India, worth almost US$180 million, and growing at a rate of 10% to 15% annually (Shevde, 2008). A typical advertisement for a fairness cream in India begins with portrayal of the (many) problems of a dark-skinned person. Majority of skin-lightening cosmetics target women. They are likely to portray a darkskinned woman as an underconfident and insecure person who is unable to secure success in life until someone suggests the use of a fairness cream (Sylvia, 2014). The advertisement then shows her radical transformation; not only does she become several shades lighter, but her entire personality undergoes a drastic change. She emerges as a successful woman, confident, and self-assured because of her "whiteness." The advertisement ends with her clinching a desirable marriage proposal; the camera zooms in on her "fair" and happy face. The implied message is "fair is lovely and dark is ugly." An industry that began selling fairness creams exclusively for women now offers whiteness to Indian men too. Fair & Lovely cream has found its counterpart in Fair & Handsome cream for men. Today, the market offers fairness creams, face washes, and makeup for both genders. Most advertisements carry an implicit message that darkness is the root of all women's (and now men's) problems. Consequently, skin-lightening is the key to our success in all spheres of life. One can find means to lighten each and every part of the body. The latest addition to the list is whitening creams for one's private areas. Cosmetics industry sells fairness along with promises of popularity, success, and happiness to many.
Underlying the growing popularity of skin-lightening or fairness cosmetics in India is one of the most baseless biases experienced and practiced. Colorism or skin-color stratification is a process that privileges light-skinned people over dark in areas such as income, education, housing, and the marriage market (Hunter, 2007). In context of race, studies have found that light-skinned people are financially better off, more educated, live in better neighborhoods, and marry higher status people than darker-skinned people (Murguia & Telles, 1996;Rondilla & Spickard, 2007). Elevated rates of blood pressure are also well documented in Blacks (see Brondolo, Rieppi, Kelly, & Gerin, 2003, for a review). Although studies examining colorism have restricted their research to African American individuals, there is growing evidence for an overwhelming preference for fair skin over dark-skin in Asian countries. "Whiteness" or having white skin is considered an important element in constructing female beauty in Asian cultures (Baumann, 2008;Krishen, LaTour, & Alishah, 2014;Li, Min, Belk, Kimura, & Bahl, 2008). According to Glenn (2008), skin-lightening products are common across many Asian cultures. In India, a woman with dark complexion is likely to be considered unattractive from the point of view of cultural constructions of beauty ideals (Arif, 2004;Parameswaran, 2015). A focused group discussion with Indian women found that obsession with fairness and the related attributes of beauty, attractiveness, and marital success continues to prevail (Karan, 2008). This is not surprising as fair skin has a colonial connotation of power and superiority (Shankar, Giri, & Palaian, 2006). Badruddoja (2005) also reports that for many Hindu-Indian women, attractiveness and beauty are linked to their color of skin. Nagar and colleagues (2017) found that exposure to fair and attractive media images led to activation of appearancerelated schemas in young Indian women. Swami, Furnham, and Joshi (2008) also found that women with lighter skincolor were rated to be more attractive, healthy, and fertile. Furthermore, fairer skin-color is related to better prospects of finding a mate in arranged marriage context (Badruddoja, 2005;Jha & Adelman, 2009).
While Black scholars in the Unites States have thoroughly examined the link between racism and colorism, there is paucity of information tracing the historical roots of skin-color discrimination in India (Parameswaran & Cardoza, 2009a). Internalization of superiority of fair/white skin has been related to the combined influences of colonialism, caste system, and globalization. Many South-Asian countries like India, Pakistan, Sri Lanka, Nepal, and so on were ruled by the British for around 200 years; "white" race was the ruler and the "dark" native were the ruled. This led to internalization of superiority and power of the "white" skin and inferiority and powerlessness of the "dark skin" (Speight, 2007). Internalized racism reveals itself in a variety of situations from work environment to social situations where people of color reject or denigrate those with dark-skin. The caste system in India is likely to have given impetus to the notion of superiority of fair skin-color brought by colonial rule (Parameswaran & Cardoza, 2009b;Shankar & Subish, 2016). Higher castes have been perceived to be "fairer" and superior while lower castes have been perceived to be "darker" and inferior. Today, in postcolonial world, globalization has led to increased spread and acceptance of Western beauty ideals in Asian and African cultures (Hunter, 2011;Peltzer, Pengpid, & James, 2016). An increasing number of young Indian women are routinely exposed to Indian and Western media, which overwhelmingly portray Western beauty ideals. Rampant capitalism has made whiteness an international beauty ideal and market commodity (Goon & Craven, 2003). Skin-color remains a predominant criterion along which many individuals are evaluated in our society, especially in context of mate selection (Haq, 2013). Overall, combined influences of the caste system, colonialism, and globalization have made fair/light skin-color a social capital that enables upward social mobility in India (Parameswaran & Cardoza, 2009b).
As opposed to the individualistic Western society, where mate selection is the prerogative of the individual in question, in collectivistic cultures, arranged marriages refer to selection of mate accomplished by family rather than the individual (Chang & Myers, 1997;Gupta, 1976). While "romantic love" is likely to be criteria and legitimate basis for marriage in individualistic cultures, collectivistic cultures often use a number of criteria including subsistence skills, family, alliances, economic arrangements between families, and health (Udry, 1974). Matrimonial advertisements are found to request correspondence for arranged marriages based on age, religion, caste, regional ancestry, education, professional qualifications, income, and skin tone (Gist, 1953;Haq, 2013). As is the case in other cultures, compared with women, men place more importance on physical attractiveness and women place more importance than men do on the earning capacity of a potential mate (Wiederman & Allgeier, 1992). In a study, content analysis of matrimonial advertisements in India affirms the same, that is, social exchange of men's financial stability for women's physical attractiveness is common (Ramasubramanian & Jain, 2009). Deeply embedded cultural norms prescribe light/fair as an essential feature of women's beauty in India and other Asian countries (Gist, 1953;Goon & Craven, 2003;Leong, 2006;Parameswaran & Cardoza, 2009b). Even in context of marriage, a preference for fair-colored women is omnipresent across matrimonial advertisements in newspapers and websites (Pandey, 2004;Ramasubramanian & Jain, 2009). "Looking for a slim, homely, and 'fair' girl for our son" is a common feature in matrimonial advertisements.
Although dark-skin is perceived to be disadvantageous for both men and women, studies have reported that Black women seem to suffer from "double jeopardy" (Beal, 2008). That is, women are likely to be at a more disadvantageous position than men where color of skin is concerned. In context of work settings, dark-skinned women receive more rejections than equally qualified dark-skinned men (Berdahl & Moore, 2006;Harrison & Thomas, 2009). In context of matrimony, although dark-skin color proves to be disadvantageous for both men and women, women are likely to be discriminated more than men (Gist, 1953). Researchers have found that in looking for mates in arranged marriages, men were more likely than women to state a preference for skincolor (Jha & Adelman, 2009;Ramasubramanian & Jain, 2009). Utley and Darity (2016) reported that prospective brides are likely to describe themselves in terms of color of skin (fair) than prospective grooms.
South-Asian immigrants in the West are also found to desire lighter skin and have lower body satisfaction compared with their White counterparts (Sahay & Piran, 1997). Grewal (2008) found that non-White immigrants in West experience feelings of discrimination based on their skin-color. Experience of rejection or discrimination on the basis of color of skin is likely to have implications for socio/psychological well-being of individuals, specifically body image, selfesteem, and dissatisfaction with self and others. Okazawa-Ray, Robinson, and Ward (1987) found a significant positive correlation between self-esteem and liking of own skin-color; greater liking for skin-color was associated with higher score on self-esteem measure for African American adolescents. Dark-skinned women may suffer from low self-esteem as they have to cope up with a number of negative messages from the larger society, such as the perception that "fair is lovely and dark is ugly" (Jain, 2005). Studies on women have replicated the findings and suggest dark-skin-color is related to greater negative ratings of overall appearance and facial satisfaction (Bond & Cash, 1992) as well as symptoms of depression and poor self-esteem among African American adolescents (Williams, 2006).

Present Study
This unfair importance of "fairness" in determining life outcomes widely studied for the Black culture (Johnson, 2002) remains an unaddressed discourse in India. Although experience of bias and discrimination is a normalized everyday experience in lives of many Indian men and women, empirical research is scarce. The primary purpose of the study was to examine the presence of preference of fair skin over darkskin color in context of matrimony. As parents play an important role in mate selection in arranged marriages in India, the participants of the present study were women with children of marriageable age. It was hypothesized as follows: Hypothesis 1: Photograph of dark-skinned individual will receive lower ratings than equally attractive and qualified fair-skinned individual.
Hypothesis 2: Photograph of dark-skinned woman will receive lower ratings than dark-skinned men.

Participants
Urban mothers (N = 108) of individuals of marriageable age group (24-30 years) from New Delhi and NCR were selected for the present study. The demographic details of the participants have been provided in Table 1. As can be seen, the two treatment conditions (fair and dark) were matched on the following self-reported variables: age of mothers, education, and socioeconomic status (SES).

Procedure
A 2 × 2 (gender × skin-color) between-group experimental design was used. The participants were presented with an option of five marital profiles containing education and work information. The participants were shown profile of either males or females depending on whether they had a son or a daughter. Once a profile was chosen, photograph of highly attractive fair girl/boy was shown or highly attractive dark girl/boy was shown. The light-skinned and dark-skinned photograph was of the same person, except their skin tones were manipulated with the use of computer software. Participants were asked (a) to rate the competency of the profile selected by them on a rating scale of 1 to 7 (highly incompetent to highly competent) and (b) to rate how strongly they would recommend the girl/boy as potential bride/groom for their children on a rating scale of 1 to 7 (strongly NOT recommend to strongly recommend). Informed consent was taken from the participants and confidentiality was ensured to all. At the end of the study, all participants were debriefed regarding the true nature of the study.
A pilot study was conducted previously for the profiles and photographs used in the study proper. First, for the photographs, 12 passport size photographs, each of males and females judged to be physically attractive, were selected. Ten mothers of children of marriageable age were asked to rate the perceived attractiveness of the photographs using a 7-point Likert-type rating scale ranging from very Note. Coding for educational qualifications: 1 = graduation, 2 = post-graduation, 3 = higher studies; SES: 1 = lower middle, 2 = middle, 3 = upper middle, 4 = upper class. SES = socioeconomic status.
unattractive (1) to very attractive (7). Top-rated photograph for attractiveness was selected for both the genders to control for the variable physical attractiveness. The male and female photographs were then turned highly fair and highly dark to assess their suitability for the two skin-color experimental conditions. This was done with the help of computer software. The photographs were rated by the same sample for the perceived skin-color from very dark (1) to very fair (7). This prerating was done to ensure that the highly attractive photographs selected were also perceived as highly fair/dark. The same pilot sample (mothers) as the one used for the selection of photographs also independently rated the competency levels of eight profiles for males and females. They rated the profiles for adequacy of the educational qualifications, work experience, and overall competency of the applicant. Apart from these, three HR professionals also rated the profiles for their competency. Pilot study results found that the profiles received high ratings for competency by mothers and HR professionals. This was done to ensure that the profiles used in the primary study were perceived as highly competent.

Results
A 2 (gender) × 2 (skin-color) between-group ANOVA was computed to assess the likelihood for recommendation of the girl/boy as potential bride/groom for their children. Competency ratings for profile selected by participants have been presented in Table 2. As can be seen, the profiles received high ratings for competency from the participants in all conditions. There were no significant differences between the competency ratings attributed to the selected profile between the fair and dark treatment conditions. This was true for both the genders. Table 3 presents the descriptive statistics for ANOVA for the dependent variable recommendation ratings. It was expected that fair-skin photograph would receive significantly higher ratings for recommendation (i.e., recommendation for suitability as life partners for their sons/daughters) relative to dark skin-color photograph (Hypothesis 1). Figure 1 presents recommendation ratings for fair and dark-skin color photographs.
Consistent with our hypothesis, the ANOVA results (Table  4) for recommendation yielded a significant main effect of skin-color with a large effect size, F(1, 104) = 32.21, p = .000, η 2 = .236. There was no gender effect; that is, there was no significant difference in the ratings given to the two photographs. It was also expected that the photograph of darkskinned woman would receive lower ratings than the photograph of man with dark-colored skin (Hypothesis 2). Contrary to our hypothesis, skin-color by time was not significant, F(1, 108) = 0.019, p = .890, η 2 = .00, implying that the pattern of ratings did not differ according to whether the participants were males or females. That is, dark photograph of the highly attractive and competent woman did not receive significantly lower ratings than the photograph of a highly attractive and competent dark man.

Discussion
This exploratory study sought to shed light on skin-color bias in context of arranged Indian marriages. Results of the study indicate that there appears to be a skin-color preference with regard to making decisions about the suitability of an individual as a potential partner. Given that the study was manipulated in such a way that the influence of the two other important variables, namely, attractiveness and competency, was controlled in a manner to minimize their influence, the significant mean differences in ratings can only be attributed to the skin-color variation. Individuals perceived as highly competent but dark-skinned were rated lower than their highly competent light-skinned counterparts. The findings of the study establish the presence of colorism in context of matrimony. Sadly, these findings are not terribly shocking but are in fact in line with our expectations.
White and black are not dichotomous categories of skincolor in India. To the West, Indians are brown skinned. However, Indians find their skin-color to be a unique dimension of social comparison, a scale on which they can rate themselves as relatively fairer/darker than others. Malhotra (2004) called this as a "whiteness spectrum," defined as a social ladder with "pure" whiteness on the top exercising control and superiority over those lower on the spectrum. Thus, skin-color is related to social hierarchy in India; fair skin is often considered to be a mark of higher social standing. However, it is important to note that historically and culturally, dark not white skin was considered to be ideal and desirable in India. Some notable examples are the popularity of God Krishna (literally black) and Draupadi (also called  Krishnaa), a character from the epic Mahabharata. Krishna is worshipped in many parts of India whereas Draupadi was considered to be one of the most desirable women in the world. The transformation of ideal skin-color from dark to fair can be traced to the influence of caste system, British imperialism, and global hegemony of whiteness. The caste system also called varna (literally color) accounts for the perceived superiority of fair skin over dark. Owing to the association of fairer skin with upper caste and darker skin with lower castes, skin-color came to signify the social position of an individual in our society. In addition, the racist construction of "dark native" by the British seems to have become a part of our unconscious and is often projected as strong dislike for the "dark other" (Parameswaran & Cardoza, 2009b). Moreover, the global hegemony of whiteness has also led to internalization of the values of the White privileged class in non-Whites. A qualitative study by Bakhshi and Baker (2011) found that Indian men and women living in the United Kingdom report that relative to men, women were more likely to face pressures and criticisms related to their appearance. Traditional Indian beliefs about feminine beauty including fair skin and slimness were perceived to be promoted by media and maternal influence and linked to finding a suitable partner in arranged Indian marriage. Women in particular indicated how fair skin-color was associated with beauty and found that that internalization of Indian body ideals (including fair skin and slimness) was linked to finding a suitable life partner for women, and this pressure was largely perceived to be influenced by maternal encouragement and the media. Whiteness is communicated through mass advertising, movies, TV shows, Internet images, magazines, and celebrity culture (Baumann, 2008). For instance, even though MTV aimed to "Indianise" its programming after it was introduced in 1994, Cullity (2002) found that the channel promoted many Western values, notably the beauty ideals of "whiteness" and "thinness." The fairness cosmetics industry is guilty of cashing in on the life-long internalization of whiteideal and rejection of the dark-other. In addition, it is also responsible for the construction of the concept of skin-color as modifiable. The advertisements of skin-lightening cosmetics perpetuate the belief that fair-skin, a mark of higher social standing in India, is indeed attainable. In most cases, they promise to protect against the harmful and darkening ultraviolet (UV) rays of the sun and reduce the melanin secretion, thus leading to the emergence of fairer skin and more "beautiful" skin. Thus, merging of new technologies with old colonial and caste ideologies has created a context in which "whiteness" can be sold and bought (Ashikari, 2005;Hunter, 2007).
Turning to our second hypothesis, even though we found that fair skin was preferred over dark skin, we did not find evidence for "double jeopardy," that is, dark-skinned woman was not found to be in a more disadvantaged position than dark-skinned man. This is a surprising finding as studies have found that women are judged more harshly for darker color of  skin than are men (Harrison & Thomas, 2009;Jha & Adelman, 2009;Utley & Darity, 2016). The concept of "double jeopardy" has been criticized as overly simplistic as it does not take into account several and simultaneous oppressions that women of color face (King, 1988). For instance, personal lives of Indian women are impacted by a number of factors including caste, class, ethnicity, religion, and skin-color (Haq, 2013). Women's lives are not a simple algorithm where colorism and gender are additive in their impact. The complex lived experiences of Indian women can be better understood through the concept of intersectionality. Intersectionality is defined as "the interaction between gender, race, and other categories of difference in individual lives, social practices, institutional arrangements, and cultural ideologies and the outcomes of these interactions in terms of power" (Davis, 2008, p. 68). The present study aimed to examine the possible (disadvantageous) effect of gender and skin-color on preference in context of arranged marriages while controlling for the effect of perceived attractiveness and competencies. We presented participants with profiles of women judged to be highly competent with respect to their educational qualifications, work experience, and overall competency. The participants believed that the woman's photograph that they were viewing was a highly competent lawyer/doctor/engineer. Also, the photograph selected received highest ratings for attractiveness in our pilot study. It is possible that while gender and skin-color may create disadvantage conditions for a woman, perceived attractiveness and competency are likely to work in their favor. In support, a study by Thompson and Keith (2001) found that the impact for skin-color was weaker for women from higher social class; physical attractiveness and SES moderate the relationship between skin tone and self-esteem. Literature is scarce on whether the perception of physical attractiveness and competency is likely to reduce skin-color biases. The popularity of some dark-skin colored Bollywood actresses like Bipasha Basu, Kajol, Deepika Padukone, and so on suggests that masses are likely to accept a dark-skinned woman if she is perceived as highly attractive. Therefore, highly attractive woman with superior educational and professional qualifications may not suffer from "double jeopardy." The interactive oppressions that characterize the lives of Indian women form a case for "multiple jeopardy" (King, 1988). Future studies can examine the interactional effects of skin-color, gender, attractiveness, and competencies in selection of mates in arranged marriages. Overall, the present study is an exploratory research and the findings of the study must be examined in light of the following limitations. First, we conducted this study on urban middle/upper-middle-class mothers. High professional competency and educational qualifications in a woman may offset the potential skin-color bias. It would be interesting to replicate this study in societies where educational and professional competency of women matter considerably lesser in context of arranged marriages. Second, the setting of the study was north India (Delhi NCR) where people generally have fairer skin compared with south India, thereby limiting the generalizability of the findings. Third, the participants of the study were mothers who generally play an important role in selection of bride/groom in arranged marriages. It would be interesting to see if their children also exhibit the same bias as they do. Third, perception of one's own skin-color can also play an important role since. However, it was difficult to gather this information from the participants without revealing the purpose of the study. Future researches should also focus on investigating the possible mediating effects of varying levels of competency and attractiveness on skincolor bias across a diverse sample of participants.
Although there is plenty of work on colorism in the United States, there is very little work on skin-color discrimination in India. Marginalization of the dark skin is a phenomenon rampant in our society, but scholars have failed to acknowledge and address this. While content analyses of matrimonial advertisements have indicated a general preference for fair skin, empirical evidence is scarce. The findings of the exploratory study suggest that skin-color may play an important role in mate selection in arranged Indian marriages. Recognizing this bias is the first step toward making amends.