The Likely Persistence of a White Majority

The Likely Persistence of a White Majority

How Census Bureau statistics have misled thinking about the American future. 

January 11, 2016

You may also like

This article appears in the Winter 2016 issue of The American Prospect magazineSubscribe here.

Has the notion of demography as destiny ever enjoyed so much credence? The disappearance of a white majority in the United States by the middle of this century is now widely accepted as if it were an established fact. Projections by the Census Bureau have encouraged those expectations, and people on both the right and left have seized on them in support of their views. On the right, the anxieties about the end of white majority status have fueled a conservative backlash against the growing diversity of the country. On the left, many progressives anticipate an inexorable change in the ethno-racial power hierarchy. Numerous sites on the web offer advice and counsel on how whites can handle their imminent minority status.

But what if these different reactions are based on a false premise—actually two false premises? The first stems from the Census Bureau’s way of classifying people by ethnicity and race, which produces the smallest possible estimate of the size of the non-Hispanic white population. Whenever there is ambiguity about ethno-racial identity, the statistics publicized by the bureau count an individual as minority. This statistical choice is particularly important for population projections because of the growing number of children from mixed families, most of whom have one white parent and one from a minority group. In the Census Bureau’s projections, children with one Hispanic, Asian, or black parent are counted as minority (that is, as Hispanic or nonwhite). The United States has historically followed a “one-drop” rule in classifying people with any black ancestry as black. The census projections, in effect, extend the one-drop rule to the descendants of other mixed families. A great deal of evidence shows, however, that many children growing up today in mixed families are integrating into a still largely white mainstream society and likely to think of themselves as part of that mainstream, rather than as minorities excluded from it.

Under alternative ways of counting, the potential range of variation in the size of the white population is quite large. In unpublicized tables, the Census Bureau itself provides a measure of how wide that variation is. If we were to go to the opposite extreme from the bureau’s official projections and adopt a white one-drop rule—that is, to classify anyone with some white ancestry as white—the data show that whites would make up three-quarters of the population at mid-century, when the Census publicly claims that whites will be in the minority. To be sure, neither extreme is credible; the white share of the population will lie somewhere in between these poles.

A second reason to be skeptical about the excited talk about the end of a white majority is that it ignores the potential for blurring the boundary between mainstream and minority. The United States has previously seen excluded minorities such as the Irish, Italians, and Jews assimilate into the mainstream. Although the channels of assimilation are narrower today because of heightened inequality, many recent immigrant families seem to be on the same path as their predecessors. The likely result will be to enlarge the mainstream and alter the circumstances under which individuals are seen as belonging to marginalized minorities.

To raise these questions is not to minimize the significance of the growing diversity of American society, stimulated above all by the mass immigration since the late 1960s. Nor is it to deny the loss by whites of majority status in many parts of the United States, such as California. But it is to peer beneath the surface of census data to the little-understood processes behind their construction and to correct distortions of social reality that the official population statistics encourage. And it is to suggest that longstanding processes of assimilation could produce a white-dominated mainstream at the national level and in many regions for the foreseeable future.


The Census and Children of Mixed Families

Population projections, whether by the Census Bureau or anyone else, are not properly understood as forecasts. As any demography textbook can tell you, a projection is a numerical exercise for working out the implications of a set of assumptions about demographic variables such as birth and death rates. Over an extended period, the assumptions are virtually certain to diverge from demographic realities. Nonetheless, the Census Bureau and, more recently, the Pew Research Center have not hesitated to project an exact year when whites will lose majority status—2044 and 2055, respectively—encouraging the popular view that they are making forecasts.

Projected changes in the racial and ethnic composition of a society also depend on the rules for classifying people. The Census Bureau assumes that every individual is either white or minority and resolves the complexities of multiple heritages by assigning people with any mixed background to the minority side of the white/nonwhite divide. Individuals with both Hispanic and non-Hispanic ancestries cannot even be identified as such in census data; according to the bureau’s rules, they are only Hispanic. Since 2000, individuals with a mixed racial heritage can claim it on census forms, but they are, without exception, considered minorities in population statistics. The Census Bureau’s report on its latest projections declares that a minority group is “any group other than non-Hispanic white alone.” Most readers probably cannot decode the full implications of this formulation.

The bias in the Census Bureau’s practices has its largest impact in estimating the ethno-racial backgrounds of children, because it is among children that we first see the consequences of the rapidly rising number of mixed families. To estimate how many children are from mixed families, I used data for 2013 on the families of infants from the American Community Survey, which is conducted by the Census Bureau. (I focused on infants because the chance of obtaining data on both parents is greatest when a child has recently been born, and we need to look at parents’ own reports to avoid the biases inherent in the census data for the children themselves.) Although the Census Bureau declared in 2012 that nonwhite births for the first time outnumbered white ones, 60 percent of the 2013 infants have a white parent. About 10 percent, then, have both a white and minority parent. These infants are counted as minorities in census statistics. And they are regarded as permanently so in census projections.

Overall, about one of every seven infants comes now from an ethno-racially mixed family. The largest group by far of these infants, as the pie chart shows, consists of those with one non-Hispanic white and one Hispanic parent. They are nearly 40 percent of the total. Other sizable groups are mixed white and Asian, white and black, and white and mixed-race parentage.

Social-science research on children from mixed families is limited. But three kinds of evidence, in concert, indicate persuasively that the Census Bureau data exaggerate the decline of the white population by failing to take into account that many children from mixed backgrounds will likely be integrated into largely white social milieus and identify, at least some of the time, as white. This evidence involves the incomes of mixed families, the social identities of individuals with mixed backgrounds, and their marriage patterns.

Income of mixed families: The household incomes of many mixed families indicate that they are closer socioeconomically to mainstream white families than to disadvantaged minority ones.

As the chart above shows, unmixed Asian families have higher median incomes than unmixed white families, and mixed Asian and white families have the highest incomes of all. The incomes of mixed Hispanic and white families are also very different from those of unmixed Hispanic parents. When the non-Hispanic white parent is the father (true for about half of the white-Hispanic mixed infants), the median family income is scarcely any different from that of white-only families. When the Hispanic parent is the father, the average income is lower but still much higher than is typical for Hispanic-only families.

Since income is a primary determinant of where families reside, these patterns imply that many of the mixed white and minority children are growing up in neighborhoods where many whites also reside, and outside of areas of minority concentration. These ethno-racially mixed children will have peers from white families and likely learn to get on with them from an early age.

There is one major exception to the pattern in family incomes. Although white-black families have incomes close to those of whites when the father is white, that is not true when the father is black, which is the more common situation in black-white families. The infants with mixed, only-minority parentage also come predominantly from families with incomes between those of black-only and white-only families.

Social identities: Unlike census data, data from a 2015 survey by the Pew Research Center illuminate the feelings and experiences of individuals with mixed backgrounds. According to “Multiracial in America,” the Pew report based on the survey, most Americans from mixed backgrounds do not think of themselves as multiracial. For those who are white and Asian, the affinities with the white group are strong. By a two-to-one margin, they say that they have more in common with whites than with Asians. They report, in addition, feeling more accepted by whites and having more white than Asian friends. Those who are white and black exhibit a very different profile. The majority believe that others see them as black. They also have much closer ties to their black relatives and are very likely to report encountering discrimination, including being “unfairly stopped by the police.”

The Pew survey found many more multiracial Americans than the census does, which implies that adults with mixed backgrounds often appear in the census in single-group categories (that is, as unmixed white, black, and so on). In fact, an internal Census study, “America’s Churning Races,” shows that large numbers of people who identify themselves as multiracial in one census identify themselves as white in another.

For instance, of those who reported a mixture of Asian and white parentage in either 2000 or 2010, 36 percent appeared as only white in the other census (and 22 percent as only Asian). Those who report mixed white and Hispanic family backgrounds to the census are counted as Hispanics of white race, a huge group that includes about half of all the nation’s Hispanics. Yet, of the individuals who appeared in this group in either 2000 or 2010, 12 percent said they were non-Hispanic and white in the other year. In other words, many Americans with mixed Asian or Hispanic family origins identify with the white majority some of the time.

AP Photo/Idaho Press-Tribune, Charlie Litchfield

In this April 1, 2010 photo, a copy of a 2010 Census form is shown at a Census Day event at the Caldwell Housing Authority in Caldwell, Idaho. 

The one exception to this pattern of “leaning” white involves individuals of mixed white and African American heritages, who are much more likely to indicate that they are only black than only white (33 percent versus 16 percent). This exception conforms to the consistent research finding that Americans with visible African ancestry confront more virulent everyday prejudice and discrimination than other minorities do. The one-drop rule appears not to have lost its power in their case.

Intermarriage: Finally, there are the marriage patterns associated with mixed backgrounds, which are indicative of the social milieus into which individuals have been integrated. For individuals who are partly white and partly minority, the likelihood of choosing a white spouse is much higher than it is for those with the same minority ancestry only.

A powerful demonstration of this pattern comes from a unique study of Mexican Americans, which followed families from 1965, when the original survey was conducted, to the late 1990s, when researchers tracked down the original participants who were still alive, as well as their children. In their book, Generations of Exclusion, based on these data, the sociologists Edward Telles and Vilma Ortiz report that the odds of intermarriage were five times higher for the children of intermarriages than for those from Mexican-only backgrounds. These intermarriages were overwhelmingly with non-Hispanic whites. Scholars of intermarriage have also found higher rates of marriage with whites among individuals who are mixed white and Asian compared with those who are Asian only. In sum, many partly white adults appear to have been integrated into largely white social worlds.

These data about family income, social identity, and intermarriage raise serious questions about the Census Bureau’s practice of counting children of mixed families as members of ethno-racial minorities. In collecting data, the Census relies on what parents say about their children’s ethno-racial backgrounds. Parents are very likely in doing so to try to honor both sides of their offspring’s family origins. When their children grow up, however, many of them may view themselves as whites. Future white counts are therefore likely to be substantially larger than one would predict from current census data.


Assimilation into the Mainstream

Children from mixed backgrounds are only one aspect of a broader social process under way that is mixing together different groups in American society. Multicultural critics of assimilation have rejected it as a goal, but the concept remains essential for sociological analysis if we are to understand important changes taking place. Some people from minority as well as mixed backgrounds are being attracted into a still heavily white mainstream, changing the mainstream even as it continues to be dominated by whites. 

The mainstream, of course, is not the whole of American society. Rather, it is the part that mistakes itself for the whole. In a society where racial and ethnic origins historically have confined Americans to different social strata, the mainstream has been long associated with the social spaces and cultural practices of white Americans. That is now changing as the boundaries of the mainstream expand.

One momentous change involves the rapidly growing presence of Americans from recent waves of immigration at the top of the U.S. workforce, in domains that were previously monopolized by native whites. Guillermo Yrizar Barbosa and I have analyzed this change in an article that appears on the website of the journal Ethnic and Racial Studies. We focus especially on the upper quarter of all workers who hold top occupations as defined by annual earnings. Until recently, 85 percent to 90 percent of these workers were whites (and non-Hispanic). But among the young workers who have entered this tier since about 2000, the share represented by whites has dipped below 70 percent. The Great Recession did not reverse this trend. Since the socioeconomic ascent by minorities closely correlates with the declining white share of young adults, it is a safe prediction that it will continue as ethno-racial diversity rises among youth.

Ted Cruz Campaign

A Minority Majority? The Census Bureau classifies children of mixed marriags as members of the "minority" population. Under those definitions, three of the four people in this picture—Senator Ted Cruz and the two children he has had with his wife Heidi Nelson Cruz—are part of the minority population that the Census projects to be a majority in 2044. In popular media, this projection is routinely equated with the end of a white majority and the emerging majority of "people of color." Cruz himself is the son of a mother of Irish and Italian extraction and a father born in Cuba. 

The expanding groups in the upper ranks of the workforce are Asians, both immigrant and U.S.-born, and U.S.-born Latinos. Black Americans are also increasing their numbers, but not to the same extent as the others.

The growing diversity at or near the top of the occupational ladder does not mean that whites on these rungs have lost all their advantages, at least not yet. With educational level taken into account, we found that whites generally are better placed occupationally than minorities, and when compared with minorities in the same occupation, they earn more on average. White advantages could decline as the numbers of individuals from minority backgrounds increase in the top tiers, bringing more of them into positions of authority where they can make decisions about hiring and promotions. But it is premature to predict how the struggles of whites to hold onto their diminishing advantages will turn out.

Individuals from minority backgrounds who hold prestigious and visible posts in the workforce, or positions of civic leadership, are part of the mainstream, in any sensible definition of it. Their ascent echoes an earlier transformative moment in our history. In the quarter-century after World War II, the mainstream was joined en masse by the descendants of Irish and southern and eastern European immigrants, Jews, Catholics, and Orthodox Christians. Assimilation today is more selective, not as massive. But as in the earlier period, when hyphenated identities became acceptable, it does not require the obliteration of ethnic and racial identities, just their muting, to allow individuals to function in social worlds that, while increasingly diverse, are still home to many whites. The motive for assimilation is, as before, to gain for oneself and one’s children access to the greater opportunities that are available in mainstream settings.

This assimilation should disabuse us of the fantasy of the imminent demise of the white majority and its loss of power. Not all the newcomers to the mainstream will identify as whites, and its visibly growing diversity will be a key development of the early 21st century, as the election of the nation’s first black president unmistakably signals. “Whiteness,” however, has never been fixed; it is a malleable concept, and it is on its way to changing again, as it has before.

Yet it is critical not to lapse immediately into another fantasy, namely, the belief that assimilation will prove a panacea for still-glaring ethno-racial disparities. Contemporary assimilation is simply not on the same scale as that of the mid-20th century, when, for example, Italians caught up to other whites in education and socioeconomic attainment in just a 25-year period after World War II. Assimilation today is crimped by greatly heightened inequalities and is leaving many outside its reach, including many Hispanics, such as the undocumented and their children, even those who are U.S. citizens because they were born here. In one respect, however, the earlier and current patterns of assimilation are similar: African Americans are participating only to a limited extent. Indeed, one could even say they are being bypassed.

To think clearly about the American future, we need not only the right concepts but also accurate data. The Census Bureau, the public agency we all rely on for neutral representations of social realities, is failing us. Not only do its rigid and illogical classifications distort important new realities, the bureau is also not forthcoming about the errors and uncertainties involved. Instead, it continues to promulgate “firsts”—in June, it declared that for the first time minorities are the majority of children under the age of 5—as if the data were unimpeachable. Given the political resonance of its statistics, which reverberate on the right and left of the spectrum, there is not a moment to lose in demanding that, in its official projections and pronouncements, the Census present a more nuanced view of the nation’s demographic future and acknowledge the alternative ways in which Americans may come to think about themselves.