ClassOn the ContraryRace

How Census Data Mislead Us about Ethno-Racial Change in the United States: A Response to Mora and Rodríguez-Muñiz

I am pleased to open a conversation with G. Cristina Mora and Michael Rodríguez-Muñiz about census data and what they indicate about ethno-racial change in this issue of New Labor Forum. To forestall misunderstandings, I think it advisable at the outset to make clear the framework within which I am operating. I take it from the way that Mora and Rodríguez-Muñiz formulate their critique that their starting point is critical race theory, with its normatively inflected concerns about the deep and persisting structures of American racism and the pathways to eventual racial justice. That is fine. But I am operating from a different standpoint, that of sociological realism, which has the goal of identifying and understanding important ongoing social processes and discerning their implications. This, it should be obvious, does not mean that I am unconcerned about racial justice, just as critical race theorists generally are not unconcerned about empirical patterns and their consequences.

It does not help the conversation that Mora and Rodríguez-Muñiz tend throughout to downplay the significance of the concerns behind my analysis, which they characterize as narrowing “debates to the issue of ‘methodological accuracy’.” I find this an unfortunate attempt to reduce my argument to mainly technical issues (granted, these are part of the story); they miss that I, too, am talking ultimately about social power, even if I do not place it in the foreground in the piece I wrote for The American Prospect (it is more clear in other writings, some currently under review [1]).

As if in further challenge to my arguments, Mora and Rodríguez-Muñiz caution that “demo- graphic projections must be handled with care.” But I could not agree more; that is in fact why I wrote The American Prospect article. Has any demographic projection received more widespread credence than the one that America is heading inevitably toward a majority-minority society, in which whites will finally be reduced to numerical minority status? According to the Census Bureau, this outcome is destined to occur before the middle of this century [2]. At the end of the last century, this transition was already heralded by then-President Bill Clinton, who declared in a widely quoted 1998 speech that in fifty years America would not have a “majority race.” Since 2000, Americans have grown more confident about this outcome, supported by regular communications from the Census Bureau that seem to support it. For instance, whites are now a minority of U.S. babies, according to one Census Bureau report (the claim in my view is false, or at least deeply misleading, because the majority of U.S. babies have a white parent) [3]. The confidence that a majority-minority society is on the near horizon has produced websites advising whites how to handle this new situation, given white supremacist groups an appeal for recruiting, and perhaps spurred over-confidence among some multi-culturally disposed intellectuals.

To clarify my problems with this vision of the American future, I have to correct some impressions conveyed by Mora and Rodríguez-Muñiz. They appear to think that I am arguing that Hispanics who report their race as white on the census should be regarded as white, and so should individuals from mixed racial backgrounds (more specifically, those whose families include white and minority parentage). Neither supposition is correct. And, in the case of Latinos who check the “white” race box on the census, I agree with Mora and Rodríguez- Muñiz that the reasons for this choice are variable, and it can be a sign of resistance rather than of membership.

Yet one of the big stories of the early twenty-first century has to be the growing group of Americans, disproportionate among children and young adults because of the recent rises in the frequency of mixed unions, who come from mixed majority-minority families. Due to the limitations in census data, which are particularly severe for Latinos, this group can be most unambiguously identified among infants because one can in most cases examine the ethno-racial backgrounds of the parents (or, in the case of single-parent families, compare what the parent says about the child with what she, in the usual case, says about herself). Analyzing census data from 2013 with PhD candidates Brenden Beck and Duygu Basaran Sahin, I estimate that this group amounts to 10 to 11 percent of infants; about half are part Anglo and part Latino, a group central to the exchange with Mora and Rodríguez-Muñiz [4]. These are larger numbers than one would obtain by looking only at the reports parents provide for infants. This is true for two reasons: parents sometimes report only a single race for mixed-race children, and, more important, the ethno-racial questions on the census do not permit reports that individuals are partly Latino and partly something else. (A proposed change for the 2020 census—merging the race and Hispanic-origin questions—might ameliorate this problem.)

I contend that the group of mixed-[race]origins is of great significance for discerning ways that ethno-racial boundaries may be blurring and for contemplating the American future, at least that of the next several decades. Its members straddle these salient boundaries and possess close family ties on multiple sides of them. Their experiences can reflect a potent new form of conviviality, which brings together family members from across ethno-racial divides, at least for important symbolic occasions such as weddings and funerals. Yet these experiences may also be deeply colored by power differentials, in which the less powerful experience forms of distancing such as micro-aggressions. This seems particularly apparent in the cases of individuals who are partly black.

I acknowledge that we do not yet know a great deal about the mixed-origins group. But what evidence we have is highly suggestive: to start with, an examination (in census data) of the income and residential characteristics of the families of mixed infants indicates that, on the whole, they resemble much more all-white families than they do the all-minority families that share the same minority origin. Families that have a white mother and a black father, which make up the great majority of white-black unions, are the exception to these generalizations [5].

If we examine the adult characteristics of individuals from mixed majority-minority family backgrounds—and, admittedly, the evidence is sparser here—we find a picture consistent with integration into the white mainstream for most, with those of partly black ancestry again the prominent exception. Social identities appear to be more fluid and contingent than are the identities of individuals with unmixed backgrounds. For individuals who are partly white but not black, this fluidity often “tilts white,” in the sense that they incline more to the white side of their ancestry than to the minority side—for instance, in their sense of acceptance by others [6]. The social worlds of individuals with white and non-black minority parentage also tilt white. For example, individuals whose parentage is partly white mostly marry all-white partners. In the case of individuals who are white and either American-Indian or Asian, about 70 percent do. Even in the case of adults who are partly white and partly black, a majority takes white partners [7].

This picture of the group from mixed majority-minority backgrounds leads to the hypothesis, empirically verifiable in the future, that a white majority is likely to persist longer than the widely believed census projections indicate. Indeed, I believe that we should understand the majority-minority society the Census Bureau projects for the future also as a hypothesis, rather than a predictable certainty.

What is the rationale for my hypothesis? It does not, as Mora and Rodríguez-Muñiz seem to believe, rely on analytic reclassifications (i.e., forcing individuals of mixed background into the white category). Rather, the instability of the social identities of these individuals implies that they will appear in the census more often as white and non-Hispanic than we would expect based on current data. We have direct evidence on this point: in a meticulous analysis that takes advantage of a unique Census Bureau data set linking individual records from 2000 and 2010, sociologists Carolyn Liebler, Sonya Porter, and their colleagues show that individuals of mixed background frequently appear as mixed (or white Hispanic) on one census and unmixed—most often as white and non-Hispanic—on another [8]. Granted, we do not understand precisely the mechanisms of this instability, which could reflect shifting individual identities or the varying perceptions of other family members (those who complete the census form), but it occurs on a sizable scale. Moreover, for Mexican-Americans, who are the majority of the Hispanic population, we have in addition the finding of labor economists Brian Duncan and Stephen Trejo that, in the third generation, the identification with Mexican origins falls off sharply in accordance with the number of Mexican-born grandparents, a relationship that reflects the occurrence of inter-marriage in the second generation [9].

This identity instability is likely to produce more non-Hispanic whites in future censuses than one would expect because Census Bureau projections start with current census data. Individuals of mixed family background are over-represented among children, whose ethno- racial backgrounds are reported by parents, who in turn are highly likely to indicate mixed backgrounds for them. But the Census Bureau classifies these children then as non-white, and this exaggerates the growth of the minority portion of the population and ramifies through the projections (since non-whites beget non-whites). In criticizing these census practices, I never envisioned, despite what Mora and Rodríguez-Muñiz suggest, that these children should be reclassified as whites. Rather, I have urged the Census Bureau to abandon binary thinking, which is not appropriate for this important and growing group. Preferable, in my view, would be to treat its members as neither white nor minority; they deserve a new category. In the event, speaking strongly against viewing them as simply minority is the degree of integration with whites that they typically evince, as described above.

Let me come now briefly to the question of power, on which Mora and Rodríguez-Muñiz focus. I argue that the patterns of identity and social affiliation evidenced by individuals with mixed backgrounds are one indication that the white-dominated mainstream is expanding, taking in some individuals of minority origins. (Although space limitations prevent me from exploring historical analogies, it is worthwhile to point out that this has happened before, most notably, in the post–World War II period, when what was originally a white Christian, that is, Protestant, mainstream absorbed large numbers of Jewish and Catholic ethnics.) This expansion is not only limited to those of mixed background, but also includes many socially mobile minorities, especially from immigration backgrounds, who are entering white-dominated spaces—such as workplaces and neighborhoods—and integrating there. This is not a frictionless process—it certainly was not historically—but it is expanding and altering the mainstream, making it visibly more diverse, at least in some regions of the United States.10 Perhaps these new members of the mainstream continue to suffer some disadvantages vis-à-vis established whites, but that was true also of the white ethnics in the beginning.

This expansion is likely to have political consequences, effectively extending the political power of the white majority, if the electoral behavior of new members of the mainstream comes to resemble that of whites—determined less by minority origins than by such factors as income, education, and residential location. Based on history, this seems a plausible hypothesis [11].

In closing, I urge Mora and Rodríguez- Muñiz to rethink their argument in light of some of their own concerns. They are, for instance, concerned with the ability to use census data to measure ethno-racial disparities; so am I. Under current census practices, minority categories are increasingly heterogeneous—in the case of Latinos, census data prevent the identification and separate classification of individuals who are partly Hispanic and partly non-Hispanic white (except when they are children living with parents), and consequently, the Hispanic category includes a rising share of individuals who are integrated with whites and, in many cases, not very different from them. This heterogeneity gets in the way of measurements of disparity, which would be sharper if the portion of a group most exposed to systemic disadvantage could be identified.[12]

Their concern for social justice leads me to a different, quite political point. One of the forces that has driven our electoral politics far rightward is the anxiety many whites feel about demographic change and the prospect of losing power as they become a numerical minority in parts of the country and in the United States as a whole. Indeed, white supremacy is feeding on these anxieties, which have in turn been nourished by the Census Bureau’s projections of a majority-minority future. But the projections, I argue, are based on absurd classification decisions, such as treating children with one white and one non-white parent as non-white, as if the one-drop rule of nineteenth- century racism was still operational. It is time for census data—and other data, for that matter—to reflect more accurately the social mixing across the white-minority boundaries that is taking place on the ground in many parts of the United States. Whether greater accuracy in this respect will allay the concerns of whites, I cannot say. I have no doubt, however, that the current distortions in demographic data contribute to the toxic miasma in our political discourse enveloping questions of diversity.


  1. Richard Alba, “The Likely Persistence of White Majority,” The American Prospect,  January 2016. Cf. Richard Alba, Brenden Beck, and Duygu Basaran Sahin, “The American Mainstream Expands—Again,” under review; Richard Alba and Jan-Willem Duyvendak, “What about the Mainstream? Assimilation in Super-Diverse Times,” under review; Richard Alba, “The US Is Becoming More Racially Diverse. But Democrats May Not Benefit.” The Monkey Cage, The Washington Post, January 6, 2017.
  1. Sandra Colby and Jennifer Ortman, “Projections of the Size and Composition of the U.S. Population: 2014 to 2060” (current population report, U.S. Census Bureau, Washington, DC, 2015).
  2. D’Vera Cohn, “It’s Official: Minority Babies Are the Majority among the Nation’s Infants, but Only Just” (Washington, DC: Pew Research Center, 2016); cf. National Vital Statistics Report, “Births: Preliminary Data for 2015” (Atlanta, GA: National Center for Health Statistics, 2016).
  3. Alba, “The Likely Persistence of White Majority.”
  4. Alba, Beck, and Sahin, “The American Mainstream Expands—Again.”
  5. Jennifer Lee and Frank Bean, The Diversity Paradox: Immigration and the Color Line in Twenty-First Century America (New York: Russell Sage, 2010); Pew Research Center, Multiracial in America: Proud, Diverse and Growing in Numbers (Washington, DC, 2015); Edward Telles and Vilma Ortiz, Generations of Exclusion: Mexican Americans, Assimilation, and Race (New York: Russell Sage, 2008).
  6. Michael Miyawaki, “Expanding Boundaries of Whiteness? A Look at the Marital Patterns of Part-White Multiracial Groups,” Sociological Forum 30, no. 4 (2015): 995-1016; Telles and Ortiz, Generations of Exclusion.
  7. Carolyn Liebler, Sonja Rastogi, Leticia Fernandez, James Noon, and Sharon Ennis, “America’s Churning Races: Racial and Ethnic Response Changes between Census 2000 and the 2010 Census” (Center for Administrative Records Research and Applications, U.S. Census Bureau, Washington, DC, 2014).
  8. Brian Duncan and Stephen Trejo, “Intermarriage and the Intergenerational Transmission of Ethnic Identity and Human Capital for Mexican Americans,” Journal of Labor Economics 29, no. 2 (2011): 195-227.
  9. Alba, Beck, and Sahin, “The American Mainstream Expands—Again.”

11. Alba, “The US Is Becoming More Racially Diverse.”

12. Richard Alba, Tomás Jiménez, and Helen Marrow, “Mexican Americans as a Paradigm for Contemporary Intergroup Heterogeneity,” Ethnic and Racial Studies 37 (2014): 446-66.