Those of us who study racial and ethnic relations in the United States recognize that race is a social construction. What race means, the characteristics and features that we attach to it and the classifications within it (whether Black, White, Asian, and the like), is not static or primordial, but dynamic and changeable. The meaning of race, then, is conditioned on and by an always shifting, societal context. For example, at the turn of the previous century, race was constructed as biological. Distinct racial classifications were understood as reflecting genetic and morphological differences, observable by phenotype. Racial disparities and inequities were explained in biological terms linked to ideas of racial inferiority and superiority.
The notion of race as rooted in biology, with consequent outcomes linked to ascribed deficiencies, or racism, is understood today as an attempt by the dominant (white) group to protect their material interests, like Southern plantation owners who relied on slave labor to maximize their profits during the pre-industrial era. In this way, the social construction of race as “biological” — in the absence of any hard proof or genetic evidence — emerged as a social fact to reproduce racial inequality.
Today we are much less likely to associate racial group membership with genetic endowments.
At the same time, the concept and category of race as a distinct social group persists in the contemporary period. An individual’s racial group membership or identity is still conditioned in part, on phenotype. What this means is that racial classification is both self-defined and externally-imposed. How an individual racially identifies and how he or she is racially identified by others, both matter. Moreover, an individual’s own, personal racial identity “choice” is often but not always, consistent with that which is assigned to him or her by the outside world. For example, although Tiger Woods identifies racially as Cablinasian (Caucasian, Black, Indian and Asian) (Kamiya 1997), most Americans racially identify him as Black only.
Why race as a social construction matters for ethnicity
The dialectical fluidity of race — between self-definition and other-definition, between an individual’s chosen racial identity versus society’s imposed racial identity — facilitates an understanding of race as a social construction. After all, if Tiger Wood’s racial identity does not match that ascribed by the vast majority of American society, then racial identity (although a social fact), is crap beyond the meaning that is attached to it by an individual on the one hand and society on the other. The racial identity mismatch observed in the case of Tiger Woods encourages us to understand race as a less salient, “made up” category of identity, especially when compared against ethnicity, which is self-defined only.
For example, if I was walking down a public street, most Americans would identify me racially as “Latina” but would be less likely to identify me as Mexican-origin. My ethnicity, whether of Mexican, Salvadoran, Puerto Rican, or other Latin American-origin, is indeterminate; they’d have to ask. Additionally, because ethnicity is self-defined, it presumably has meaning for the person identifying with a particular ethnic group. From this perspective, ethnic self-identity matters for individuals and society in a way that racial self-identity doesn’t.
Ironically, this idea is related to a counterintuitive conception of ethnicity that characterizes it as more fluid than race, because one’s ethnicity is always “optional” (Waters 1990). Here the idea is simply that ethnicity is dynamic, fluid and self-defined; as such, anyone can assert any ethnic identity they choose to. And yet, the ethnic identity that they choose, because they choose it, must matter.
The salience of ethnicity when compared to race is also highlighted in the work of some racial and ethnic scholars who refer to race as a “secondary” category of identity, whereas ethnicity is referred to as an “anchoring” or “primary identity” (Itzigsohn and Dore-Cabral 2000). Moreover, the “maintenance” of ethnicity is thought to foster immigrant group cohesion, which may offer some material protection against a negative societal reception context. In contrast, racial identity formation may take place during a process of assimilation, as immigrants and their descendants “lose” their ethnicity, and with it, close-knit ties and sociocultural support. The characterization of ethnicity as a primary, anchoring identity, the maintenance of which offers protection to group members (whereas racial identity does not), underscores the greater importance and salience that American scholars of race relations place on ethnicity.
The perception of ethnicity as a more salient feature of identity is related to its conception as a socially constructed, self-defined identity. Because there is no other option but the option that is chosen by the individual, whether the option makes (common)sense or not, the option selected is accepted without comment. On the other hand, racial identity may be self-defined but is also other-defined. The person who racially identifies one way may or may not be racially identified that way by everyone else.
The problem with this construction of ethnicity is that it tends to reinforce the idea that ethnicity is somehow more “real” (if more fluid) than race. And if ethnicity is more real than race, then for some scholars, it becomes inherent, primordial. This is the slippery slope of ethnic identity formation and why at times, we may forget that ethnicity is as socially constructed as race.
Remember, ethnicity is a social construction too
I started thinking of the relationship between ethnic self-identification and the tendency to interpret ethnicity as more “real” than race when, as part of a new research project, I read a transcribed interview of a self-identified Mexican-origin entrepreneur. This entrepreneur was born in Mexico by immigrant Lebanese parents, she went to boarding school in France, eventually moved back to Mexico for a short time, then moved to the United States where she has been ever since, and where she married a White American man. Her self-defined ethnicity is Mexican, although she speaks Arabic, Spanish and French. Her racial identity is White, and claims that she doesn’t “feel Hispanic,” even though she has applied for and won awards for being a successful “Hispanic” entrepreneur.
In other words, her racial identity is White (she mentions that “No one considers [me] Hispanic”), although she has on occasion identified as Hispanic for instrumental reasons. Either way, the social construction of race is apparent here. Her ethnicity, on the other hand, was confusing to me. Although she self-identifies as Mexican, her parents are Lebanese, she maintains cultural features that are Mexican and Lebanese, she spent a lot of time outside of Mexico when she was growing up, and she racially identifies as non-Hispanic White.
When I finished reading the interview, I wondered about my easy acceptance of her White or Hispanic racial identity, depending, and my confusion and even discomfort about her ethnic self-identity. Why didn’t I readily accept her as ethnically Mexican, when she said she was? At that point I reached out to some colleague-friends and asked them to weigh in on her ethnic identity. Every one of them said she was Mexican; basically, because she said so. Yet, couldn’t we argue that she is also Lebanese? Why didn’t we stop to consider whether she was “more Lebanese” than Mexican, or both? Then again, since ethnicity is socially constructed as self-identified, why shouldn’t she be classified as Mexican if she says so? The point here is that ethnicity as a category of identity is arguably as messy as the category of race is, and yet, we often take ethnic identity at face value. Regarding this entrepreneur, my colleagues were willing to accept her as Mexican because she said so. This belies a salience not to ethnicity, per se, but rather, to the salience of self-identification.
The acceptance of self-identification as real deserves explicit acknowledgement, because it is the reason why we accept ethnic self-identity choices or options without question. Ethnicity could be just as messy as racial identity, if we constructed it as such. But we didn’t, so it isn’t. In fact, we don’t really care about racial self-identity at all, because whether it converges with society’s externally-imposed identity or not doesn’t really matter. A racial mismatch between an individual’s self-identity and society’s is acceptable, while an ethnic mismatch is not. In the end, the only ethnic identity that matters is the one the individual ascribes to. And yet, this doesn’t mean it is more important or anchoring or inherent than racial identity, just that it is socially constructed as self-identified, so it is perceived as such.
 Although this conception also continues to persist. For example, the neoconservative argument that affirmative action is “reverse racism” highlights the undeserved, merit-less, advantages of “less-qualified” racial minorities who benefit from “unfair government” “set-asides” in education and the labor market. More recently, however, observed racial inequality is commonly explained using a “color blind” framework, which is an attempt to explain racial inequality through non-racial means. In other words, to blame “anything but racism” for persistent racial inequality (Bonilla Silva 2003).
 With the exception of critical race scholars, who, in contrast to traditional or mainstream approaches (i.e., the research on assimilation or immigrant incorporation), tend to emphasize race as a systemic or structural force (see the works of Feagin, Moore, Bonilla-Silva).