social sciences Mixed-Race Ancestry (cid:54) = Multiracial Identiﬁcation: The Role Racial Discrimination, Linked Fate, and Skin Tone Have on the Racial Identiﬁcation of People with Mixed-Race Ancestry

: Mixed-race identiﬁcation may be complex, in that people with mixed-race ancestry may or may not identify as multiracial. Social experiences, such as experiencing racial discrimination, believing that your fate is connected with speciﬁc racialized others, and personal characteristics, such as skin color, all have been theorized to play a role in identiﬁcation. The Mixed-Race Ancestry Survey (2019) conducted on Mechanical Turk allows me to ask unique questions with a large enough sample of this understudied population to disaggregate by racial ancestries. Only people with mixed-race ancestry are included in this study, but respondents may identify mono- or multiracially. Binary logistic regression models reveal that increased linked fate with a speciﬁc racial group is associated with greater odds of racially identifying, at least in part, with that group (e.g., among Asians, greater linked fate with Asians is associated with greater odds of identifying as mono- or multiracially Asian). Increased linked fate with multiracials as a group is also connected to greater odds of identifying as multiracial. In addition, personally experiencing racial discrimination is associated with a greater likelihood of identifying as Black and slightly lower odds of identifying as White or as Latinx. Finally, as skin tone darkens the odds of identifying as Black increase three-fold and the odds of identifying as multiracial increase by 1.3 times. I discuss these ﬁndings by racial ancestry groups, noting that being aware of having mixed-race ancestry does not in and of itself predict multiracial identiﬁcation. Rather, in a social structure that uplifts Whiteness, feeling linked fate, experiencing discrimination, and having darker skin tone are important predictors of identiﬁcation. These ﬁndings highlight the mechanisms connected to racial identiﬁcation for a population that may feel tied to multiple racial groups and is navigating identiﬁcation within a White-centric nation. The present study focuses on one point in time, but recognizes the importance of and need for future longitudinal work.


Introduction
As a mix of multiple races, multiracials blur the boundaries imposed by race (Lee and Bean 2004;Miyawaki 2015;Morning 2000). What constitutes a group, who gains access, and who becomes a member may fluctuate for multiracials more than monoracials. Multiracials may have weaker boundaries compared to monoracial peers, in that multiracials may be more able to transgress spaces and access various racial groups . For example, twenty years into the twenty-first century, as a nation the U.S. is having conversations regarding the fluidity of racial identification "choices" 1 made by high profile figures such as former President Barack Obama (Masuoka 2017) and Vice President Kamala Harris, and the impact of which race(s) they mark on a form. The multiracial movement pushed for the U.S. Census to allow respondents to choose more than one racial category; the movement raised public awareness, discussions, and conceptualization of multiracials as a unique group. Multiracial identification is now available on surveys, and this opportunity to "choose one or more" racial category is being utilized at increasing rates (Jones and Bullock 2012).
Historically, people with one Black parent and one White parent were classified in the U.S. as "Black" (Davis 1991), but in contemporary society may assert a "biracial" identity,

Variations among Biracials by Racial Composition
Given the divergent histories of racial groups in the U.S., researchers must consider how racial identity varies by component group (e.g., a Black/White individual compared to an American Indian/White individual). Scholars have advocated for disaggregating biracials, noting that subgroups of biracials have different processes when choosing to identify as biracial or monoracial (Gullickson and Morning 2011;Liebler 2016;Liebler and Halpern-Manners 2008). For example, Black/White biracials are most likely to identify as Black or biracial, Latinx/White biracials' identity is influenced by skin tone and class, and Asian/White biracials are most likely to identify as White or biracial (Davenport 2018;Golash-Boza and Darity 2008;Quiros and Dawson 2013).
The absence of U.S.-based slavery among other historically underrepresented groups affects contemporary individuals' racialized understanding and racial affiliations. American Indian/Whites, Asian/Whites, Native Hawaiian or Pacific Islander (NHOPI)/Whites, Latinx/Whites, Middle Eastern or North African (MENA)/Whites tend to have more variation in their racialized self-understanding than Black/White biracials. For example, historically, high rates of consensual and nonconsensual relationships between American Indians and Whites increased the number of American Indian/Whites, simultaneously diminishing the American Indian population if American Indian/Whites are understood as not "fully" American Indian (Dunbar-Ortiz 2015). Unlike enslaved Africans who historically had various types of relations with Whites and tended to retain (except in instances of passing, in which White-passing people with Black ancestry performed Whiteness in order to receive social, political, and economic benefits) a Black identity and slave status, American Indians were forcibly assimilated to Whiteness. This hiding of identities was used as a form of cultural erasure and disconnection, thus facilitating the assimilation to White culture. The continuation of this historical interpretation of race is shown in contemporary understandings of racial connectedness. For example, Pew Research Center (2015) reports that 61 percent of American Indian/White adults state they have a lot in common with Whites, while 22 percent believe they have a lot in common with American Indians. Variations in how different racial groups have and continue to be constructed undoubtedly influence the identity options available in present-day society and the likelihood of racial identification with certain groups.
Given the varying dynamics of racial identity and racial composition groups, some scholars have emphasized the importance of street race and reflected race. Street race refers to the race that people "on the street" see someone as ; this dimension of racial identity emphasizes how others racialize someone. This is important when analyzing discrimination, as the way that others racially categorize someone may impact their treatment of the individual. In a related vein, "reflected race" refers to the race that one believes others see them as (Khanna 2004;Brunsma 2004, 2008). This may be influenced by how they are treated and by what other people inform them their race is, but, ultimately, reflected race (also called "reflected appraisals") focuses on how an individual believes others racially classify them. These concepts tie into racial identification, as people of varying racial component groups may have a street race and a reflected race that may or may not comport with their own racial identity, but that, nonetheless, may influence their racialized experiences in the world Mills 2017;Rockquemore and Brunsma 2004).

Linked Fate
The theory of linked fate purports that an individual's closeness to people who identify with a specific racial group is associated with the individual's belief that their life chances are tied to the group (Dawson 1994;Gay and Tate 1998;Jaynes and Williams 1989). Linked fate-believing what happens to a certain racial group impacts you-influences one's politics, ideologies, and self-concept (Dawson 1994;Sellers et al. 1997;Simien 2005). Thus, linked fate "reflect(s) a sense of belonging or conscious loyalty to the group in question" (Simien 2005). Linked fate emphasizes how discrimination is central to Black solidarity (Gay 2004;Shelby 2005) and has been shown to lead to increased panethnic group consciousness among Asian Americans and Latinxs (Masuoka 2006). Davenport extends the idea of linked fate by applying this concept to people with mixed-race ancestry to determine how this influences their political attitudes and group consciousness, finding that Asian/White and Black/White multiracials may assert greater solidarity and linked fate with their minority race (Davenport et al. 2021;Davenport et al. 2022). In addition, for biracials with non-White/White ancestries who identify as multiracial or as monoracial people of color, discrimination increases their connections to racial minority backgrounds and influences their political identities (Davenport 2018, p. 105). In this way, multiracials tend to align themselves with their minority component group, and discrimination is associated with a strengthened linked fate with racial minorities.
Linked fate is frequently connected to political participation as an indicator of racial connectedness (Jones-Correa 2011;McConnaughy et al. 2010;Simien 2005;Wilkinson 2014;Wong et al. 2005). Feeling a "common fate" with a specific racial group is connected to behavioral outcomes, such as variations in voter turnout (Chong and Rogers 2005). Given this, researchers have broadened the applicability of linked fate to the intersecting identities of race and gender (Simien 2005), and race, class, and religion (Gay et al. 2016), showing the significance of racialized identity even when other identities are considered.
In addition to considering the significance of linked fate on identification and political participation, scholars have contemplated the connection between linked fate and discrimination. For example, Dawson (1994) asserts that when Black people experience racial discrimination this leads to linked fate with other Black people. Lu and Jones (2019) add that, even if racial minorities do not experience racial discrimination themselves, perceiving discrimination against their group is an equally strong predictor in their feelings of racially-linked fate. Furthermore, Herman (2004) finds that racial discrimination increases the likelihood of a racial minority identity. Multiracial boundary transgression generally emphasizes entry into monoracial spaces and/or acceptance of multiracials by monoracials. People who know of mixed-race ancestry could theoretically feel connected to (and affected by) at least two different monoracial groups. Alternatively, they may feel social solidarity with biracials.
Linked fate may also be connected to racial discrimination, in that discrimination can impact the linked fate one feels not only with their own groups, but also with other racial minoritized groups. The oppressed minority ideology (Sellers et al. 1998) is a belief system that asserts one racial minority group's oppression is unique and simultaneously has commonalities with the oppressions of other racial minorities. The common ingroup identity model (Gaertner and Dovidio 2000) emphasizes that racial discrimination against one's own group increases positive attitudes toward and feelings of closeness with other minoritized groups. Thus, racial discrimination may increase linked fate with multiple racial minoritized groups.

Racial Discrimination
Everyday discrimination is the routine and seemingly minor discrimination people experience that results in chronic discrimination (Essed 1991;Williams et al. 1997). While the majority (55 percent) of U.S. adults with mixed-race ancestry report perceiving to be racially discriminated against (Pew Research Center 2015), racial discrimination still reflects the current U.S. racial hierarchy. People who identify as biracial or as a monoracial person of color tend to report similar percentages of discrimination, which are higher than respondents who are aware of mixed ancestry and identify as White. Perceiving discrimination emphasizes the awareness and internal processing of events, which is related to how people believe others see them (Sellers and Shelton 2003) and strength of racial/ethnic identity (Gong et al. 2017). In addition, as perceived discrimination increases the likelihood that people of mixed-race ancestry report solidarity with multiracials as a group increases (Giamo et al. 2012). Building on this work, the present study emphasizes the relationship between perceived discrimination and identification, highlighting that discrimination may be associated with monoracial identification, particularly when factoring in skin tone.
Perceiving discrimination may impact an individual's closeness to a particular racial group and an individual's self-identity. For example, Dowling (2015) found that Mexican American interviewees who identified as "Chicano" or "Brown" or "Other" were more likely to perceive their experiences as discriminatory, while interviewees who had negative experiences but did not interpret them as discrimination tended to identify as "White" or "American" on a form. Thus, identity was not related to the likelihood of experiencing exclusion or racism, but rather, was associated with the ways interviewees interpreted and responded to these interactions. Specifically, Dowling (2015, pp. 52-53) states: "What distinguishes 'other race' Mexican American respondents from those who identify as 'white' on the census is . . . how they see racism and the strategies they employ to deal with discrimination." Dowling's findings provide further evidence that perceived discrimination is a powerful tool that can influence racial identification. Furthermore, interviews with multiracial activists explicitly emphasize the importance of discrimination and how racial discrimination is against all people of color broadly (Masuoka 2017). According to an interviewee who self-identifies as multiracial, and whose racial makeup is Black, White, and Native American: "Those guys are not thinking that way, in terms of he's multiracial [or] he's African . . . No, they're just saying this guy has brown skin, I don't like him. You know, and they don't really care about what color brown skin, you know, got brown skin . . . if he's not white he's not pure, he should go back to wherever he came from. Forgetting, of course, that he came from somewhere else. No, I think racism doesn't go for finer points" (Masuoka 2017, p. 67).
Believing one experiences discrimination due to their race has the power to influence the racial group(s) one feels connected to and excluded from. In this way, perceived discrimination can impact which racial groups one feels they can claim or not, and the spaces in which they may or may not be welcome. This example also shows how perceived discrimination is directly connected to skin color and has the potential to impact one's racial identification. Discrimination and skin color may both be important factors at play regarding racial identification.

Skin Tone
Skin tone is both connected to discrimination and is impactful to racial identity in its own right. Perceptions of discrimination tend to increase as skin tone darkens (Hersch 2006(Hersch , 2008Rondilla and Spickard 2007). Klonoff and Landrine (2000) go as far as asserting skin tone is a marker of discrimination, an assertion supported by their findings that darker-skinned Blacks were 11 times more likely than lighter-skinned Blacks to report high levels of racial discrimination. These perceptions are connected to lived experiences, such as darker-skinned Black people tending to have lower levels of educational attainment, household income, and occupational status (Monk 2014). Many Asians receive messages from family members, media representations, and beauty campaigns that lighter-skin is preferable (Rondilla and Spickard 2007). Latinxs, Arab Americans, and South Asian Americans, while tending to reject Whiteness, often idolize lighter-skin and experience racialized outcomes that worsens as skin tone darkens (Grewal 2009;Quiros and Dawson 2013). Latinxs with darker-skin are rated as having lower occupational prestige, lower wages, and higher levels of depression than lighter-skinned Latinxs or Whites (Bonilla-Silva 2004;Espino and Franz 2002;Hochschild and Weaver 2007); yet lighter-skinned Latinxs are more likely to identify as White (though they do not always identify as such) and darker-skinned Latinxs receive benefits from not identifying as Black (Golash-Boza and Darity 2008;Quiros and Dawson 2013). Among multiracials, Herman (2004) finds that those who are darker-skinned perceived higher amounts of discrimination, and linked this perceived discrimination with a racial minority identity. While other physical attributes, such as hair and body shape, are purported as more or less desirable, skin tone is the master marker of difference (Herring et al. 2004).
In terms of skin tone's impact on racial identification, contemporary studies reveal that people with mixed-race ancestry who have a more "prototypical" phenotype-in particular, darker skin-are more likely to identify as multi-or monoracial people of color. Davenport (2018, p. 87) explains that darker skinned Asian/Whites and Latinx/Whites are more likely to assert a non-White identity than their lighter-skinned counterparts specifically because "their darker skin heightens the visibility of minority background." Lighter-skinned Latinxs are more likely to identify as White (though they do not always identify as such) and darker-skinned Latinxs receive benefits from not identifying as Black (Golash-Boza and Darity 2008;Quiros and Dawson 2013). Khanna (2011) and Mills (2017) both find that Black multiracials with darker skin are more likely to assert a Black identity.
Appearance, and in particular skin tone, has been found to be directly connected to racial identification. Regarding the significance of physical appearance broadly, Rockquemore and Brunsma (2004) describe how physical appearance is often used to assert a connection to Black identity; Ahnallen et al. (2006) indicate how multiracials with Japanese/White ancestry connect physical appearance with a sense of belonging and racial identification; and Gonzales-Backen and Umaña-Taylor (2011) indicate that Latinx youths associate appearance and darker skin with Latinx identity affirmations. While other physical attributes, such as hair and body shape, are purported as more or less desirable, skin tone is the master marker of difference (Herring et al. 2004). People of color with darker skin continue to be perceived more negatively and are more likely to experience multiple forms of discrimination (Hebl et al. 2012;Monk 2014Monk , 2015Rondilla and Spickard 2007). Given the importance of skin tone, scholars have considered how appearance broadly, and skin tone in particular, influences identification.
Although Brunsma and Rockquemore (2001) find that it is not skin tone, but rather one's assumptions of how others perceive them, that impact Black/White biracials' identity, their foundational research was conducted just as the U.S. Census was allowing citizens to choose "more than one" racial category for the first time in the 2000 Census. Thus, the identification options available may have felt more stringent to respondents in the early 2000s, compared to respondents on the cusp of a new decade in 2019, when the present study was conducted. This suggests further need to test the potential connection skin tone may have in the identification choices of people with mixed-race ancestry today.

Hypotheses
Given the current literature on variations among biracials, linked fate, racial discrimination, and skin tone, I hypothesize: Hypotheses 1 (H1). A one unit increase in linked fate with a specific racial group will be associated with greater odds of identification as a member of that racial group, for people within that racial category (e.g., among respondents with Asian ancestry, feeling linked fate with Asians will be associated with greater odds of identifying as multi-or monoracially Asian compared to not identifying as Asian).

Hypotheses 2 (H2).
A one unit increase in personally experiencing discrimination will be associated with increased odds of identification as a multi-or monoracial person of color within that group (e.g., among people with Black ancestry, increased racial discrimination will be associated with greater likelihood of identifying as mono-or multiracially Black, compared to not identifying as Black in any way).

Hypotheses 3 (H3).
A one-unit increase in skin tone darkness will be associated with greater likelihood of identification as a multi-or monoracial person of color within that group (e.g., among people with Latinx ancestry, increases in skin darkness will be associated with greater likelihood of identifying as mono-or multiracially Latinx, compared to not identifying as Latinx).

Data & Methods
In order to address the study hypotheses, I administered an online survey titled The Mixed-Race Ancestry Survey (Gonlin 2019) to determine the relationship between discrimination, linked fate, skin tone, and racial identification. I used Mechanical Turk (MTurk) because this online platform is available throughout the U.S., making my sample broader than a locally based sample, such as a survey of one institution. In addition, a growing body of literature indicates that MTurk respondents are less likely to fail attention checks than college students (Anson 2018) and are more socioeconomically and ethnically diverse than college students (Casler et al. 2013). MTurk respondents also have a higher-than-average completion rate, likely because not completing a survey could lead to them receiving a bad evaluation, so MTurk users who begin a survey tend to have high completion rates and do well on attention checks (Anson 2018). Using MTurk allowed me to have a large enough sample of people with mixed-race ancestry to analyze by racial component groups. I dropped respondents who did not finish completing the survey and those who had missing data on key or control variables, bringing the sample size to N = 486. Though this sample constituted people who indicate having mixed ancestry, they were not required to identify as "mixed" or "multiracial" in order to be included.

Variables
I asked respondents to self-report the amount of racial discrimination they experienced by using the Everyday Discrimination Scale (EDS) (Williams et al. 1997). Afterwards, I asked questions indicating the amount of linked fate they experienced with every racial group, including with bi/multiracials. I then instructed respondents to racially self-identify. Finally, in addition to the aforementioned key variables, respondents were asked to report their gender identity, personal pronouns, age, education, full-time employment status, median household income, and geography.
Racial ancestry/background. In order to gain entry into the survey, participants were required to check off having two-or-more racial groups in their ancestry. Participants who did not indicate having multiple racial groups in their ancestry were not able to take the survey. Those who were aware of having multiple racial ancestral groups were able to complete the questionnaire, which included questions about their own racial identification later in the survey.
To determine their racial ancestry, participants were asked to: "Please indicate one or more races that are part of your racial heritage or background," and are provided the options White, Black/African American, American Indian/Alaska Native, Asian, Native Hawaiian or Pacific Islander, Hispanic or Latina/o/x, 3 and/or Middle Eastern or Northern African (MENA). The last two racial categories, though not included in the U.S. Census question on race, offered respondents racially identifying with either of these groups the option to disclose that in survey format. Including these last two groups allowed me to consider the experiences of Latinxs and MENAs with mixed-race ancestry, in addition to the traditional racial groups studied.
Linked fate. Survey questions measuring linked fate were constructed from the American National Election Survey (ANES), chosen for its numerous detailed questions on linked fate. For decades, the ANES has provided quantitative data on voting trends, public opinion, and political participation; particularly impactful for this current study are the questions about linked fate that ANES incorporates, which are intended to determine the impact of linked fate on political solidarity, but may be applicable to other realms of social life and connection. Adapting the ANES questions for this particular study, I asked respondents about the amount of linked fate they shared with every group in order to analyze how much variation there was in their responses. They were asked, "How much do you think what happens to [racial group] in the U.S. will affect your life?" They indicated linked fate on a four-point scale, with the options "a great deal," "a lot," "a moderate amount," "a little," or "none at all." Respondents were asked this same question for every racial group in this survey, including bi-and multiracials.
Racial discrimination. To determine how often respondents experience discrimination due to their race or ethnicity, they were asked a series of questions from the Everyday Discrimination Scale (EDS). Constructed with the goal of measuring everyday discriminationchronic, routine, and seemingly minor instances of unfair treatment (Essed 1991)-the EDS consists of nine questions that indicate the frequency of discriminatory interactions. The original EDS (Williams et al. 1997) did not explicitly ask about discrimination due to race; in this study, the EDS was adapted to highlight racial/ethnic discrimination (previously scholars have found that the EDS is largely very adaptable to analyzing racial/ethnic discrimination (Kim et al. 2014)). Respondents were asked how often they: (1) were treated with less courtesy than other people because of their race or ethnicity, (2) were treated with less respect than other people because of their race or ethnicity, (3) received poorer service than other people in restaurants or stores because of their race or ethnicity, (4) thought people acted as if they thought the respondent is not smart because of their race or ethnicity, (5) thought people acted as if they were better than the respondent because of their race or ethnicity, (6) thought people acted as if they were afraid of the respondent because of their race or ethnicity, (7) were treated as if they thought the respondent was dishonest because of their race or ethnicity, (8) were called names or insulted because of their race or ethnicity, and (9) were threatened or harassed because of their race or ethnicity (Cronbach's alpha 0.96). Respondents chose between the options "everyday," "a few times a week," "a few times a month," "a few times a year," "less than a year," or "never." Tabulations of each individual EDS item revealed that respondents answered different EDS questions in a variety of ways; for example, respondents tended to answer "a few times a year" on some questions and "never" on other questions, indicating that respondents did not simply choose the same response to every item.
Skin Tone. Respondents may have had varying ideas of what constituted "light" and "dark" for different groups. Therefore, respondents were asked to rate their own skin tone on a 69-item skin tone palette created by researchers at the Racial and Ethnic Studies Institute at Texas A&M University . Unlike more common skin tone measures that focus on light to dark skin shades, this palette included variations in undertone as well as darkness of skin color, making it applicable across racial groups. In addition, Brunsma and Rockquemore (2001) find that biracials' self-reported skin tone may vary based on the racial makeup of their social settings; this gives further credence to the notion that a skin tone measure in which individuals look at their hand then select their skin color, rather than stating they are very light, light, medium, dark, or very dark, is particularly important for people of mixed-race ancestry. Respondents were provided this palette and instructed to: "Look at your hand. Which color shown below comes closest to your skin color?" Categories A through G were undertones and 1 through 10 were skin lightness or darkness. I used categories 1 through 10 when discussing variations in responses by skin tone, then recoded categories 1-2 as very light, 3-4 light, 5-7 medium, 8-9 dark, and 10-12 very dark. Due to low sample size of respondents who indicated their skin tone was dark (N = 23) or very dark (N = 15), I combined dark and very dark participants together (N = 38).
Racial identification. While this survey incorporates people who are aware of personally having mixed-race ancestry, this does not mean respondents identify as multiracial. Therefore, after completing the aforementioned questions, participants were instructed to: "Please choose one or more races that you consider yourself to be" (emphasis added). They were given the same racial categories as provided earlier (i.e., White, Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, Hispanic or Latina/o/x, and/or Arab or Middle Eastern or Northern African). Respondents had the option of choosing more than one racial category, and those who chose two-or-more racial groups were considered to have a multiracial identification. Again, this question focused on participants' own racial identification, rather than whether or not they were aware of multiracial ancestry, which distinguishes this from the earlier question regarding racial ancestry. This question was informed by the Multidimensional Model of Racial Identity (MMRI; Sellers et al. 1998), which provided useful questions on racial identity. The MMRI is a model constructed to analyze four dimensions of Black racial identity (but has since then been applied more broadly to additional racial and ethnic groups).
Analysis. I used Stata 16 to analyze the data collected. I tested binary logistic regression models predicting the likelihood of racial identification, and removed missing data via listwise deletion. Results that are statistically significant at p < 0.05, p < 0.01, and p < 0.001 levels are presented.
My positionality. Critical quantitative methodologists address the frequent critique of quantitative methods as incorrectly assuming objectivity on the part of the researcher by encouraging quantitative scholars to acknowledge their own social identities within a hierarchical social structure, which influence the questions they ask and the ways they analyze data (Carter and Hurtado 2007;Garcia et al. 2018). My identity as a Black biracial woman impacted this research as it lead me to question which racial groups people of mixed-race ancestry feel tied to and to expect that for some people this could be "multiple groups." This identity and witnessing others' experiences also prompted me to recognize how racial discrimination and skin color are often connected to racial identification, leading me to question the salience of these variables in the racial identification for a broad array of mixed-race ancestry people, noting that while there are similarities across BIPOC groups, the overlapping yet divergent histories of racial groups in the U.S. shape the contemporary experiences of people with mixed-race ancestry. Acknowledging my positionality provides necessary context for this work. Table 1   The majority of respondents were aware of having part-White ancestry (72 percent). Respondents tended to be age 18-29 (56 percent), hold a bachelor's degree (42 percent), and be employed full-time (70 percent). Respondents were equally divided between selfidentified men (46 percent) and women (46 percent) and tended to use he/him (46 percent) or she/her (46 percent) pronouns. Respondents most often fell into the $30,001-$45,000 income bracket (19 percent). The most common states for respondents to live in were California (12 percent), Indiana (12 percent), and Texas (11 percent).

Descriptive Statistics
Seventy-seven percent of participants in this survey identify as multiracial rather than monoracial. This high percentage stands regardless of the ancestry group involved: 77% of people with White ancestry, 78% of people with Black ancestry, 69% of people with AIAN ancestry, 77% with Asian ancestry, 85% with NHOPI ancestry, 83% with Latinx ancestry, and 86% with MEAN ancestry identify as multiracial.

Linked Fate on Racial Identification
As shown in Table 2, linked fate is a significant predictor for identification with all but one (MENA) respective groups. The magnitude of this relationship is greatest for the likelihood of identifying in whole or in part as Black, the odds of which more than double (OR: 2.15; p < 0.001) for each one unit increase in linked fate with Black people. This level of magnitude is similar for identification as mono-or multiracial AIAN (OR: 1.60; p < 0.001), Asian (OR: 1.82; p < 0.01), or Latinx (OR: 1.58; p < 0.05). Respondents with White ancestry have a positive relationship between linked fate and racial identification with that group, but at a slightly lower magnitude (OR: 1.52; p < 0.01). Finally, a one unit increase in linked fate with multiracials is associated with 1.37 (p < 0.01) greater odds of identifying as multiracial (rather than as monoracial 5 ). These results indicate the strong connection between linked fate and identification with each specific racial group. Overall, these results show linked fate is clearly tied to racial identification, and in particular, to mono-or multiracial Black identification. Data: Mixed-Race Ancestry Survey, 2019. Robust standard errors in parentheses. *** p < 0.001, ** p < 0.01, * p < 0.05. Note that respondents may be in both the multiracial identity column as well as a monoracial identity column.

Racial Discrimination on Racial Identification
Reports of experiencing everyday discrimination are statistically significantly connected to racial identification with three groups: White, Latinx, and Black people. Table 3 indicates that with one unit increase in racial discrimination, the odds of identifying as White slightly decrease (OR: 0.96; p < 0.01) or as Latinx slightly decrease (OR: 0.96; p < 0.05). We see the opposite effect for Black identification: a one unit increase in experiencing everyday racial discrimination is slightly associated with greater odds of identifying as mono-or multiracially Black (OR: 1.07; p < 0.01). Notably, a one unit increase in skin darkness is also associated with higher odds of identifying as Black (OR: 2.74; p < 0.001). Thus, experiencing discrimination is associated with lower odds of identifying as White, greater odds of identifying as Black, and is not significantly associated with identification with any other racial group. Data: Mixed-Race Ancestry Survey, 2019. Robust standard errors in parentheses. *** p < 0.001, ** p < 0.01, * p < 0.05. Note that respondents may be in both the multiracial identity column as well as a monoracial identity column.
As a sensitivity test, given that both factor analysis and principle components analysis revealed that only the first variable in the Everyday Discrimination Scale was necessary to include (factor analysis: the first factor accounted for 0.99 proportion of variance; PCA: the first component accounted for 0.76 proportion of variance), I reran the aforementioned model using only the first variable in the EDS. The results were remarkably similar, with the first EDS variable (being treated with less courtesy than others due to race/ethnicity) negatively significantly associated with White identification (OR: 0.79; p < 0.05) and positively significantly associated with Black identification to an even stronger degree (OR: 1.84; p < 0.001). The odds of Latinx identification, which was marginally significantly associated with EDS, are not significantly associated with this particular EDS variable. As with the original results, skin tone was a major predictor for Black identity: among people with part-Black ancestry, a one unit increase in skin tone depth was significantly associated with two and a half times greater odds of Black identification (OR: 2.56, p < 0.001). Please see Appendix B, Table A1 to view this sensitivity test.

Skin Tone on Racial Identification
Skin tone was associated with predicting racial identification as Black and as multiracial in general. Table 4 shows that skin tone was one of the most significant factors in predicting Black identification-a one unit increase in skin tone darkness was associated with respondents being three times more likely to identify mono-or multiracially as Black (OR: 3.03, p < 0.001), even when controlling for gender, age, education, employment, income, and living in the southeast. Skin tone depth was also associated with multiracial identity: a one unit increase in skin darkness was associated with 1.29 greater odds of identifying as multiracial (p < 0.05). However, skin tone was not significantly associated with racial identification, in whole or in part, as any other racial group. Data: Mixed-Race Ancestry Survey, 2019. Robust standard errors in parentheses. *** p < 0.001, * p < 0.05. Note that respondents may be in both the multiracial identity column as well as a monoracial identity column.

Discussion
Linked fate, racial discrimination, and skin tone all work to influence the racial identification of people of mixed-race ancestry. The significance of these interconnected phenomena are representative of how the oppressed minority ideology (Sellers et al. 1998) and the common ingroup identity model (Gaertner and Dovidio 2000) may be at work here and may be connected to racial identification, as both increased linked fate and increased racial discrimination are associated with greater likelihood of identification with racial minorities. Skin tone, which is associated with increased discrimination, is clearly connected to the likelihood of Black and multiracial identification. Working in tandem, linked fate, racial discrimination, and skin tone help predict the identification outcomes of people of mixedrace ancestry within a White hegemonic society.

Linked Fate
As predicted, linked fate is a significant predictor for identification across the board, with nearly all respective groups (the one exception being that linked fate with MENA is not significantly associated with increased likelihood of MENA identification). This highlights the importance of linked fate-the more one feels their fate is tied to a racial group, the more likely they are to identify with that group. This magnitude of this relationship is particularly strong for identification as Black. Black identity may be especially associated with linked fate due to the historical categorization and treatment of any person with "one drop of Black blood" as Black (Davis 1991), which likely strengthened the idea that what happens to Black people at large impacts a Black individual. Although declining in usage over more recent decades, the one-drop rule persists in how others racialize people with Black ancestry, and thus influences how these individuals racially self-identify (Khanna 2010). The idea of linked fate itself was introduced with a focus on Black people (Dawson 1994), and later was expanded to incorporate additional racial and/or marginalized groups (Gay et al. 2016;Lu and Jones 2019;Sanchez and Vargas 2016;Schildkraut 2017;Simien 2005).
The finding that increased linked fate is associated with greater likelihood of racial identification across the board except with MENA is intriguing. MENA may have been the only group in which linked fate was not significantly associated with racial identification, due, in part, to the struggle of determining whether or not MENA should be considered a racial category. For example, Middle Eastern/North Africans are not currently constructed as their own racial group on the U.S. Census; rather, MENA are combined in the White category. This may be related to MENA being a relatively newly constructed category in U.S. history. While MENA have been integral to the U.S. for over a century (Bakalian and Bozorgmehr 2011;Beydoun 2013), the construction of a "MENA" racial group has, as of yet, been thwarted (Maghbouleh 2020). This may, for the time being, prevent a cohesive group identification from fully forming, leading participants to be less likely to identify with MENA than other racial groups that are more prominently constructed.
When undertaking analyses of identification for the mixed-race population, the racial group(s) respondents feel linked fate with should be considered as a key variable. Within a White-centric social structure, the trends in who mixed-race ancestry people feel connected to can indicate their proximity (imagined or real) to Whiteness, to groups of color, and to other mixed-race people. Who respondents feel linked to may be indicative of where racial boundaries are shifting, where boundaries are transgressed, and variations in how likely this is to happen for people of mixed-race ancestry with different racial component groups.

Racial Discrimination
Personally experiencing racial discrimination is significantly associated with three of the racial groups presented herein. As predicted, increased racial discrimination is connected to lower odds of identifying as White and higher odds of identifying as Black; increased discrimination was marginally associated with slightly decreased Latinx identification. As reports of anti-Black discrimination in the U.S. remain high, this finding of increased discrimination being associated with Black identity would lead us to anticipate that, for respondents with multiple identity options potentially available, identifying as Black may be the outcome for more and more people. A caveat to this potential outcome, though, is that while discrimination was significantly associated with Black identification, the effect size was relatively small at an odds ratio of 1.07. Within a racially structured society that places White and Whiteness at the top and Black and Blackness at the bottom of this hierarchy, respondents may find that everyday discrimination is just one of the ways their lives are constrained. Black identification is associated with discrimination, but is also connected to a myriad of other factors, including phenotype, cultural norms, and the persistence of the one-drop rule (Buggs 2019;Khanna 2010).
Interestingly, racial discrimination is not associated with higher odds of identifying as AIAN, Asian, Latinx, or MENA. There are a few potential explanations here. First, given that this sample is of respondents who personally have mixed-race ancestry, those with AIAN, Asian, Latinx, or MENA ancestry tend to also have White ancestry. For example, fourteen percent of the entire sample is comprised of people with White/AIAN ancestry and thirteen percent is comprised of people with White/Latinx ancestry. Given the history of colonization of AIANs and Latinxs and of subsuming MENA into the White category, respondents with these ancestry groups may tend to experience the world as White-passing or White-adjacent individuals (Bonilla-Silva 2004), impacting the amounts of racial discrimination they do or do not receive.
Second, groups may be experiencing discrimination due to not being seen as BIPOC "enough." It is possible and likely that mixed-race respondents experience discrimination due to their race and also discrimination due to not being seen as "real" BIPOC. This may lead some respondents to more strongly assert a BIPOC identity, while others may feel they cannot claim a BIPOC identity (Mills 2017). This potentially "cancels out" statistical results. Unfortunately, the data available do not allow me to determine where the discrimination is coming from or why respondents feel they are being discriminated against. Yet it is noteworthy that in a society predicated on the supremacy of Whiteness, experiences of racial discrimination may be weaponized to otherize some BIPOC while still avoiding incorporating them into the "White" category, as described in Bonilla-Silva's (2004) assertion of a pigmentocracy with "honorary Whites" as a middle category between Whites and "collective Blacks".
Given the U.S.'s centering and construction of Whiteness as "ideal" and "desirable," it is plausible that some respondents experiencing racial discrimination may move away from a BIPOC identification as a strategy to avoid persecution. Discrimination might lead them to work towards assimilating into Whiteness, a strategy we have seen historically used by Eastern and Southern European immigrant groups to the U.S. attempting to avoid discrimination (Roediger 2005), Mexican Americans claiming rights as White citizens under the Treaty of Guadalupe Hidalgo (Del Castillo 1992;Hernández 2001), and Arab nationals' ability to immigrate to and become citizens of the U.S. under a White racial label (Beydoun 2013).

Skin Tone
Skin tone is significantly connected to Black identification. As expected, the darker one's skin tone is, the greater odds of Black identification. These results are in line with interviews conducted by Khanna (2011) andMills (2017, p. 171), both of whom found that Black multiracials with darker skin were more likely to assert a Black identity. Lighterskinned Black people are more likely to have their Black authenticity questioned than medium or darker-skinned Black people (Hunter 2007), increasing the likelihood that their reflected race (the way they believe others see them) and their street race (the way that strangers racialize them) may not match the way people with Black ancestry racialize themselves. The significance of skin tone in impacting multi-or monoracial Black identification highlights the potential that darker skin tone is associated with Blackness, leading those with lighter skin to be more likely to have their racial identity challenged (Hunter 2007), which may impact their own self-classification (Khanna 2011;Mills 2017).
In addition, I find darker skin is associated with a greater likelihood of claiming a multiracial identification. This is true for people of all racial component groups, not just multiracials with Black ancestry. The greater likelihood of multiracial identification as skin tone darkens may be due in part to how people are perceived based on skin color. Those with lighter or medium skin may be seen as more racially ambiguous, allowing them greater variety in identity options that are deemed viable by others.
More surprising is that skin tone is not significantly associated with the racial identification trends of any other group. Stephens and Fernández (2012), for example, found that self-described White Hispanic women preferred "tan" skin over "pale" skin because they believed the former increased desirability among peers, value in dating contexts, sexual appeal to men, and was seen as a marker of Hispanic identity in social settings. It is possible that the perception of lighter skin may lead some racial group members to feel they are not "enough," whereas dark skin may be seen as, in this example, "Black" rather than "Latinx". Thus, perceptions of medium skin tone (as compared to light or darker skin) may be the overarching phenomenon at play.
While skin tone is significantly associated with Black identification and multiracial identification in this study, and has been shown to be connected to Latinx identification (Stephens and Fernández 2012), this is not necessarily the case for Asian Americans. Rondilla and Spickard (2007) determine that Asian Americans experience colorism in that lighter skin is lauded and darker skin is often rejected, but having light or dark skin is not necessarily associated with being "more" or "less" Asian. This may account for the lack of significance between skin tone and Asian identification. These skin tone results highlight that, while people of mixed-race ancestry often have overlapping experiences, the social world we live in racializes people of different racial component groups such that the experiences of some mixed-race people vastly differ from other mixed-race people. Skin tone is one example of how a variable can be incredibly important for the racial identification of some groups, but not significantly associated with the racial identification of other mixed-race groups.

Conclusions
People of mixed-race ancestry may have complex identities that are often formed within racially reductive societies. Rather than attempting to "box" people in, scholars may consider this increased identity complexity in more nuanced ways. The work herein highlights the need for acknowledging the differences people of mixed-race ancestry experience by racial component groups allows us to consider unique struggles and privileges various groups are offered, which has implications for how we analyze racial group categories at large. Addressing three factors that influence mixed-race identification-linked fate, racial discrimination, and skin tone-highlights how people of mixed-race ancestry may come to a racial identification, and that these variables have important differences by group.
My findings that linked fate and racial discrimination are significantly associated with racial identification with racial minority groups evidences how identity may be formed around stigma-based solidarity, as perceived discrimination shapes intergroup relations among stigmatized groups (Craig and Richeson 2016). Structured within a racialized social structure that places Whiteness at the top, experiencing racial discrimination can increase feelings of commonality and identification with multiple minoritized groups. That skin tone is particularly significantly associated with racial identification as mono-or multiracially Black furthers scholarship indicating that, while the one-drop rule may be connected to Black identification particularly in the south (Khanna 2010), as a nation we may be moving away from hypodescent and towards an emphasis on phenotypic features being major predictors of racial identification (O'Connell et al. 2020). As the U.S. approaches becoming a majority-minority society (Colby and Ortman 2015), scholars must carefully consider the racial identification "options" for people who may feel connected to multiple racial groups to prevent miscategorization and to increase our understanding of complex racial identities. Considering how people may identify with multiple racial groups challenges the White hegemonic norm that classifies people as either one-or-the-other, and opens up the possibility for a both-and identity formation.
I encourage researchers to strongly consider incorporating measures of linked fate, racial discrimination, and skin tone in future models intended to predict racial identification, particularly when considering identification for the mixed-race ancestry population. As shown, all three variables are important predictors of racial identification. Building on this study, future research may work to address some of the limitations of the present study. First, for all its merits, Mechanical Turk still has draw backs-respondents have to know about MTurk's existence, self-select to participate, and tend to have higher education levels than the general population. Second, sample size restricted me from certain analyses. For example, I did not run models on Native Hawaiian and Other Pacific Islander (NHOPI) identification due to the small number of people (N = 28) with NHOPI ancestry in this study. Sample size also prevented me from conducting interaction effects between skin tone and other variables. Given my emphasis on comparing the experiences of mixed-race people with different racial component groups, I did not aggregate groups together. Data with a larger sample size may allow for interaction effects by component groups. Third, this manuscript highlights that, in addition to people with mixed-race ancestry potentially identifying monoracially, they may also identify with multiple racial groups. Many contemporary scholars have taken care to avoid assigning monolithic racial categorizations and requiring people of mixed-ancestry to identify with one racial group or another. In line with this work, respondents described in the present research may be represented in multiple categories. For example, a respondent who reports their identification as Black and Asian is included in the "Black/African American" as well as the "Asian" categories. Additionally, this data set does not explicitly ask respondents about street race-how people "on the street" classify someone else's race or ethnicity )-nor does this data set explicitly address reflected appraisals (Khanna 2004). It is, therefore, not possible to conclude if street race or reflected race were the mechanisms leading to a lower or higher likelihood of identifying as Black or multiracial. Street race and reflected race may be tied in with the racial discrimination discussed herein, and further analyses may help disentangle the two. Finally, it is noteworthy that one's racial identification may shift over time or by context (Pauker et al. 2018;Rockquemore and Brunsma 2001). For example, Brunsma (2006) finds that even within the same country (the United States), geographic region plays an important role in identity and identification. The present study focuses on one point in time, but recognizes the importance of and need for future longitudinal work.
As people of mixed-race ancestry navigate the social world, who they feel connected to, the racial discrimination they experience, and their skin tone all have a factor in their racial identification. These under-analyzed aspects connected to identification may be particularly important for people with mixed-race ancestry, who arguably have a greater range of identification options available to them. Connecting linked fate, racial discrimination, and skin tone to identification trends allows us to see nuances in how people of mixed-race ancestry assert their racial identity within the constraints of a White-centered racial order.
Funding: This research was made possible in part by financial assistance from the Ruth Landes Memorial research fund, a program of the Reed Foundation.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Texas A&M University (protocol code IRB2019-0186M, approved 3 April 2019).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available as to stay in line with the author's IRB data management plan.

Conflicts of Interest:
The author declares no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the result.

Appendix A
Soc. Sci. 2021, 10, x FOR PEER REVIEW 18 of 23 range of identification options available to them. Connecting linked fate, racial discrimination, and skin tone to identification trends allows us to see nuances in how people of mixed-race ancestry assert their racial identity within the constraints of a White-centered racial order.
Funding: This research was made possible in part by financial assistance from the Ruth Landes Memorial research fund, a program of the Reed Foundation.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Texas A&M University (protocol code IRB2019-0186M, approved 3 April 2019).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available as to stay in line with the author's IRB data management plan.

Conflicts of Interest:
The author declares no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the result.
Appendix A Figure A1. This two-axis color grid was created by Campbell et al. (2020) to include a greater variety of skin tones than those presented in the White-to-Black skin tone continuum provided by many skin tone measures. In addition to more adequately including other racial and ethnic groups' range of skin tones, this measure offers a greater variety of darker skin tones than most contemporary measures incorporate. Figure A1. This two-axis color grid was created by Campbell et al. (2020) to include a greater variety of skin tones than those presented in the White-to-Black skin tone continuum provided by many skin tone measures. In addition to more adequately including other racial and ethnic groups' range of skin tones, this measure offers a greater variety of darker skin tones than most contemporary measures incorporate. Data: Mixed-Race Ancestry Survey, 2019. Robust standard errors in parentheses. *** p < 0.001, * p < 0.05. Note that respondents may be in both the multiracial identity column as well as a monoracial identity column.

Appendix B
Notes 1 I use quotation marks around "choices" to emphasize that many individuals with mixed ancestry feel their racial identity is not something they choose, but rather an identity that is expected of them or that they are treated as (Khanna 2010). Many other individuals with mixed ancestry feel that they do have multiple identification options available (Rockquemore and Arend 2002), and still others report racial identities that vary over time (Doyle and Kao 2007). 2 Racial identification is also known as "self-classification" (Roth 2016), "expressed race" (Roth 2010), "expressed internal race" (Harris and Sim 2002), and "self-reported race" (Saperstein 2006).

3
While I use "Latinx" throughout my writing to be inclusive of all genders, in this survey I use "Latina/o/x" in order to be clear to both academic and nonacademic respondents. 4 By cell I mean the subgroup within a subpopulation. For example, Black/Asians are a subgroup of multiracials.

5
To clarify how these racial identifications were determined: all respondents were first asked about their racial ancestry and later asked about their own racial identification. If, for example, someone with Black ancestry identified as Black alone or in part, they were recorded in the Black category. If someone with Black ancestry selected multiple racial groups to describe their racial identification, they were put in the multiracial category. Thus, one respondent may be in both the "mono-or multiracial Black" group and also in the "multiracial" group. A respondent may also be represented only in the "mono-or multiracial Black" group and not in the "multiracial" group.