Of Course Race and Color Matter: A Response
Response to the memo (“Do Race and Color Still Matter? Considerations on the Combined Question”) written by Howard Hogan, former Chief Demographer for the U.S. Census Bureau.
By Alan Aja PhD, Professor in Public & Urban Policy, Chair, Department of Puerto Rican & Latino Studies. Brooklyn College (CUNY).
Herewith I will respond to the statistical validity and key concerns expressed by Dr. Hogan in his memo questioning the proposed Census methodology to merge Hispanic/Latinos into the larger racial category, in lieu of the current arrangement of separate questions that decipher between ethnicity and race, and that have assisted researchers and advocates in understanding not only how race is understood by a socioeconomically diverse Latinx in a U.S. context, but has provided statistical voice to the growing Afro-Latinx community.
As context, I will share that in my first monograph, I used the separated Census race, ethnicity and ancestry variables to distinguish between Black and white Cubans in South Florida by age, gender, wave/era of arrival and other demographic and economic variables (and using each Census 5% sample since 1970 to 2010). I found that despite similar levels of education (and similar access to state-level services and legal rights per refugee status for a large segment of the group), local Black Cubans yielded lower personal and household incomes, higher poverty rates, higher unemployment and prevalence in lower paying jobs. The statistical findings provided context for the personal narratives I collected through interviews, whereas Black Cubans described the extent and contours of their everyday housing to labor market discrimination. In fact, when Dr. Hogan questions why self-identification, instead of community attachment and/or lived experience (see page 2), is most appropriate for Federal statistical analysis, he speaks to the statistical dangers in enveloping all into the race question. In the aforementioned work, a key finding was that Black and white Cubans in South Florida, especially materially, belong to different racialized communities.
In another piece published in Critical Sociology (with Brandon Martínez), we examined home values and the odds of home ownership among Hispanics/Latinos by self-reported race and ancestry groups across the country. We found that white and Asian-identified Hispanics/Latinos were more likely to report higher home values and higher levels of homeownership, while Native/American Indian Latinos and Black Latinos had the lowest shares. Hogan (page 3) asks an important set of questions regarding how race is meaningful among Hispanics, and considers “Are white Hispanics treated differently from Brown or Black Hispanics?” The short answer is yes, it is, and in our own statistical findings, a graduated pattern was found among/within the Latinx community, with serious implications for how we consider anti-discrimination measures in housing to how we view it as a commodity.
Furthermore, in a collaborative work (The Color of Wealth in Miami) with Danielle Clealand, Darrick Hamiton, William Darity, Jr, Mark Paul, Anne Price, Khaing Zaw, Daniel Bustillo and Gretchen Beesing, we examined the role of wealth (a separate and more robust measure than income), in the South Florida region among local racial and ethnic groups. In accompanying ACS (American Community Survey 2013-15) descriptive statistics to the wealth-asset survey instrument, we were intentional in separating out race, ancestry and nationality to assess economic differences among local Black communities, including Afro-Latinos, African Americans and West Indian/Afro-Caribbean communities. The local “collective Black” community was uniform in demonstrating asset-level inequities compared to whites, white Latinos and Asians, but varied across asset-types and forms of debt. The more we were able to disaggregate from differently measurable categories, per Hogan’s point, the more we were able to understand how race and racism works materially across different ancestry and nationality groups.
Lastly, in a recent book (Afro-Latinos in the US Economy) co-authored with economist Michelle Holder, the first national statistical profile of the Afro-Latinx community in the field of economics, we used results from separated questions to look at those who identified as “Hispanic/Latino” and “African American/Black” and by ancestry group so as to make inter and intra-group comparisons. Again, we found consistent patterns of high or similar levels of education for Black Latinos compared to white Latinos, yet socioeconomic disparities as measured by income, earnings, employment and occupational status, home ownership, incarceration status, and other variables were demonstrated for the Afro-Latinx community.
Given that the current question arrangement provided a robust statistical technique in understanding how Afro-Latinos are treated at the everyday market/material level, I want to specifically respond to additional points Hogan makes that would compromise the ability to generate such crucial sociological understanding.
-Hogan underscores (page 3) the danger of “racializing” Latines by pointing out the term “Hispanic/Latino” is often loosely applied to the large percentage of U.S. Latinx of native/indigenous ancestry (often South and Central American native peoples). From the get-go, Hogan speaks to the potential “middle categorization” that perpetuates erasure of Latinx Afro-descendancy (and reproduces vestiges of caste-like structures where non-Black identities are socially desirable regardless of one’s phenotype). Moreover, by methodology, it invisibilizes the “street race” (see Nancy López’s extensive research) identities and experiences of U.S. Afro-Latinos. The late great Miriam Jiménez Román spoke to this in an important essay in NACLA (https://nacla.org/article/check-both-afro-latins-and-census), when she wrote:
“But even as mestizo, or mixed identity—expressed variably as raza, “rainbow people," or "mutts "—is a commonplace collective designation, Latin@s are also understood to be “of any race.” This apparent contradiction can be traced to the convergence of two seemingly distinct racial formations. On the one hand, the national ideologies of our countries of origin emphasize racial mixture and equate it with racial democracy—even as whiteness continues to be privileged, and indigenous and African ancestry are viewed as something to be overcome or ignored. On the other hand, in the United States Latin@s have been allocated an ambiguous racial middle ground that invisibilizes those too dark to conform to the mestizo ideal, while simultaneously distancing them from other communities of color, particularly African Americans.”
The second important point Hogan speaks to is that the Census proposal assumes that by making Hispanic ethnicity co-equal with race de facto racializing Latines, it will allow Hispanic/Latinos to check all the boxes appropriate that represent the group’s multitude of racial backgrounds - white/Hispanic, Black/Hispanic, Asian/Hispanic, Native/Hispanic, etc. However, this approach doesn’t consider the sociological evidence and credible fear that the combined question will reduce the number of those reporting as Black, with serious implications for law and public policy (resource distribution, voting rights, labor market anti-discrimination enforcement, etc.) - see the extensive work of Tanya Hernandez, for instance as well as recent evidence that the 2019 and 2021 American Community Survey Conducted by the Census Bureau annually are incomparable and lead to reductions in Black (Non-Hispanic alone); Native American (Non-Hispanic alone) and Black Hispanic and impede our ability to document the color line in child poverty due to the fact that national origins were placed next to each race box.
-Hogan (p. 7) points to the argument that a combined question will collect data similarly “as good” and at a “lower cost” and “burden” - while I question data collection methods based on analytical to fiscal savings, given the need for rigorous, disaggregated analysis regardless of cost as matter of public good, Hogan is correct that this assumption is premature. In my own review of the 2015 ACT, I see evidence that the combined question likely reduced the number of Latinos who reported as Black and as Hogan clearly points out: the absence of evidence is not evidence of absence. The fact that that sampling strategy for Hispanic strata did not engage in intersectional inquiry about the analytical distinction between race and ethnicity, and did not target Black Latinos in their criteria for determining the Hispanic Strata, questions were tested by respondents by panels that included mail only and internet and did not include a non-response follow-up sample, given the correlation of race, ethnicity, income/socioeconomic status and likelihood of returned response. This in the absence of strategic non-response follow-up, it default skews responders “away” from people most likely to be racialized as Black. In addition, as per John Logan’s 2010 study published in The Afro-Latin@ Reader, evident was that Hispanics/Latinos who were more likely to identify as Black lived in Northeastern states, and are more likely to identify themselves alongside (Afro)-Caribbean ancestries (Dominican, Puerto Rican, Cuban, Panamanian, etc.). The majority of Latinos who identified as “Some other race” live in Southwestern and Western states (and in spatial realities where Black Latines are less likely to live), and they represent the largest portion of the US Latinx community. By virtue of the “geographic universe” represented in the ACS, with no evidence of oversampling documented regions with larger numbers of Afro-Latinos, nor specific ancestry groups where more Black Latines are found (Puerto Ricans, Dominicans, Panamanians, etc.), it is highly likely that non-Afro Latinx statistical realities are skewing the “testing” data. There’s already evidence this is happening when as previously mentioned we compare the 2019 and 2020 ACS (see Table 1 in this crucial report).
Another assumption Hogan challenges is that this new method will lessen the number of Latines who identify as white, or that mixed-race Latines who are racialized as non-white will instead choose “Hispanic/Latino” and something else. This seems to be interpreted as a good thing and is likely based on 2015 ACT (table 7). Not only will this conflate race and ethnicity by giving us multiracial categories of people, but this ignores the fact that Latinos do not act neutral with regard to race. By moving SOR (Some other race) respondents to the Hispanic/Latino race category, it will create a kind of “statistical mestizaje.” Put differently, we are creating the likelihood that SOR becomes a more solidified substitute for Black identity, and if Latines with different experiences based on colorism are captured under the same racial categorization, we risk diluting intra-group material experiences. As argued by González, López, Karpman, Furtado, Kenney, McDaniel and O’Brien in this key Urban Institute study, to address the number of Latines that choose some other race, other racial status could be added (e.g., Brown, moreno claro/oscuro, etc.) so as to avoid the conflation of race (phenotype) with Ethnicity (cultural background) and ancestry (familial/distant heritage). Altogether, absent more nuanced alternatives like the above, we will undermine the work of advocacy groups who have been long creating Black affirmative spaces within the Latinx community (as in embracing one’s Afro-Latine racial status). I would also argue that this new method proposed by the Census will prolong the likelihood of misunderstanding how respondents may be racialized by others, in that the respondents will choose sole or combined categories they “ascribe” to, rather than what they experience at an everyday level situated in anti-Blackness. The difference between self-identification and social categorization/treatment for civil rights use must be considered by Census methodologists/enumerators.
In essence, Hogan’s analysis yields a missed point amid the current proposal, especially given the presence of what some scholars view as a Latinx “flight to whiteness” (see for instance: Darity, Hamilton and Dietrich, “Bleach in the Rainbow in Jimenez Roman and Flores, eds. 2010) whereas the difference between self-identity and social classification are underscored[1] The Census argues that if “Hispanic/Latino” is a race then respondents who choose “other” for their race will choose Hispanic/Latino instead thus providing us more accurate data. Accurate, for whom? If Hispanic/Latino is marked as a race, then those that also choose Black or White as their race will now be measured as two or more races, Black and Latino, White and Latino, etc. Given that one is an ethnicity and another is a racial category, this will dilute Latinos instead as two or more races. Put differently, if the intention is to report those who report as “Hispanic” and “Black” into “more than one race” category, this would be the most dangerous act the Census would do, given that those racialized as Black or white may be grouped into identities that do not correspond to their everyday treatment. At the least, the Census needs to assure that all levels of tabulation - the total number of those who identify as Black and Hispanic are reported individually, and in aggregate, which is still problematic because we are conflating race and ancestry (e.g., to repeat, many people of African ancestry are not racialized as Black).
In closing, the insistence that this new proposal will provide more accurate data is not supported by the literature/findings across many disciplines on how race matters with the Latinx community for interrogating social outcomes. The separate questions available for intersectional disaggregated data allowed us to capture not only the statistical economic disparities evident of a racialized group WITHIN a larger ethnic group whereas not all members are racialized equally at the material level, equally challenging the myth of a putative post-racial America. How else will we know if we have eradicated the connection between race and the color line in social inequalities? We need to be clear when we are asking about race we are asking about phenotype/colorline and street race, which is not the same as ethnicity or cultural background, and is definitely not the same as familial/distant ancestral heritage. As proposed, listing Dominican or Panamanian next to the Black box and Argentinian next to the white box is not the solution. It's a falsehood and regardless of intent, it may actually contribute to nativist/racist ontologies and policies to suggest that race and national origin ethnicity ancestry should be concordant.
Some in the Latinx civil rights community have argued that including Hispanic ethnicity as a co-equal category with race in one question format will rectify injustices like states that didn't report COVID-19 mortality for Hispanic ethnicity or as other administrative data like gun violence where Hispanic ethnicity is not reported. This is short-sighted and problematic. The reality is Hispanic ethnicity is a minimum standard and a protected category for civil rights enforcement in the current OMB guidelines. If a federal, state or local government agency or other institution violated their responsibilities in reporting Hispanic ethnicity and race as per current Office of Management and Budget (OMB) guidelines then they should be held accountable for what can be viewed as a form of statistical malfeasance. Arguments that checking two or more is a panacea are problematic because it would lead to creating de facto multiracial individuals (e.g., reported as two or more) whose racialized experiences may not be similar. Making Hispanic ethnicity a co-equal category with race is not the answer and would nullify intersectional analysis for equity use and distribution of resources and contributes to structural anti-Blackness. Let’s keep Hispanic/Latino ethnicity and race separate and avoid linking national origins to race boxes.
The author would like to thank Nancy López for her generous comments and Tanya Hernández, Danielle Clealand, Guesnerth Josue Perea and the many folks from the Afro-Latin@ Forum for their crucial advocacy.
[1] Also see “Passing on Blackness,” Darity, Hamilton and Deitrich, 2010. For a summary, see Darity, William - https://prospect.org/civil-rights/latino-flight-whiteness/.
