What is the difference between race and ethnicity? This is a common question, and there is not an easy answer. In this post, I present the views from various academic sources and try to pull some common ideas from them.
One of the main motivations for this post is that I consistently see “race/ethnicity” used together in academic papers and data sets. Most editorial styles (APA, Chicago, etc.) discourage using slashes because they tend to create ambiguity, and indeed that often happens when race/ethnicity is lumped together without explanation.
Teasing out the terms is quite difficult if not impossible, so it’s understandable that the slashed construction is still widely used. However, I am of the opinion that we owe it to study participants, readers, policy makers, and society to be more specific than this.
Academic Sources on Race vs Ethnicity
APA on Racial and Ethnic Identity
Because this is an academic-writing site, let’s begin with an academic-writing source. APA (3.14, Race and Ethnic Identity) discusses issues of bias in language related to racial and ethnic groups. However, the discussion does not really clarify the distinction well, and definitions are not provided for the terms. Overall, the discussion favors using precise language. For instance, the section recommends that we should specify “ethnic minority” as opposed to just “minority,” and “in general, naming a nation or region of origin is helpful (e.g., Cuban, Salvadoran, or Guatemalan is more specific than Central American or Hispanic)” (p. 75).
Let’s turn to another source, the Chicago Manual of Style. There is a noticeable lack of discussion on this subject, which I find interesting because racial and ethnic differences are commonly measured in academic studies (for instance, a search of “racial” turns up zero results in the manual, and a search of “ethnic” returns a few hits but really only a couple of relevant sections, none of which provide definitions or distinctions between the terms).
Section 8.37 (Ethnic and national groups and associated adjectives) lists terms such as Arabian, Chicano, and Italian American. The section does not list any designations by color, but a later section (8.39, Color) refers to “Common designations of ethnic groups by color,” which further confuses the issue of what constitutes ethnicity and what constitutes race. (As a side note, APA capitalizes racial and ethnic group designations by color, whereas Chicago sets them in lowercase). These two sections do not give much clarity on definitions of or distinctions between the two terms.
Another source common to both APA and Chicago is Merriam-Webster. M-W’s definition 2a for the term ethnic is
Because the word racial is used within the definition for ethnic, we have another source that shows overlap between the terms; that is, ethnicity refers to some kind of shared characteristic, and race may be one of those shared characteristics.
Merriam-Webster’s definitions of race (second set of definitions for the noun, further down the page) also mention shared characteristics in a broad way in a couple of definitions; however, one definition does refer specifically to physical traits (whereas physical characteristics are not mentioned in the definitions for ethnic). This slight difference in definitions indicates a connotation of physicality for one term (race) but not the other (ethnic).
U.S. Census and the CPS
The U.S. Census explains that its definitions of racial categories are not based on biological or genetic characteristics, but the Census’s definition of “Black or African American” refers to only Black people from Africa (therefore, not White people from Africa). The other racial groups are defined exclusively by origin from geographical regions and do not refer to color). Including color for one group but not others is inconsistent, to say the least (note that the Census states that it “has” to follow the government’s 1997 OMB definitions, which to me is an indirect reference that some people in the Census want this to change, so we shall see). Also note that the definition for White includes a mention of “Northern Africa,” which includes countries such as Tunisia, Algeria, Morocco, and Egypt, where people tend to be (but are not exclusively) lighter skinned; of course, Sudan is often grouped with North African countries, and many people there tend to be darker skinned, so this is just another limitation in the Census groupings. Putting all of this together though, when one reads the Census definitions, it is very difficult not to perceive race as referring in whole or in part to color, especially given the definition for Black.
In any case, the 2010 Census measured six racial categories (White, Black or African American, American Indian and Alaska Native, Asian, Native Hawaiian and Other Pacific Islander, Some Other Race) and two ethnic groupings (Hispanic or Latino and Not Hispanic or Latino). Additionally, in the Current Population Survey (CPS), which is a “joint effort” between the Census and Bureau of Labor Statistics that collects extensive demographic data and is a common source in social and policy research, respondents can select multiple racial groups (White, Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander). Before identifying race, however, the CPS asks the respondent “Are you Spanish, Hispanic, or Latino?” (prior to 2003 ethnicity was asked after the race question, and in a slightly different way).
A report on changes to the CPS by Bowler, Ilg, Miller, Robison, and Polivka explains that these three ethnic terms are meant interchangeably, and they are all used to measure ethnicity, not race. The dichotomous grouping for ethnicity (essentially Latino or non-Latino) seems quite limited in both the Census and CPS. For instance, individuals of Asian background do not have the opportunity in the race question to specify their ethnicity as South Korean, Taiwanese, and so forth, and if I were of Middle Eastern descent, I imagine it would be difficult to feel comfortable classifying myself in any of the categories. To me, this is more evidence that the Census’s racial categories are indeed tied more tightly to physical traits than to what I perceive as ethnicity. I’m not saying that this is a bad thing–actually, I like the specificity of having two words for two different concepts (race for shared physical characteristics and ethnicity for shared cultural characteristics). Also, it is important to keep in mind that because the Latino groupings are not used as a racial group in the Census, when using Census or CPS categories for race and ethnicity (either directly from Census or CPS data or indirectly through a data set that uses the Census or CPS groupings), then it seems that we should refer to Hispanic as an ethnic group rather than a racial group.
American Anthropology Association on Ethnicity and Race
Let’s now turn to perhaps a more suitable academic authority on the subjects of race and ethnicity, the American Anthropology Association. Race and ethnicity are common topics in sociocultural anthropology, so I began to search the AAA site to see what I could dig up. The American Anthropology Association’s Statement on Race argues that race is not just biological and that there are not clear-cut definitions. In another page on the AAA site, the organization discusses race vs ethnicity. AAA notes that most people tend to perceive the two terms in a similar way and that, because the Census is unable to clearly distinguish the terms, that the term race be eliminated from “government parlance” altogether. AAA’s stated preference is for ethnic group or ethnic origins.
A particular passage on the second AAA page clarifies a bit:
The idea that the term ethnic refers to culture, language, and national origin sits well with me because I like preciseness in language, and as academic writers, we should be precise and clear when writing. Indeed, in academic studies, it seems more useful to distinguish between, for instance, Mexican Americans and Puerto Rican Americans because their pre- and postimmigration experiences (and access to government benefits) are so different. Indeed, although the two groups share a similar language (with slight variations), they have cultural differences in food, music, dance, celebrations, and other areas. Of course, data sets often do not give us the opportunity to make these kinds of detailed distinctions.
American Sociological Association
I also searched the American Sociological Association website but had difficulty finding posts specifically dedicated to defining the terms. I only note this here because it seems like a subject that ASA could address on the site in the future, given that the concepts are so common in the sociological literature.
Putting It All Together
All of this discussion could leave us just as confused as when we started, so let’s try to pull some threads together.
Race vs Ethnicity: Some Overlap, But Also Some Distinctions
The sources above (among many others not mentioned here) indicate that the term ethnicity refers to shared characteristics such as culture, language, and national origin, whereas the term race has common connotations of genetics or physical characteristics, such as color.
As much as some scholars will argue that the concept of race does not or should not have to do with color, dictionary definitions, Census definitions, and common usage still give race a connotation of color. Also, of course, each country has people with varying skin colors, so, for instance, some people from the Dominican Republic might identify as Black and others as White, but most will likely identify as “Hispanic” or Latino in terms of ethnicity.
So the the question becomes, should we use skin color as a measure in academic studies at all? I argue that we should (in addition to measuring ethnic background) because, to this day, skin color does affect our experiences in society in one way or another. Apartheid in South Africa was an extreme example of this, but of course subtler distinctions exist in each country today. Sometimes one’s skin color affects one’s experience in a given situation more than one’s cultural background and vice versa. Taking a hypothetical example, imagine a situation of housing discrimination in a residential suburb. If studies were to find that, for instance, all Latino groups were discriminated against, we might have an instance of ethnic discrimination, racial discrimination, or both. But if studies were to find that only dark-skinned Latinos were discriminated against, then we would have a strong case for racial discrimination. This is a simplified example, but the practical application for scholars is that, if you are measuring differences in experiences by skin color, ethnic background, or a mix of the two, state your methods clearly so that readers understand which experiences your racial or ethnic categories are meant to capture.
As AAA laments, it might take a while for the government to change its formal classifications of race and ethnicity to be more precise and inclusive of various groups. Those changes, if and when they happen, would probably begin to affect public and media discourse, which in turn will affect how people think of the terms. For now, we have to be aware how terms are perceived by study participants and readers and use terms accordingly.
Be Specific and Clear With Your Language
Lumping groups together in broad categories should be avoided whenever possible. Although many people might view the terms race and ethnicity similarly, as AAA argues, you can avoid confusion by being specific with your language. If your sample consists mostly of Chinese immigrants to New York, state that in your Method section. Explain distinctions between national origin, immigration status, generation status, region of residence before and/or after migration, or other factors that apply to your sample. For instance, a second-generation daughter of Egyptian-born parents will be influenced by two cultures while growing up in the United States, and compared to, for instance, a first-generation Egyptian immigrant, she might identify her ethnicity differently and might have slightly (or very) different ideas about culture, gender relations, religion, work values, and a range of other issues. Including generation status gives us a more detailed understanding of the effects of race, ethnicity, and national origin.
These are just a couple of quick examples to illustrate the importance of specificity and clarity. You can shorten your language in your study after you first explain the ethnic, racial, or other groupings, but just make sure everything is clear to the reader from the outset. Explain what you are really trying to measure, describe the categories and definitions for all groupings, mention limitations to those categorizations so that the reader understands, take the necessary steps at all phases of the research process to ensure that you are measuring true group differences and not methodological flaws, use large enough samples to really measure those effects, and make recommendations for future studies (and perhaps more important, for future data sets, to allow for more nuanced research).
Also, sometimes it is a good idea to clarify your sample briefly again in the Discussion section because some readers glance over the Method and might draw inaccurate conclusions about the results.
The bottom line is that it is best to be specific with your language. My personal take is that you should favor the term ethnicity when you are referring to shared characteristics such as culture and language. For example, Latino would be an ethnicity in my view. Use the term national origin or country of origin if you are exclusively referring to country where a person was born. If you have a data set that allows for this, the distinction will give you richer information on group differences than lumping, for instance, all Asians or Africans together, because there is substantial diversity within those regions. Of course, there is diversity within each country. When possible, use various measures. If you can measure the Census groupings of race and then let participants fill in their ethnicity and national origin in two other spots, you will have three ways to run comparisons to see what is driving the results; again, this might not always be possible, but do the best you can and note the limitations so that future research can take another step.
Finally, of course there is overlap between race, ethnicity, culture, and national origin. Whole bodies of literature have been dedicated to all of these terms. But ultimately, researchers need to be clear with their questions to their participants to obtain the data they need to measure what they want to measure. Researchers then also need to be clear in their articles, books, and other publications to communicate what is really being measured to readers. This clarity will move the literature forward in a more nuanced, informative way than simply saying “race/ethnicity” and allowing the readers to figure the rest out. If your data set forces you to lump groups together, state that limitation in your Limitations section, and call for more detailed data sets in the future. This can be done in one sentence, and the more that the limitation is mentioned, the greater likelihood that data sets will be more specific in the future. Ultimately, with more detailed data, scholars can create stronger analyses and produce more insightful findings. This in turn should make policy and programs more informed and effective. It’s amazing what a little preciseness in our language can do to create change.