Rurals of Nevada: Race, the 2030 Census, and the Three Rurals (art. 13)
"But comparing race data from the 2020 census with data from earlier counts can be a bit like comparing 'apples and oranges'." - Hansi Lo Wang and Ruth Talbot, NPR, 22 August 2021
The reference to the 2030 Census in the title is not a typo. On Friday, 27 January 2023, the Office of Management and Budget (OMB) published in the Federal Register an “Initial Proposal for Updating OMB’s Race and Ethnicity Statistical Standards.” In short, OMB wants to revise Statistical Policy Directive No. 15 (SPD 15), the 1997 standard for racial and ethnic standards used for all federal agencies, including the U.S. Census Bureau. The proposal is a call for public comment, but the changes will have a significant impact on future statistic collections—including the 2030 Census.
The SPD 15 policies originally were proposed in 1977 and first used in the 1980 Census. The intent was to create consistency across all federal agencies for reporting purposes—particularly in relation to the numerous civil rights laws passed in the 1960s and 1970s. Rather than having each agency develop its own definitions and reporting mechanism, SPD 15 created a standard list used by all federal agencies. It is from SPD 15 that the racial “checkboxes” on various federal forms originated. The directive has been revised once, in 1997, to further refine the Hispanic/Latino ethnicity question (into a separate category), to allow the selection of multiple racial categories, and to add a “Some Other Race” option (which became required only for the decennial census and the American Community Survey). The 1997 OMB Standards, as they became known, are the current reporting standard used in the 2020 Census.
Anyone who has spent any time studying the Census knows the convoluted and contentious history of trying to establish race metrics. I discussed a bit of this history in the first article on Native American communities in Nevada. Anna Brown of the Pew Research Center has a short, broader history of the Census race categories here. Given that race is a social and political construct, it is not too surprising that there has been significant change since the 1790s as American demography has changed.
It is not surprising the Census Bureau’s race categories have changed over the decades. But more often than not, the change has been in attempting to address contemporary issues than establish a neutral overview of American society.
Yet the change has not been consistently towards some model of total measurement, but rather addressing contemporary issues. The decision to add Native Americans to the 1860 Census, a Chinese race to the 1870 Census, and a Mexican race to the 1930 Census (discontinued until 1970) are examples. Even the “Other Race” category, introduced in the 1910 Census, initially was used primarily by Koreans, Filipinos, and other non-Chinese Asians who might have been exempted from the discriminatory Chinese Exclusion Act of 1882.
But in reality, much of American demographic thinking historically has been concerned with demarcating white and black populations. Over the decades, the Census devoted significant resources to most clearly defining African-American populations, most notably the 1850 and 1890 efforts to utilize “gradations” such as “mulatto” or “quadroon”—only to be dropped with the implementation of the “one-drop rule” in the 1930 Census. This fixation in part explains why the Census Bureau’s efforts to define other groups have often been haphazard until the 1970s and the increased activism for civil rights outside the African-American community.
It also explains another tension with the Census data: the shrinkage of the white population. The 2020 Census results showing a decline have been criticized in part as the result of changes in definitions, particularly the increased use of multi-racial options which may make the “white decline” appear more pronounced (see the NPR story, quoted at the top of the page). As sociologist Ann Morning notes in the article, it should be expected that any changes to the Census race categories will result in variations to the white population, since it is the one category consistently used since 1790. But this also makes it very difficult to compare categories across various census datasets. There likely is no good solution at the moment, but we can be assured the media at least will continue to obsess over the “white decline” issue (see this recent article from NPR).
So, with this convoluted history in mind, what changes in the new standards, and what might they reflect about contemporary American concerns? They closely resemble a 2017 proposal from the Census Bureau which was too late to be considered for the 2020 Census. Three fundamental revisions are being proposed:
Create a Single Question that Collects Race and Ethnicity Data: This change would do away with the current separate “Hispanic/Latino” question and move the category to the main race block. This issue is intended to address the increasing tendency of those who select “Hispanic/Latino” not to answer the race question (apparently considering it as already answered).
Create a Middle Eastern/North African (MENA) Category: Over the last 20 years, there have been repeated calls for a separate category. While strongly supported by members of Congress such as Rep. Rashida Talib (D-MI 12) and some activist groups, there is a long history (including a number of admittedly dated court cases) of members of this group resisting being classified in a separate category.
Detailed Race and Ethnicity Data Collected by Default: Finally, the revisions would basically require not only the collection of the racial category but also specific ethnicity/country of origin. Up until now, this had been an option. Including it back in can help address not only the issues of AIAN tribal identity and more detailed Hispanic/Latino data (for example, Mexican versus Cuban versus Venezuelan), but also differentiate African-American descendants of enslaved persons from groups such as Nigerians or Somalis. This will be the most expensive change since many forms which previously required just the race categories will have to be changed.
The end result will likely be a new “block” of seven core races with common ethnicity checkboxes and write-in spaces under each. One possibility presented in the Federal Register notice gives an idea of what the 2030 Census questionnaire could ask:
One item to note is the absence of a “Some Other Race” category. Although not stated directly, the category is the root of many of the issues in understanding the Hispanic/Latino community, one of the primary drivers of these proposed changes. The shift to adding a separate Hispanic/Latino racial category would encourage (it is hoped) more “complete” racial information. For example, one concern is the extent to which just answering Hispanic/Latino and skipping race or answering “Other” means Hispanics/Latinos descended from Caribbean enslaved populations might be undercounted in the African-American populations. Of course, the assumption is that it is primarily Hispanics/Latinos that choose “Some Other Race,” which may be problematic in and of itself.
For Nevada generally and the Three Rurals of Nevada in particular, the most impact of these new proposed standards will be the move to treating Hispanics/Latinos as a single racial category rather than an ethnicity. So to examine the impact of these proposed changes, the relationship between Hispanic/Latino identity and Some Other Race identity is a good place to start.
Hispanics, Latinos, and Some Other Race
The terms “Hispanic” and “Latino” are complicated both because they are relatively new, dating from the 1970s, and also somewhat vague in definition. The original 1976 Congressional law, “Relating to the publication of economic and social statistics for Americans of Spanish origin or descent” (Public Law 94-311), emphasized the centrality of the Spanish language and descent from “Mexico, Puerto Rico, Cuba, Central and South America, and other Spanish-speaking countries.” This law was the direct impetus for the original SPD-15 model.
The emphasis on the Spanish language was intentional. The law was passed in response to growing civic activism from widely different communities ranging from Mexican immigrants in the Southwest to Cuban refugees in Florida to multi-generational Puerto Ricans in New York. At the same time this political activism was happening, the economic power of these communities was growing. Businesses and emerging media companies such as Univision were looking for a means to reach these markets. The creation of a Spanish-language identity under the umbrella of “Hispanic” appeared to fit the need of all these groups, as this article explores.
While the Hispanic/Latino definition was initially based on Spansih language and culture, it has become a problematic construct.
But, as always, the details became complicated. The term “Hispanic” brought a range of complications. First, it was more used and accepted on the east coast. In the west United States, the term “Latino” was more common, because it more accurately reflected the origin of this population among Latin American countries. There was also the issue of the term “Hispanic” in many Latin American countries being used specifically for those people of direct European descent as opposed to those descended from indigenous communities even if they spoke Spanish. So some combination of “Hispanic or Latino” came to be used to reach both groups.
But note the emphasis on Spanish adds an additional wrinkle which has recently blown up publicly. Technically, people from Spain would be counted as “Hispanic/Latino” under the prevailing definition. Hence the scandal in 2020 when the Oscars deemed Spanish-born and raised Antonio Banderas as a “person of color” when he was nominated for Best Actor (read the Los Angeles Times’ take here). The problem is that the combination term “Hispanic/Latino” hides what many consider a vital distinction. Banderas may be Hispanic, but he is no Latino.
But even if we accept people from Spain as “Hispanic/Latino,” note that it technically would exclude Brazilians, since the primary language is Portuguese. So Sérgio Mendes, for one, would not be considered Hispanic/Latino (seja ainda meu coração amante da bossa nova!); neither would sixth-generation Brazilian Gisele Bündchen. Of course, others argue that the term “Latino” should embrace Brazilians as well—but not necessarily people from Spain. So Mendes and Bündchen would be Latino, but not Banderas.
This inexactitude has led the Pew Research Center and others to attempt to determine how this population sees itself. The results are . . . unclear. For instance, a 2019 survey by the Pew Research Center found that nearly half (47%) of Hispanics/Latinos found the identity stemming from their families country of origin, and only 17% said in a 2015 survey that the identity is primarily racial (the results of these and other surveys are summarized in Pew Research’s “Who is Hispanic?” article here). Moreover, their research indicates that the Hispanic/Latino self-identification declines across generations—although I view the data as more questionable here.
So where are we now? The current standards (the 1997 OMB Standards) defines Hispanic (which is short-hand for “Hispanic/Latino” as “persons who trace their origin or descent to Mexico, Puerto Rico, Cuba, Central and South America, and other Spanish cultures.” Except for the absence of an explicit reference to the Spanish language, this is not much different from the 1976 definition. It certainly is not going to resolve the issues which the Pew Research Center has identified. It might help with one issue: the Census Bureau’s use of the two-tiered question to resolve identification. According to Pew Research, someone who checks “Hispanic/Latino” but then only includes a “non-Spanish” ancestry (the example they use is Irish) would be recorded as “non-Hispanic/Latino.” Likewise, a person who selects “non-Hispanic/Latino” but then lists, for example, a “Mexican” ancestry would be classified as Hispanic/Latino. The Pew Research Center offers a handy flowchart for this here. Although Pew Research estimates the impact of this at around 1% of the population, it seems to bring the schema into question.
The more interesting part of the question is the association between Hispanic/Latino identity and the “Some Other Race” category which has arisen. Despite the earlier surveys undertaken by the Pew Research Center, a significant proportion of Nevada’s Hispanic/Latino population did identify themselves as “Some Other Race” in the 2020 Census (as they did elsewhere in the country). The chart below shows a resounding correlation in Nevada county populations percentages that identify as Hispanic/Latino (in green) and as Some Other Race (in orange).
For people not identifying as Hispanic or Latino, the proportion choosing “Some Other Race” drops dramatically. With the exception of Humboldt County (at 1.6%), no county in Nevada has a non-Hispanic/Latino population that chose “Some Other Race” alone at more than 1%, and none has the non-Hispanic/Latino population identifying Some Other Race alone or in combination at more than 2.9%. The three urban counties and the fourteen rural counties are very similar in this instance.
It appears then that self-identification as Hispanic/Latino is closely associated in Nevada with Some Other Race populations. So I suspect that the impact of moving the Hispanic/Latino question from separate ethnic to a racial category is not going to have a significant impact for the Three Rurals, and likely only a minor impact for the urbans. The current system which conflates the “Hispanic” and “Latino” will continue and some of the complexity found within this population will be lost if a race category alone is considered.
But there is a consequence which may impact other populations. There are groups that are neither Hispanic nor Latino which also use “Some Other Race.” In the 2020 Census in Nevada, over 6% of those who chose “Some Other Race” alone or in some combination identified as not Hispanic/Latino. I suspect that many of this group belong to various communities such as Filipinos or those from India which are currently captured under an “Asian” category which has also attracted some controversy recently. Remember that it was Koreans and Filipinos who tended to use the “Other” category in the 1910 Census.
Would the proposed removal of the “Some Other Race” category to encourage a better understanding of the Hispanic/Latino population result in these groups being “hidden” in the population as well? That is a question that remains to be answered.
Hispanic/Latino Population in the Three Rurals
In earlier articles, such as the Population Under 18 in the Aging Rurals series, I looked at the question of Hispanic/Latino populations in rural Nevada. My analyses of this population were in connection with other metrics such as age or the Diversity Index. Time to back up and look specifically at the rural Hispanic/Latino population more directly.
Collectively, the 14 rural counties have a population which is 17.6% Hispanic/Latino, according to the 2020 Census (Table P2). This population is less than 6% of Nevada’s total Hispanic/Latino population, which is heavily centered in the three urban counties of Clark, Washoe, and Carson City. But, as those readers playing along at home know, this hides a rather stark variation across the state that is hidden by the use of the singular “rural Nevada” model.
The table below provides the Hispanic/Latino population in Nevada divided along the major regions we have been discussing, as well as some additional data:
The main item which stands out to me in this data is that the I-80 Corridor counties look much more similar to the three urban counties (Clark, Washoe, and Carson City) than to the other two rural regions. In fact, the five counties in Nevada which have the highest Hispanic/Latino population are:
Clark: 31.0%
Humboldt: 26.1%
Carson City: 25.4%
Elko: 25.3%
Washoe: 25.1%
In addition, Lander and Pershing Counties (both in the I-80 Corridor) are the only two other counties in Nevada with more than 20% of the population identifying as Hispanic/Latino. As I have argued previously, I suspect this difference is the result of the more dynamic, expanding economies of the I-80 Corridor counties compared to the other rural counties.
But there are some significant differences that mark the urban and rural Hispanic/Latino populations as distinct. These show up in two columns to the chart, looking the percent of the population which was foreign-born and the percentage of the population which speak Spanish primarily at home. These statistics are drawn from the 2021 American Community Survey (ACS) 5-Year Estimates, so they are not entirely comparable to the 2020 Census data and remain based on samples rather than direct measures. However, I believe they are representative enough to highlight some key dynamics.
The foreign-born percentage of the population is probably the most interesting, especially since a lot of current political discourse concerns Latin American immigration. While the foreign-born population includes all races and ethnicities, Hispanics/Latinos comprise the largest recent immigrant group in Nevada and “Latin America” as the place of birth is by far the highest number of those foreign-born in all Nevada counties. Even with the sampling errors of the ACS (and remember that the ACS asks about citizenship, unlike the 2020 decennial Census), this is a good proxy for immigration.
What is interesting is that the percentage of foreign-born compared to the Hispanic/Latino population generally. In the urban counties, the percentage is relatively high, indicating a significant percentage (over half) of the Hispanic/Latino population having been born elsewhere. But in the rurals, this percentage drops to almost a third—even in the growing I-80 Corridor counties (although the percentages are higher in Pershing and Humboldt Counties, perhaps reflecting a changing pattern). It appears that direct foreign immigration is a less significant factor in the population growth here.
The percentage of the population speaking Spanish at home, however, remains relatively high across the state. Only the Western Rural and I-80 Corridor regions are below 60% of the Hispanic/Latino population and even they are above 50%. These are relatively high numbers for a population with relatively small percentages of foreign-born.
I think what this data shows is a slightly longer-established Hispanic/Latino community in the I-80 Corridor—but not one which has totally assimilated. I suspect (and I might dive into this in future articles) that the Hispanic/Latino communities in the I-80 Corridor especially probably represent 2nd and 3rd generation families where the children have been born here but parents and grandparents continue to speak Spanish. For the Western Rurals, this population might be even older.
But if this analysis is correct, I think it indicates something else: a dynamism in the economy that is attracting long-term family settlement.
Conclusion
My primary question on the proposed changes is the same one I had for the diversity index: what does this tell us in terms of policy planning? While civil rights legislation that drove the original SDP 15 process in the late 1970s was driven by voting concerns, many of our contemporary social issues require a much finer distinction between the experiences of communities than can be handled by broad racial categories. Complexity in problems requires complexity in the data for solutions—and broad categorizations are horrible for complexity and can lead to inefficient policies.
We can take one example policy: English-as-a-Second-Language programs. Under the current Pupil-Centric Funding regime in Nevada, additional funds are awarded based on the need for English-language acquisition. Can better tracking of Hispanic/Latino populations provide some guidance here? Perhaps—but children eligible for the program would be tested for placement anyway. Moreover, older Hispanic/Latino populations may not correlate as closely with this paradigm anyway. Moreover, emerging immigrant groups which may require increased ESL funds such as Haïtians, Afghanis, or Somalis are not Hispanic/Latino and are poorly tracked by current racial categories.
Similar arguments can be found in a range of policies. For instance, even before 9/11 there was interest in tracking anti-Muslim discrimination and the creation of a Middle Eastern/North African (MENA) racial category as a means to do so (as discussed in the Federal Register as part of the 1997 revision or this NBC article). But can this category help identify anti-Muslim discrimination directed against Somalis (Black of African-American), Pakistanis and Indonesians (both Asian), or even separate anti-Muslim from anti-Jewish discrimination (both are in the proposed MENA category)? Again, the specifics needed are embedded below the racial category level.
In my view, the real benefit of the proposed changes to the OMB standards may lie not in the revision of racial categories but the obligatory collection of place-of-origin/ethnicity data. The specifics of whether someone is from Venezeula or Afghanistan or Haïti or Somalia provide far more information than racial categories can provide. Better use of more complex data will lead to better outcomes.
In any case, the whole purpose of the OMB notice in the Federal Register is that the new proposed scheme is open for public comment. Until 14 April 2023, you can submit your comments directly here at the Federal Register or even read the submitted comments.