Introduction: Immigration and Changing Identities

In the last half century, the United States has undergone a profound demographic transformation in the wake of a massive inflow of immigrants. In 2016, immigrants represented approximately 14 percent of the U.S. population; together with their U.S.-born children the figure was more than 25 percent, a remarkable eighty-six million people. This growth in immigration, mainly from Latin America, Asia, and the Caribbean, has altered the racial and ethnic composition of the nation. The non-Hispanic white population in the United States declined from 83 to 62 percent between 1970 and 2015, and the Hispanic population grew from 4 to 18 percent in the same period. Asians, fewer than 1 percent of the U.S. population in 1970, are now 6 percent. The number of black immigrants (from Africa and the Caribbean) has also increased, and close to 10 percent of blacks in the United States are now foreign born. The result has been greater racial and ethnic diversity in a wide swath of both urban and rural neighborhoods across the country. While policymakers deliberate about controlling future immigration or dealing with those who are already here (particularly those who lack documentation), the reality of a changed ethnic-racial population plays out in the lives of millions of Americans, immigrants and nonimmigrants alike.

In distinction from earlier immigration issues in RSF that have highlighted questions of legal status (Gonzales and Rafael 2017) and political representation (McCann and Jones-Correa 2016), we examine fundamental issues of identity definition and group categorization with a focus on the U.S. context: how immigrants, the children of immigrants, and long-established Americans see themselves in terms of race and ethnicity, as well as how members of each group view and label the others. What has increased diversity meant for the way people define their ethnicity and that of others? How do these conceptions influence the ways in which members of different ethnic groups interact with one another? To put it another way, how has the massive immigration inflow of the last half century altered the social and psychological reality that people experience in the United States today?

In exploring these questions, this introduction brings together insights and perspectives from a variety of social science disciplines, from social psychology to demography to social history. Questions about identity and identity change, intergroup relations, and indeed the field of immigration itself, cannot be relegated to any single discipline. We firmly believe that a multidisciplinary strategy is the most productive choice. Some topics and questions do of course fall more easily into one domain than another. But we have consistently tried to link the domains and broaden the perspective, in the hope that readers from each discipline will find both the familiar and the new.

We begin by addressing identity, a concept that can be defined in terms of basic cognitive and social psychological processes, but at the same time depends on social norms, accepted demographic classifications, and historical developments. Later, in discussing the historical context for identity development and change, we draw heavily on work from history and sociology, at the same time pointing to processes that may have implications for individual functioning. In examining changing identities and intergroup relations, we again bring in work from several disciplines, including demographic and sociological analyses of intermarriage, social psychological work on identity and intergroup contact, and the work of sociologists and political scientists on panethnicity. Finally, we take stock of where we are and we look ahead to possible future developments. How might immigration, ethnic and racial identity, and relations between diverse ethnic and racial groups continue to develop and change? What additional factors need to be taken into account and what new research is needed?

IDENTITY AS CONCEPT AND PROCESS

Fundamental to our analysis is the recognition that racial and ethnic identities are socially constructed and amenable to change. Although these categories are often assumed to be fixed and constant, the historical, sociological, and psychological evidence convincingly document their dynamic and flexible nature. Historically, both the categories of usage and the accepted conventions for assigning certain groups to certain categories have changed over time (and place) and continue to do so as large-scale post-1965 immigration influences the demographic profile of the United States. Given that flow and shifting demographic patterns, how do individuals see themselves and how do others see them? We know from psychological research that people have many identity options open to them (for example, age, gender, social class, and religion). Although these identity categories are surely not irrelevant to immigration, in this issue we focus primarily on ethnic, racial, and national identities. From the census to the media to the marches on the street, these identities are what are most debated and contested in the context of immigration. On the one hand, historical precedents and assumptions establish a framework for contemporary usage of identity as a subject for ethnic, racial, and national discussion. Yet, at the same time, social forces and current experiences create a process by which previous notions of identity may be challenged, redefined, and individually selected and used. These possibilities for change, both over time and circumstance, impel us to give greater attention to the processes of ethnic and racial identification and to the forces that promote one choice over another.

Constructing Identity Categories

Ethnic and racial identities are, like other social identities such as gender, religion, or occupation, categories that people use to place themselves and others in the social world (Gleason 1983; Deaux 2015). As such, they not only are forms of self-definition, articulating the aspects of self that one sees as most central to who they are, but also serve to organize a person's social world, viewing others as members of identifiable categories in which members share some common features (Vignoles, forthcoming). These categories are more than mere labels. They carry beliefs about content and meaning, have evaluative connotations, and are affirmed with varying degrees of importance. Moreover, they link individuals to social groups, shape interactions within and between groups, and provide viable links to past histories and future goals (Ashmore, Deaux, and McLaughlin-Volpe 2004).

How and why people choose particular identity categories for self-definition or for the description of others have been increasingly explored in recent years, opening up possibilities previously thought to be settled issues. A major shift in the conversation has been the recognition that identity is not an objectively determined category but instead a construction, reflecting and substantiating a particular set of assumptions about the world or a particular position for viewing the social structure in which one exists.

Across time and continents, racial categories have often been an accepted way of sorting people into groups. From the early eighteenth-century proposal by the Swedish botanist Carl Linnaeus of four basic races, defined in terms of physical features (such as color of skin, texture of hair) and elaborated by assumed traits and behaviors (such as haughty, indolent, inventive, or capricious) through contemporary census categories, race has continued to be a demarcation tool (Prewitt 2013). Extended to five categories by the German scientist Johann Blumenbach, racial classification soon became a simple color chart of white, yellow, brown, red, and black—the ethnoracial pentagon, a term offered by David Hollinger (2006). In various forms, these color-coded categories have long been used and continue to be used in this country, both in official and bureaucratic documents and in everyday usage, both descriptive and sometimes pejorative.

Although racial categorization has been ubiquitous in the United States, and in many other countries as well, the finer distinctions of ethnicity and culture have also been pervasive. The felt need to make these distinctions in the United States has often been prompted by the increasing inflow of new immigrant groups, exemplified by Robert Park and Herbert Miller's Old World Traits Transplanted (1921). As in the case of racial categories, ethnic distinctions have continued to be used, both by researchers and in common parlance, the latter documented in a series of studies of ethnic stereotypes begun by Daniel Katz and Kenneth Braly in the 1930s and continuing to the present day (Katz and Braly 1933; Fiske and Lee 2012).

Currently, the U.S. Census contains a mixture of racial and ethnic bases of classification (Prewitt 2013). A question on race asks people whether they are one of three races—white, black-African American-Negro, or American Indian-Alaska native—or some other race, which respondents are asked to specify. A fifth option, covering the broad Asian category, offers more specificity, listing ten possibilities (such as Chinese, Filipino, Samoan) and an additional Other category. The race question on the census is preceded by a question focusing specifically on Hispanic-origin respondents ("Is person of Hispanic, Latino, or Spanish origin?"), with a yes response allowing four choices (for example, "yes, Puerto Rican"), one of which is again an invitation to state a specific heritage not included in the answers provided.

As Kenneth Prewitt discusses in detail, the census exemplifies the statistical definition of race, "constructed and reconstructed by the government" (2013, 4). The categories are both cause and consequence: their inception is typically the result of political decisions and pressures, and their existence in turn influences the way that our understandings are shaped and political policies are enacted. Yet, as Richard Alba convincingly argues, the seeming impartiality of statistics does not preclude arbitrary coding and allocation practices that affect the interpretation and use of these statistics (2016a). Such biases are particularly likely to emerge in the classification of the growing multiracial population in the country.

To use a general multiracial category for the myriad of combinations that are possible in a diverse society has serious limitations, as the particular combinations are often distinctive and consequential. For example, those with an Asian-white background are likely to claim a multiracial identity whereas black-white mixed-race individuals tilt more often toward a black identity (Alba 2016b; Bratter 2016). Gender can further complicate the picture, in that women of mixed-race heritage—particularly younger women—are more likely to identity as multiracial than men are (Morning and Saperstein 2016; Pew Research Center 2015). These variations, when coupled with evidence that the multiracial population is growing (though still small in absolute terms) provide further reason to take notions of identity seriously.

Selecting an Identity: What and Why

Societal formulations make certain categories more available for regular usage, both by those who want to describe others and by individuals and groups who seek to define their own position in the society that they inhabit. In some cases, the labels commonly used in the United States are unfamiliar to immigrants or were not used (or less often used) in their countries of origin. Immigrants from Jamaica, for example, may shift from an identification with their specific country or ethnic origin to a more general West Indian or black label when they are assumed by others to be part of these larger groups in the United States (Waters 1999). Similar issues arise for many migrants from Latin America, who are categorized as Hispanic when they settle in the United States and less often distinguished by their country or culture of origin.

A variety of categories are available for a person to use in self-attribution, but clearly some are more suitable or more desirable than others. How the individual comes to define his or her identity within the set of possibilities is a multidetermined process. Certainly, the immediate environment in which a person lives—the family constellation and traditions and the neighborhood, for example—plays a role in making certain identity options both available and feasible. A person's physical features can also be influential: individuals who look more prototypical of their ethnic group in terms of physical features are more likely to be assigned to that ethnic category by others and to identify with that category themselves (Wilkins, Kaiser, and Rieck 2010). The choice of a particular racial-ethnic identity can also reflect personal motives and desired goals. Dominican immigrants, for example, often adopt the identity of Hispanic as a way of placing themselves in a category distinct from black and white (Itzigsohn 2009). Individuals from Vietnam, China, or the Philippines may opt for the more general label of Asian American because they see strategic value in adopting a panethnic identity (Okamoto 2014).

In discussions of race and ethnicity, the emphasis is often on the nonwhite sectors of the population, that is, those who are differentiated from the normative white category. Yet, in recent years, white identity has also been recognized as an important racial as well as ethnic identity category (McDermott and Samson 2005), and one that has become more prominent in the current U.S. political climate. Whereas only a decade ago some social scientists could confidently say that "Whites' whiteness is usually likely to be no more noteworthy to them than is breathing the air around them" (Sears and Savalei 2006, 901), more recent research has shown an array of behaviors that are moderated by white ethnic identification (Knowles and Peng 2005), including voting patterns in the 2016 U.S. presidential election (Major, Blodorn, and Blascovich 2016).

My Choice Versus Your Assumption: Discrepancies in Category Usage

Categories used for classification by others do not always match the identity categories that the person uses. Although both outsiders and insiders are likely to draw from the same available set of societally defined options, criteria for a particular endorsement can differ. As noted earlier, visibility of some physical features such as skin color can make agreement more likely (Wilkins, Kaiser, and Rieck 2010). Yet in other cases, physical features carry more weight for observers than for the targets of their view. For example, people of Asian descent born in the United States are often assumed by others to be immigrants rather than natural-born citizens, illustrating the outsider's preference for an ethnic-racial category, whereas the Asian insider might want to be seen in terms of the national identity of American (Tuan 1998). This difference in priorities might seem relatively harmless; however, research has shown that the target person is likely to be angered by what is seen as an offensive mistake in categorization (Cheryan and Monin 2005; Wang, Minerva, and Cheryan 2013).

Distinctions within a panethnic grouping can also be more important for self-identification than for categorization by others. Panethnic identities, such as those made available in census forms, can be accepted and used, particularly for political mobilization (Okamoto 2014; Schildkraut 2011). At the same time, interethnic distinctions remain important. The failure of others to correctly identify a person's national-origin identity—such as by assuming that a person speaking Vietnamese is Japanese, or that a Venezuelan would celebrate Mexican Independence Day—can elicit adverse reactions (Flores and Huo 2013).

Discrepancies such as these between identities preferred by an individual (or group) and categorical assignment by others have numerous implications for interactions between the two. At a minimum, one cannot assume equivalence of the label and the interpretations that follow from the use of that label when shifting between the two vantage points. More broadly, discrepancies in the preference for and use of identity categories can potentially disrupt, or at least make problematic, interactions between those representing one versus the other perspective.

Identity Change

The available options for identity categorization within a society can change, as the frequently modified U.S. Census illustrates. Yet even when categories remain stable, people's claims to those categories vary over time and circumstance. Evidence of this variability in the use of census categories is provided by Carolyn Liebler and her colleagues (2017), who used matched 2000 to 2010 census files to determine the extent to which self-selected ethnoracial categories remained stable or changed over the period. They find not only that 6 percent of respondents differed in self-labeling between the two periods, but also that individuals in particular groups were most likely to show flexibility in their responses. Respondents who reported a multiracial background at one point in time, for example, frequently chose only a single category at another. This pattern was particularly true of those who combined white and Asian or white and Hispanic labels in one of their self-descriptions.

Possible reasons for changes in an identity category during a ten-year period are numerous, from methodological (such as unreliability of categories that were not common-use terms for respondents) to experiential (such as a substantial change in a person's social network during the period that makes one ethnic identity category more salient or relevant than another). More longitudinal research on the self-labeling of identity would be helpful in sorting out possible causes.

At the same time, social psychological research provides ample evidence that people can easily shift from one identity to another in shorter, more situationally dependent circumstances. Evidence for these shifts is both cognitive and behavioral. In what is termed cultural frame-switching, the availability of cognitive cues relevant to one or another identity can affect the thoughts and expressions of multicultural persons, often in wholly nonconscious and automatic ways (Benet-Martínez 2012). At a more behavioral level, in a process that has been termed bicultural (or multicultural) identity performance, it has been shown that people will strategically make choices among possible identities when presenting themselves to others (Wiley and Deaux 2011; see also Klein, Spears, and Reicher 2007). Numerous factors influence these choices. For example, people are likely to emphasize a version of self that is consistent with the situation or with the characteristics of those around them, such as identifying as a student in a classroom or as a Muslim when in a mosque. Norms and social pressures influence many of these choices: for example, the American-born daughter of Chinese immigrant parents would be likely to express her Chinese identity more strongly when at home, but her American identity in a more diverse school setting. Individual differences in the importance of a particular identification or in allegiance to an ethnic group also moderate presentational choices. Not surprisingly, those who strongly identify with a category are more likely to show consistency across situations than are those whose subjective investment in the category is weaker (Wiley and Deaux 2011). Thus change, both short-term and long-term, must be considered an essential part of any account of ethnic and racial identity.

CHANGEABILITY OF ETHNORACIAL CATEGORIES AND IDENTITIES: PUTTING THE PRESENT IN HISTORICAL CoNTExT

If our theme is immigration's impact on changing identities in the contemporary United States, it is useful to view the present in light of the past. Looking back to earlier periods in U.S. history brings out in a powerful way how racial and ethnic identities are socially constructed in the context of large-scale immigration and points to underlying factors that have led to specific changes over time. Racial differences may seem permanent and immutable—this is the essence of race, which refers to the belief that visible physical characteristics or putative ancestry define groups or categories of people in ways seen to be innate and unchangeable. But "race has only the meaning that culture gives it," and in fact race is a changeable perception (Robinson 2017, 16). So, too, are ethnic categories and ethnic identities, which not only may shift in meaning in different everyday situations, but also have been viewed in different ways in earlier eras. Ethnicity is often defined by ancestry and descent as opposed to physical markers, but such distinctions are not as easy or unambiguous as is sometimes assumed. George Fredrickson argues that race refers to what happens when ethnicity is deemed essential or indelible and made hierarchical (2002, 155); Stephen Cornell and Douglas Hartmann emphasize that race is a construction imposed on a group against its will and that ethnic consciousness is a self-construction of the group (2004). Indeed, it has been argued that the very concept of ethnic group in American discourse to refer to groups with common cultural characteristics and ancestry did not emerge until the early twentieth century, when it was used by Jewish intellectuals to resist being assimilated into the melting pot while avoiding racialization (Foner and Fredrickson 2004, 4; Hattam 2004).

As these comments suggest, debates about the use of race versus ethnicity as conceptual categories have a lengthy and continuing history, and usage of the terms can vary both within and between disciplines. Such differences of opinion and usage attest to the fluid character of identity construction and definition.

The Meaning of Whiteness: Eastern and Southern Europeans in the Past

The commonly taken-for-granted meaning of white or whiteness has undergone significant shifts over time. It is often said that a major distinction between today's immigrants and those a hundred years ago is that then they were, in the main, white Europeans and today they are, in significant numbers, people of color, but this is to impose early twenty-first century understandings of racial categories on the past. Race today is basically a color word, but when it came to the millions of southern and eastern European immigrants a century ago, race and color were not "perfect synonyms"; in the first half of the twentieth century, "one could be considered both white (color) and racially inferior to other whites (race)" (Fox and Guglielmo 2012, 334).

Southern and eastern European immigrants at the time were widely viewed as racially inferior to those with origins in northern and western Europe. At the same time, although their whiteness was sometimes questioned, they were legally white—that is, not prevented from naturalizing as Asians were, for example, or subject to the antimiscegenation laws that existed in many states. They were also placed in the white category by federal agencies, including the U.S. Census (Fox and Guglielmo 2012, 364). Historians have sought terms to describe the ambiguous racial position of Jews and Italians a century ago: "inbetween peoples," in the words of the historians James Barrett and David Roediger (1997), "probationary" whites in those of the historian Matthew Frye Jacobson (1998), or, as Thomas Guglielmo suggests (2003), racial outsiders but color insiders—that is, racially inferior to other whites on the basis of notions of stock, heredity, blood, and selectively chosen physical characteristics (Fox and Guglielmo 2012, 343). As Jacobson aptly puts it, eastern and southern Europeans were both white and racially distinct from other whites (1998).

In the early twentieth century, Jewish and Italian immigrants, who then made up more than half of the southern, central, and eastern European immigrants in the United States, were seen as belonging to inferior "mongrel" races that would alter the essential character of the United States and pollute the nation's Anglo-Saxon or Nordic stock. They were thought to have distinct biological features, mental abilities, and innate character traits and many Americans believed that they were physically identifiable: facial features (including the "Jewish" nose) were often noted in the case of Jews, "swarthy" skin in the case of Italians. Even as late as the 1930s, an American history textbook asked whether it would be possible to absorb the "millions of olive-skinned Italians and swarthy black-haired Slavs and dark-eyed Hebrews into the body of the American people" (Barker, Dodd, and Commager 1934, cited in Fitzgerald 1979, 79-80).

What factors eventually led Jews and Italians to become part of an all-encompassing white community and no longer set apart in the popular mind as inferior, in racial terms, from those with northern and western European ancestry (Alba 2009; Foner 2000, 2005)? The economic and occupational successes of Jews and Italians were critical: economic prosperity and the enormous expansion of higher education in the post-World War II years and the benefits that came with postwar policies, such as the GI Bill of 1944, provided opportunities for educational and job mobility for both the Jewish and the Italian second generations. Climbing the social and economic ladder was accompanied by increased intermixing—in neighborhoods, at work, and eventually in marriage—with those whose roots were in different parts of Europe.

A combination of other factors was also involved. That those with origins in eastern and southern Europe shared a safe haven of legal whiteness with other European groups from the very beginning—and were not subject to the same kind of systematic legal and official discrimination facing black, Asian, and Mexican immigrants—was important in their eventual racial inclusion into the white mainstream. Also, because many Jews and Italians physically resembled members of the older European groups, it was often possible for them to blend into the majority population ("to pass") if they shed cultural features such as distinctive dialects or dress.

Nor can we dismiss the ending of the massive eastern and southern European immigrant influx following restrictive U.S. legislation in the 1920s, which reduced fears of old-stock Americans about the deluge of racial inferiors and contributed to cultural assimilation. Moreover, the massive migration of African Americans from the rural South to northern cities from World War I on likely facilitated the acceptance of Jews and Italians as full-fledged whites by changing the racial order in these cities from one marked by the multiplicity of white races to one focused on race as color (see, for example, Guterl 2001). As blacks became a significant proportion of the population in cities like New York and Chicago (where they were less than 2 percent of the population in 1900), Jews and Italians often sought to distinguish themselves from (and claim superiority to) African Americans—which they did by emphasizing their whiteness.

The Nazi genocide made anti-Semitism less respectable. World War II had another effect as well, occurring when the mass inflow of immigration had receded and the army was filled with U.S.-born generations with origins in all parts of Europe. Fighting in segregated white platoons "brought about a self-conscious wartime unity that transcended ethnic lines among whites" (Alba 2009, 80; see also Gerstle 2001). Especially after the war, struggles by the groups themselves—most notably Jewish organizations such as the Anti-Defamation League—to eliminate racial exclusionary barriers in housing, higher education, resorts, and social clubs were key in the passage of laws prohibiting racial and religious bias in employment and higher education. Whether any of these factors will play a role in changing the social construction of racial and ethnic categories in the future is one of the topics we take up in the conclusion.

Whiteness in Flux Today

Whatever the course of change in the future, we already see shifts as a result of the massive immigration of the last half century, and this includes the meaning of whiteness. The scholarship on white racial identity in the United States has highlighted how much it is taken for granted as well as rooted in social and economic privileges, with these privileges often invisible to many whites who do not think of themselves as having a race at all. Although white racial identity continues to encompass European Americans' advantaged position in the ethnoracial hierarchy, it is also very much in flux (McDermott and Samson 2005).

One issue that has come into prominence, particularly in analyses of the last presidential election campaign and support for President Donald Trump, is the degree to which many whites feel that their "natural" privileges as whites are in jeopardy, indeed under siege, in the context of demographic change fueled by immigration and the economic and political gains of nonwhite groups in the post-civil rights era—something dramatically symbolized by the election of the nation's first black president. In the midst of growing ethnic and racial diversity across the country, whiteness has become more salient to many Americans who believe that nonwhites are undeservedly receiving advantages and "cutting in line" (Hochschild 2016). The historian Nell Irvin Painter argues that in the Trump era what it means to be white has fundamentally changed "from unmarked default to racially marked . . . from of course being beauty queen and of course being the cute young people selling things in ads to having to make space for other, nonwhite people to fill those roles" (2016).

In their article in this issue, Maureen Craig and Jennifer Richeson report empirical evidence of group-status threat—that is, the threat or fear that whites will lose their privileged and dominant position in America's racial hierarchy (2018). They note that many white Americans believe that antiwhite discrimination is on the rise, a view especially likely to be held by whites living in areas with relatively large racial minority populations. Deborah Schild-kraut and Satia Marotta (also this issue) ask whether white millennials, born after 1980, might be less conservative in their political views than older whites, given their greater experience with a diverse society (2018). Although white millennials are more liberal on some issues, the stronger message from Schildkraut and Marotta is that differences between whites and nonwhites are much stronger than the generation difference among whites.

Another shift, although unusual, is worth attention. In some communities where Asian immigrants are numerically dominant, highly educated, and well off (and blacks and Latinos virtually absent), whiteness may well be downgraded. This has happened in Cupertino, California, an affluent white-Asian city in Silicon Valley where high academic achievement is no longer associated with whiteness: whiteness has come to stand for lower achievement, laziness, and academic mediocrity, and Asianness is linked with academic success, hard work, and achievement (Jiménez and Horowitz 2013).

There are also intriguing questions about white identities among groups classified as white on the census but who do not see themselves, and may not be seen by others, as white. Whereas for the descendants of earlier European immigrants, ethnic identity has become optional and white ethnic distinctions have gradually blurred into a more diffuse European American identity, especially by the third and fourth generations, this may not be the case for groups such as Arab and Middle Eastern Americans. Although they are officially considered white by the U.S. Census, they often have a stronger identification with their home country, region of origin, or religion than with whiteness, in good part owing to the widespread public suspicion and discrimination they face, especially among the many who are Muslim (Tehranian 2010). In fact, advocacy groups have pressured the census to adopt a separate Middle Eastern and North African (MENA) category because of the discomfort many have with checking off white, as well as a desire to increase opportunities for federal funding and gain more political clout; in early 2018, however, the Census Bureau announced that it would not add MENA to the next census.

Asians and the Elasticity of Race

Asians, especially those with origins in East Asia, have experienced a remarkable change in racial status, undergoing a metamorphosis in the last seven decades from "yellow peril" to model minority. Indeed, when whites stereotype Asian Americans today it is often for being economically successful (Abrajano and Hajnal 2015, 54).

It is hard to imagine that the Chinese and Japanese in the United States used to be cast, as Yen Le Espiritu puts it, as "almost blacks but not black" given that now they are frequently seen as "almost whites but not whites" (1997, 109). In the past, racial prejudice against Asians was inscribed in restrictive immigration and naturalization laws. The Chinese Exclusion Act of 1882 singled out the Chinese as the first and only group to be excluded from the United States on the basis of race, ethnicity, or nationality; by 1917 Congress had also banned the immigration of most other Asians. For much of the nation's history, Asian immigrants were denied the right to become citizens; the 1882 Chinese Exclusion Act decreed that the Chinese were "aliens ineligible for citizenship," and over the next few decades the rule was extended, through a series of decisions in state and federal courts, to all other immigrants from east and south Asia (see Kurien 2018). It was not until 1943 that Chinese immigrants gained the right to become citizens and that the discriminatory immigration laws affecting Asians began to be relaxed. Only in 1952 was naturalization eligibility extended to all Asians.

Anti-Asian sentiments were particularly virulent on the West Coast, where several states adopted laws prohibiting Asian-white intermarriage. A 1913 law, targeting Japanese farmers, barred Asian immigrants from owning land. When a California court held in 1885 that the public schools had to admit Chinese children, the state legislature passed a bill allowing school districts to set up separate schools for "Mongolians" (Wollenberg 1995). Most devastating of all, during World War II more than one hundred thousand Japanese who lived on the Pacific Coast were forcibly evacuated and moved to internment camps.

Changes in U.S. immigration policy, foreign relations, and law set the stage for radically altered perceptions of Asians, including the abolition of the exclusion regime in the mid-twentieth century and revocation of Asian immigrants' ineligibility to citizenship. Also, Asian immigrants' home countries have changed over time. Americans once saw Asia as a backward region; now Japan, Taiwan, South Korea, and China are modern advanced nations and world economic powers. Of greater importance, however, in the higher racial status of Asians is that a large proportion of post-1965 Asian immigrants are highly educated and highly selected from their countries of origin. As Jennifer Lee and Min Zhou note, Chinese and Korean (as well as Asian Indian) immigrants are not only more highly educated than nonmigrants from their countries of origin, they are also more highly educated than the general U.S. population (Lee and Zhou 2015; on the Chinese, see Tran et al. 2018). Hyper-selectivity goes a long way toward explaining another significant, and related, factor in East Asians' altered racial status: the extraordinary educational success of the second generation, including their significant overrepresentation in the nation's most competitive magnet schools and elite private universities. "Despite decades of institutional discrimination and racial prejudice," Lee and Zhou observe, "the status of Asian Americans has risen dramatically in less than a century. Today, Asian Americans are the most highly-educated [racial] group in the country, have the highest median household incomes, the highest rates of intermarriage, and the lowest rates of residential segregation" (2014, 8).

The substantial rates of intermarriage between Asian Americans and whites not only reflect the more positive views of Asians in the current period but also support and strengthen them. About one in three U.S.-born Asian Americans is married to a non-Hispanic white (Alba and Foner 2015); in 2010, among new marriages contracted in the past year, the figure was comparable, nearly four in ten U.S-born Asians marrying a non-Asian, by and large a non-Hispanic white (Wang 2012).

The frequent experience of marriage to whites among the second generation does not mean that negative stereotypes of Asians have disappeared; they have not. East Asians may seem almost white to many Americans, but as Mia Tuan puts it, yellow is not white (1998, 164). One persistent negative stereotype about Asian Americans is that they are newcomers, or thought of as "forever foreign" no matter how many generations their families have been in the United States, a perception that may be reinforced by continued large-scale inflows of Asian immigrants (Tran and Valdez, n.d.). Another is the model minority stereotype, which labels Asians as a racial group distinct from the white majority at the same time as they are lauded as culturally programmed for success, "well assimilated, upwardly mobile . . . and definitively not-black" (Wu 2015, 2). Hailing Asian Americans as a model minority, it has been argued, overlooks the heterogeneity among Asian immigrant ethnic groups, diverts attention away from the existence of continued racism against Asian Americans, and pits them against blacks and Latinos. Still, that Asian Americans are often now touted as a model minority reflects a positive change from the "yellow peril caricatures" of the past that described Asians as illiterate, undesirable, and unassimilable immigrants (Lee and Zhou 2014, 8).

Another change that should be mentioned is the increasing use of the Asian label as a marker of self-identity and identity attribution by others. In the pre-1960s era—when the Asian population was overwhelmingly of Chinese, Japanese, and Filipino origin—those with roots in Asia strongly identified with their national-origin identities, but not in terms of a shared Asian label. To what extent those seen as Asian Americans today identify that way themselves is an open question; their national-origin identities, associated with distinctive languages, religions, national histories, and cultures, seem to supersede the broader Asian identity much, perhaps most, of the time. Still, as Okamoto argues, the use of the category Asian "by mainstream institutions . . . [has become] institutionalized and . . . taken for granted in everyday interactions" (2014, 48). That the census and frequently the mainstream media as well use the term Asian has had an impact. Before 1970, the U.S. Census Bureau categorized Asian groups as different "races"—Chinese, Japanese, and Filipino in 1950. In 1990, the racial category Asian and Pacific Islander was used for the first time in the decennial census, partly in response to civil rights legislation and the need for a standard system to collect data to enforce equal opportunity and affirmative action policies. Researchers, businesses, public agencies, educational institutions, foundations, hospitals, and industry adopted the category Asian to collect data, award grants, and allocate resources (Okamoto 2014), which encouraged a panethnic Asian identity. By the mid-1980s, Asian panethnicity had become well established as an organizing principle for building a community among groups of different ethnic origins, replete with many panethnic organizations and institutions (Okamoto 2014, 45-46; see also Espiritu 1992).

Creating Hispanics

The case of Hispanics also represents a contemporary sea change, in large part because the very category Hispanic is a modern-day invention. Hispanics are now the largest minority group in the nation, attributable to high levels of immigrant inflows as well as relatively high fertility among Latino immigrants and U.S.-born Latinos in recent decades (Durand, Telles, and Flashman 2006). It is now normal to hear about the Hispanic vote and Hispanic organizations, but in the mid-twentieth century the press and pundits wrote about Mexicans or Puerto Ricans, not about Hispanics (Fischer 2014). A key development occurred in 1980, when the census adopted the term Hispanic as an enumeration category; this decision both reflected changes in American society and, at the same time, entrenched Hispanic as a legitimate official category, contributing to its importance as an identity label for Hispanics themselves as well as non-Hispanics.

The emergence of Hispanic (and Latino) as categories of identity is to a large degree the result of the politics of ethnic and racial classification (Itzigsohn 2009; see also Rumbaut 2006). As Cristina Mora tells it in Making Hispanics, the creation of this new identity label in the 1970s and 1980s involved a combination of activists seeking political clout, government funds, and philanthropic support by uniting under the Hispanic banner; Spanish-language television broadcasters seeking a larger national market; and activists and politicians successfully campaigning to have the census adopt the Hispanic category (2014). To this day, Mora argues, the web of media, state, and activist networks has upheld the notion of Hispanic panethnicity (2014, 16).

A key question, of course, is the extent to which those with origins in Latin America actually identify as Hispanic or Latino. It is not an either-or situation. Although many, perhaps most, Latin American immigrants prefer to be known by and primarily identify in terms of their national origins, they also often identify as Hispanic or Latino; the two identities, in other words, are not mutually exclusive but instead complementary. What seems clear is that what started out as a statistical term of convenience or tool to bring those of Latin American origin together has been transformed into a real social entity.

Whether Hispanics should be considered a race or an ethnic group is an issue characterized by confusion and debate. On the ethnic group side is the fact that the Census Bureau classifies Hispanics in terms of ethnicity, not race, and, as a group, they have varying skin tones, many being phenotypically white and more than half checking white on the 2010 Census. Hispanics who have light skin color and European features, are well educated, or are well to do may gain acceptance as whites, at least in some contexts and places; by the same token, dark-skinned Hispanics may suffer many of the same disadvantages as African Americans (see Kibria, Bowman, and O'Leary 2013, 129; Fox and Guglielmo 2012).

Yet a scheme that treats Latinos as a race between whites and blacks, as Wendy Roth puts it, is winning out in the United States today (2012, 64). In the media, public discourse, some government reporting standards, and everyday language used by Latinos and non-Latinos alike, the view of Latinos as a separate racial group has increasingly come to dominate. The term tends to conjure up images of people who are brown or tan-skinned and foreign in speech and manner. Or as Nazli Kibria and her colleagues observe, notions of intrinsic difference from and inferiority to whites are long-standing features of the stigmatization of Latino/a populations in the United States (2014, 126; see also Brown, Jones, and Becker 2018).

The case of Mexicans, by far the largest national-origin group of Latinos, illustrates some of the complexities of their position on the white-nonwhite boundary in the past and their racial status today. Looking to the past, the census classified Mexicans as white before 1930 and official instructions in World War II called on local draft boards to classify Mexicans as white; no state miscegenation law specifically barred unions between whites and Mexicans and the federal government accepted Mexicans as white for purposes of naturalization. Yet, at the same time, the boundary between whites and Mexicans appeared bright "in the sense that a wide range of individuals and non-state institutions recognized Mexicans as non-white. Many race scientists [in the early twentieth century] categorized the vast majority of Mexicans as nonwhite. Numerous Anglos did as well, a point that became most obvious when Mexicans [in many parts of the Midwest and Southwest] found themselves excluded from white-only public accommodations, when realtors refused to sell them homes in white neighborhoods, or when school officials excluded them from white schools" (Fox and Guglielmo 2012, 367).

Fast forward to the present and Mexicans have come to be seen, as some scholars argue, as a racialized ethnic group, often stigmatized as inferior, illegal, and foreign and regarded as nonwhite (see Alba and Foner 2015, 107). Another view stresses that Mexican Americans are targets of prejudice and discrimination because of nativism, or intense opposition based on their foreign connections, rather than because of beliefs about their racial inferiority. According to one argument, Mexican Americans experience a racialized form of nativism in which their presumed foreignness is central and their right to be in the country is questioned; third- and later-generation Mexican Americans, in this perspective, may encounter discrimination because they are associated with and often mistaken for new Mexican immigrants (Jiménez 2010). Pigmentation and other physical features also appear to be involved. Skin color among Mexicans and other Latinos has been shown to matter for socioeconomic standing and residential integration, the lighter, not surprisingly, the better—light skin color (along with social class standing) enabling some, at least some of the time, to be seen as white. A recent ethnographic study shows that Mexican Americans and Mexican immigrants, including those attaining middle-class status, often experience discrimination in their daily lives because of their skin color and surnames (Dowling 2014). Indigenous ancestry, manifested in physical features such as facial appearance and height as well as skin color, also poses racial challenges. As David Lopez and Ricardo Stanton-Salazar put it, speaking of Mexicans, "those who fit the mestizo/ Indian phenotype, who 'look Mexican,' cannot escape racial stereotyping any more than African Americans, though the stigma is usually not so severe" (2001, 75).

Black Immigrants: Persistence in the Context of Change

This brings us to blacks, who are currently nearly one in ten immigrants in the nation. Blacks are the quintessentially racialized Americans with a history of special disadvantage—slavery, Jim Crow, ghettoization, and, most recently, massive incarceration (Foner and Fredrickson 2004, 8). The particular history of blacks in America has led to a highly rigid boundary surrounding them and a pattern of what might be called black exceptionalism. The historical legacy of deeply entrenched antiblack racism, combined with continued racial inequalities, has meant that the way contemporary black immigrants and their children identify, and are identified by others, is in many ways eerily like it was a hundred years ago. Overall, critical aspects of the dynamics of race, ethnicity, and identity for blacks of immigrant origin appear to be at least as much a matter of persistence as change across historical time.

Of course, not all is the same: many changes for the better have occurred since the early and mid-twentieth century. When tens of thousands of Afro Caribbean immigrants arrived in the United States a hundred years ago, a harsh regime of institutionalized racial oppression and legal segregation prevailed in the South, confining blacks there to separate (and inferior) schools, hospitals, and public places, putting them under threat of violence and lynching, and depriving most of the right to vote. Indeed, these conditions were a major reason why black immigrants at the time avoided the South and headed for northern cities, most notably New York. Although New York did not have a Jim Crow regime like the one in the South, it was no racial paradise. The housing market was rigidly segregated. Relatively few Afro Caribbean immigrant women had jobs in the manufacturing sector where so many eastern and southern European immigrant women clustered; black immigrant women overwhelmingly worked as household servants (Model 2001; Watkins-Owens 2001). During World War II, blacks in the armed forces were relegated to segregated units and after the war, as Ira Katznelson details, black veterans generally were unable to reap the benefits of the GI bill that enabled so many in the eastern and southern European second generation to attend college and buy homes (2005). In an environment of intense antiblack racism, the children and grandchildren of early twentieth-century Afro Caribbean immigrants had little choice as they incorporated into American society: they became African American (Vickerman 2016, 74-75).

Post-1965 black immigrants, who include a growing number of Africans as well as Afro Caribbeans, have entered a post-civil rights America that has seen declines in the grip of racial inequality on the life chances of blacks; the growing presence of blacks in the upper middle class and visibility in elite positions, from the heads of large companies, well-known newscasters and other personalities in the media, to the first black president; a rise in the number and proportion of mixed-race (black-white) individuals; and evidence of a greater willingness among whites to see differences among African Americans and more acceptance of middle- and upper-class African Americans (Waters and Kasinitz 2015; Edsall 2017; on the integration of African American history into school curricula, see Lash 2018).

Yet despite these improvements, stark social cleavages involving people of African ancestry remain. Rates of black-white intermarriage are much lower than those of Asian- and Hispanic-white intermarriage. Whether native or foreign-born, blacks in the United States are still highly residentially segregated from whites—more so than Hispanics and Asians—and the New York City metropolitan area, still home to the largest black immigrant population in the country, is one of the most segregated (Alba and Foner 2015; Foner 2001, 2015, 2016). In a large-scale study of young adult members of the second generation in the New York metropolitan area, West Indians reported the most discrimination (compared to Dominicans, South Americans, Chinese, and Russian Jews), especially in public places on the streets, in stores, and from the police (Kasinitz et al. 2008). Anonymous encounters with shopkeepers, security guards, and particularly the police in public spaces "are powerful because they are so purely 'racial.' In such confrontations class differences do not count . . . Nor do ethnic differences. . . . A police officer rarely has a basis for knowing if a young man on a public street is African American or West Indian, middle class or poor. If the police officer discriminates, it is on the basis of race alone" (Waters 2014, 159).

How blacks of immigrant origin are identified also shows considerable continuity over time. Much like the early and mid-twentieth century pattern, the descendants of black immigrants continue to face great difficulties today in being recognized by others in terms of their ethnicity rather than their race. Moreover, black immigrants, at least so far, have not had a significant impact on how blackness is widely seen in the United States.

To be sure, some changes are apparent. It might even be said that contemporary black immigration is contributing to "tweaking" notions of a monolithic blackness in cities like New York where black immigrants have become a considerable presence (Vickerman 2001). In early twentieth-century New York, for example, Afro Caribbean politicians played down their ethnic identity in the quest for office, putting themselves forward as representatives of the broader black community (Kasinitz 1992). Today, when the number of black immigrants is so much larger in New York City—more than ten times the 1920 figure—several sizable West Indian neighborhoods have provided the base for politicians to use the "ethnic card" to gain support in campaigns for local office. In general, the larger populations of Africans and West Indians provide more scope for ethnic identities to thrive within their communities in private interactions as well as public spaces. It has been suggested, as well, that in some instances second- and third-generation West Indians have actively tried to create a hybrid identity with younger African Americans, based in part on melding aspects of Caribbean and African American popular culture (Vickerman 2016, 77).

To tweak notions of a monolithic blackness, however, is not to create substantial change. Even when their ethnicity is recognized in certain places and contexts, blacks of immigrant origin are seen as an ethnic group within the larger black population, with all the negative stereotypes that this so often involves. Their racial status as blacks, in other words, is always salient.

Identity issues take on special significance for the second generation, born and raised in the United States, as research on Afro Caribbeans reveals. Without an accent or other cues to immediately telegraph their ethnic status to others, second-generation Afro Caribbeans, in the words of Philip Kasinitz and his colleagues, are likely to fade to black (Kasinitz, Battle, and Miyeres 2001). Second-generation Afro Carib-beans who continue to identify with their ethnic backgrounds are aware that unless they are active in conveying their ethnic identities, they are seen as African Americans and that the status of their black race is what matters in encounters with whites. Or, as Milton Vickerman puts it, "the general public has yet to discard the reflexive habit of identifying 'black' with 'African American'" (2016, 78). The crux of the problem is that being seen as black American, they are subject to the same kind of racial prejudice and exclusion that black Americans are (see Waters 1999). The children of African immigrants face the same problem, although it is unclear whether they will have an easier time establishing an ethnic identity given, among other things, their distinctive surnames and whether they will be more or less likely than the Afro Caribbean second generation to want to do so (see, for example, D'Alisera 2009; Imoagene 2017; Ludwig 2013).

If, so far, contemporary black immigration has brought no significant change in the meaning of the racial categories black or African American, there still may be shifts ahead to incorporate the large and growing populations of recent West Indian and African origin. On the one hand, the children and grandchildren of today's black immigrants are likely to assimilate as African Americans, but, on the other, the meaning of African American may expand, if only in the African American community itself. It is not inconceivable that in the years ahead African American will be seen as a product not just of centuries on American soil but also of intermixing with black foreigners (Vickerman 2016, 2001).

CHANGING IDENTITIES AND INTERGROUP RELATIONS

As the array of ethnic groups increases in number and presence, interactions between people almost inevitably are influenced by ethnic and racial group membership—and by the members' identification with those groups and their perceived membership by others. These interactions can be positive or negative, can result in cooperation or competition, and can create occasions for unifying groups under a superordinate umbrella or for sharpening group boundaries and increasing distinctiveness. Further, these encounters occur for individuals as well as groups. At the individual level, contacts between members of different ethnic groups take place in casual encounters, friendships formed at work or in the neighborhood, or, perhaps with the greatest consequences for future identity issues, in the context of the family and kinship circles as a result of intermarriage. At the group level, ethnicity can structure interactions at work, in neighborhoods, and in political organizations, sometimes pitting groups against each other in conflict over resources, either imagined or real, and at other times, laying the groundwork for the development of panethnic (superordinate) identity groups that can unite for common goals.

Intermarriage

Intermarriage is perhaps the most intimate of all intergroup relations. It is, of course, an outcome of interpersonal relations between individuals in different ethnic and racial groups. But it also both affects relations between those of different race and ethnicity within families, kinship networks, and other contexts and has an impact on the identities of the partners and their multiethnic or multiracial children. Ethnoracial intermarriage is on the rise in the United States: about one out of every six marriages in 2015 was an interracial or interethnic marriage, more than twice the rate in 1980 (Livingston and Brown 2017). According to a Pew Research Center's analysis of census data, intermarriages of Asians, Hispanics, and blacks mainly involve a white spouse (some seven in ten new intermarriages in 2010), and this is the type of mixed union we focus on here; in most cases, both partners are native born, many of the nonwhite spouses being the children or grandchildren of immigrants (Wang 2012). Immigrants are much less likely to intermarry than the second and third generations, partly because many are already married when they arrive and because they may lack English fluency and other skills that facilitate easy interaction with the native born. The children and grandchildren of immigrants, born and raised in the United States, not only speak English but also often have many opportunities to mingle with those in the white majority in such places as schools, colleges, and work (Alba and Foner 2015, 208).

Whether considering rates of intermarriage or their consequences, the boundary separating blacks and whites stands out. Although marriage between blacks and whites in the United States has increased appreciably in recent decades (antimiscegenation laws were still on the books in sixteen states when the Supreme Court declared them illegal in 1967), it falls well short of the levels between whites and Hispanics and between whites and Asians. The frequency of intermarriage among the second generation is revealing. According to a recent summary, for second-generation Hispanics and for the largest group Mexicans, the rates of marriage to non-Hispanic whites are 35 to 40 percent; for second-generation Asians in general and most Asian national-origin groups (except Indians and Vietnamese) the rates vary roughly between 30 and 45 percent, depending on gender, women being more likely to intermarry. By comparison, only about 10 percent of the Afro Caribbean second generation have white partners (Alba and Foner 2015, 209-10). These low figures for Afro Caribbeans reflect a continued deep social cleavage involving groups with visible African ancestry, including the persistent stigma attached to blacks and the high levels of residential segregation they experience, which reduce opportunities for close social contacts and intimate relations with whites to develop.

The identity options of children of mixed unions also reflect continued black exceptionalism. Admittedly, mixed-race individuals' identities are fluid and change over time, no doubt more so than among individuals with ethnoracially unmixed backgrounds (Alba, Beck, and Sahin 2018), and what we know about these identities is sparse, if only because the mixed-race population, though now rapidly growing, historically has been a small share of the U.S. population as a whole. Still, a study based on in-depth interviews with interracial couples and their children suggests greater constraints facing the children of black-white unions in how they identify (Lee and Bean 2010). The study found that Asians and Hispanics married to whites felt that their U.S.-born children had the option to identify as whites, without having their decisions questioned by outsiders or institutions. The experiences of the children of black-white unions were different. Blacks who intermarried with whites said that their children were often seen as black only, underscoring that the "one drop rule" is not altogether a relic of the past. In fact, black-white couples emphasized that nobody would take them seriously if they tried to identify their children as white, and their children chose to identify as black rather than as multiracial or nonblack (Lee and Bean 2010, 108-9). Although a recent Pew Research Center survey (2015) found that a similar percentage of adults with a white-Asian (70 percent) and white-black (61 percent) background identified as multiracial, it was a different story in terms of how they thought others saw them: six in ten with a black-white background said a person passing them on the street would see them as black (only 7 percent said strangers would see them as white); around four in ten of those with an Asian-white background said that they would be seen as white, and one in four as Asian. Experimental work in social psychology supports these findings: although both biracial Asian whites and black-whites are more likely to be perceived as minorities than as whites by white observers (a pattern that has been determined by hypodescent), the threshold for being perceived as white was higher for biracial black-whites (Ho et al. 2011). Perhaps the best-known example of the identity issues facing black-white multiracials is Barack Obama, the son of a Kenyan father and white mother, who is seen, without question, as the nation's first black president, and as an adult defined himself as a black American.

The former president aside, similar patterns emerge with regard to friendship and family relations that can be viewed as a manifestation of how multiracial adults are identified by others and see themselves. The social worlds of biracial adults with Asian and white backgrounds in the Pew Research Center survey were more likely to lean to the white side of their ancestry than were those of biracial adults with black-white backgrounds (2015). Biracial white and Asian adults had more close friends who were white than Asian and more contact with white than Asian family members; most said they felt accepted by white relatives and by whites in general. Indeed, the Asian-white biracials were more likely to feel very well accepted by whites than by Asians. In contrast, black-white biracial adults tended to tilt more to the black side: they had more close friends who were black than white, had much more contact with black than white family members, and felt a much greater sense of acceptance from black people than whites (see also Chito Childs 2005).

Not unexpectedly, multiracial adults with a black background were far more likely than those with Asian-white ancestry to say they had experienced discrimination because of their racial background. A large proportion of those who said that casual observers would describe them as black also said that they had been unfairly stopped by the police—the same proportion, it turns out, as single-race blacks who reported receiving unfair treatment from the police. One young black-white biracial man summed up the feelings of many others in the survey: "No matter how I see myself, at the end of the day I'm still black" (Pew Research Center 2015, 58). Even so, by virtue of having parents in two ethnoracial groups, biracial children, whether the nonwhite parent is black, Asian, or Hispanic, inevitably have intimate contact with members of two groups. Just how this contact, as well as relations within wider kinship and friendship circles, affects the children's identities and interactions is one of the many topics about which we have much to learn.

Intergroup Contact

Whereas intermarriage can foster contacts with an ethnic or racial community other than one's own, in some ways a by-product of original interpersonal goals and attachments, direct contacts between ethnic groups are shaped by a variety of other factors. Some of these are idiosyncratic, the product of specific individual histories and encounters with particular people; others are structured by larger social and economic forces such as patterns of residence, labor market incorporation, and political mobilization and alliances. At the same time, the form of intergroup contact is shaped by widely shared and indeed normative views based on prevailing categorization practices and the narratives associated with those categories. These "normative ideas in circulation" influence individuals' views of those in different groups and how they should be treated (Appiah 2016, 164). Fundamentally, intergroup interactions are energized by issues of identity: how strongly does a black, white, Latino, or Asian person, for example, identify with his or her ethnic and racial group, and how readily does the other party view them as representatives of that group?

The tradition of research on intergroup relations is a long one, in particular on the consequences of contact between groups on the attitudes and behaviors of the members of the different groups. Originating in large part in the work of Gordon Allport and carried on by a long line of social scientists, much of this work has focused on the black-white dynamic (Allport 1954; see also Brown and Hewstone 2005; Pettigrew and Tropp 2006). More recently, however, ethnic diversity has opened the door to exploration of numerous combinations.

In its simplest form, the question is whether greater contact between different ethnic groups facilitates or impedes positive attitudes and friendly relations; evidence can be gathered to support both sides of this argument. Recent data reported by Eric Knowles and Linda Tropp, for example, suggest that for white Americans, greater exposure to Hispanics, as indexed by their numerical share of the neighborhood population, is associated with stronger white identities (2016). Yet research also indicates that although an increase in an immigrant group in a community can seem threatening to the native residents, those numbers also increase the probability of having contacts—some potentially favorable—with members of that immigrant group (Pettigrew, Wagner, and Christ 2010). More probing analyses suggest a variety of moderating conditions that can influence the outcome of intergroup contact; these conditions include equality of status between the two groups, the existence of common goals, experiences of cooperation, and institutional support (Tropp and Molina 2012). Another condition that has been identified in recent research is the importance of cross-group friendships (Pettigrew 1997). That is, the person who has a close friendship with a member of another ethnic group is more apt to have less prejudice and more liking for that group as a whole (Pettigrew and Tropp 2006; Davies et al. 2011).

Perhaps even more interesting, particularly in a multiethnic society such as ours, is evidence for a secondary transfer effect (Pettigrew 2009b). This effect is said to occur when contact with a member of a primary outgroup (for example, of whites with blacks) affects attitudes toward members of a different ethnic-racial group. In one recent exploration of this process, the attitudes of whites and blacks toward Indian and Mexican immigrants were assessed to see if they might be influenced by the extent of contact whites had with blacks, and vice versa (Jones-Correa et al. 2018; Marrow et al. 2018). As might be expected, the degree of direct contact that both whites and blacks had with each of the two immigrant groups was a primary determinant of their attitudes toward those groups. But in addition, mainly among U.S.-born white respondents, more frequent contact with blacks was associated with more welcoming attitudes toward both Indian and Mexican immigrants. Among U.S.-born black respondents, the pattern was more limited, showing evidence for secondary transfer effects for attitudes to Mexicans but not to Indians. Given differences in status positions and possible similarities or dissimilarities between particular combinations of groups, much more work is needed to map out exactly how primary and secondary transfer effects occur. It is clear, however, that in a multiethnic society, both history and current conditions can shape the course of intergroup relations and their consequences for identity development and change among immigrants and the native born.

The interactions that immigrants have with the U.S.-born groups they encounter have implications for their American identity as well. As Michael Jones-Correa and his colleagues discuss in their article in this issue, specific characteristics of the interactions matter (2018). For Mexican immigrants, interactions with both blacks and whites influence their identification as Americans; for Indian immigrants, by contrast, identifying as American is influenced only by the nature of their interactions with whites.

An additional caveat on the influence of intergroup interactions on identity patterns is a reminder that it is often the perception of conditions, rather than an objective tally, that shapes subsequent attitudes and behaviors. As Craig and Richeson report, not only does the actual percentage of racial minorities in a community influence whites' estimates of the discrimination that they and members of their group face, but so do their (not necessarily accurate) perceptions of those proportions (2018). Clearly the dynamics of intergroup contact are complex, characterized by numerous possible combinations of racial-ethnic groups and conditions of contact, not all of them marching in lockstep together. Yet, although overall consistency of patterns cannot be automatically assumed, evidence for the impact of interaction on identity processes is already convincing.

Panethnic Identities

The rise of Hispanic or Latino and Asian panethnicity, as noted, is a modern-day development, in that multiple ethnic groups have widened their boundaries to forge broader groupings and identities. Although Latinos and Asians remain strongly attached to their more specific ethnic or national identities, individuals allied with each panethnic group also often come together under the banner of panethnic identities, particularly in situations where cooperation allows them to exert political influence to common advantage. In general, as the historian George Fredrickson observes, to the "extent that the white or Anglo majority, nationally or locally, treats either Asians or Latinos as a single group and acts in ways that affect all or most of those so designated, panethnic identities . . . tend to emerge" (Foner and Fredrickson 2004, 7).

How are panethnic identities and intergroup relations connected? In the first place, relations among ethnic groups in the Asian or Latino panethnic category can influence panethnic identities. On one side, regular, positive interactions among members of different national-origin groups can encourage or fortify panethnic identities. This may be especially likely in institutional settings and local neighborhoods that have no dominant majority ethnic group, as in the Latino community of Corona in New York City's borough of Queens, where repeated interactions among residents in neighborhood arenas—from stores, churches, and senior centers to political groups—fostered an overarching Latino identity (Ricourt and Danta 2003; Espiritu 2013). Whatever the demographic configuration, collective organizing and political mobilization can encourage and strengthen panethnic identities among both Asian Americans and Latinos (see, for example, Okamoto 2014). On the other side are instances where divisive and conflictual relations among ethnic or national-origin groups reduce the salience or strength of panethnic identities, as when these groups compete for resources and desired positions, including political office. In the Sikh American case reported by Prema Kurien, a combination of factors in recent years, including discrimination experienced since September 11 and homeland political conflicts, have led to an emphasis on a Sikh American rather than Indian or South Asian identity, as well as a movement for Sikhs to be classified as a distinctive ethnic group by the U.S. census (2018).

The connection between panethnic identities and intergroup relations also may operate in the other direction: panethnic identities can affect intergroup relations. Most notably this happens when these identities provide the foundation for political mobilizations among ethnic groups, for example, when leaders of pan-Asian or pan-Latino organizations are able to appeal for support on the basis of panethnic allegiances. In other cases, panethnic identities support alliances with other panethnic or racial groups, such as the cooperation between Latino, Asian American, and black caucuses in state legislatures and city councils. More research is clearly needed on the consequences of panethnic identities for intergroup relations in communities and institutions as well as for political participation and organization (for political science analyses of the links between panethnicity and political attitudes and behavior, see Junn and Masuoka 2008; Lee 2008; Wong et al. 2011).

CONCLUSION

The massive immigration of the past half century has given rise to growing racial and ethnic diversity throughout the United States and led to striking shifts in the way that individuals, both immigrants and the native born, identify themselves and others. Indeed, the very categories used to define racial and ethnic similarities and differences, and the meanings attached to the categories, have undergone significant change.

We have focused on the past and present in the preceding pages, but have also speculated about the shape of racial and ethnic categories and identities in the years to come. Will any of the immigrant groups currently thought of as not white but not black, for example, come to be viewed as white—in other words, will the category white widen to include new strands? Or is it misleading to pose the question this way? Just as white meant something different a hundred years ago than it does today, so, too, the category white may become outmoded, or at least less salient, in the years ahead as new ways of thinking about racial and ethnic differences, and new racial divisions emerge, for example, a black-nonblack dichotomy, as some scholars speculate (see, for example, Lee and Bean 2010). In this scenario, Asians and most Hispanics, who start out in an in-between status, neither black nor white, could become part of a new nonblack or beige majority. Other social scientists suggest the possibility of a triracial stratification system, which in Eduardo Bonilla-Silva's version is made up of whites, honorary whites, and "collective blacks" (2004). The latter category, he proposes, would include, among others, dark-skinned Latino immigrants, West Indian and African immigrants, as well as Filipinos, Vietnamese and Laotians; honorary whites would consist of light-skinned Latinos, Japanese Americans, Korean Americans, Chinese Americans, Asian Indians, Middle Eastern Americans, and most multiracial Americans. Whatever the course of change, what seems clear is that some of the forces that operated in the past for the descendants of southern and eastern Europeans are unlikely to recur in the near future, such as an end to mass immigration or a huge expansion of higher education. Intermarriage, by contrast, is likely to be more significant in the coming years. Whereas in the past, Jews and Italians were transformed from races into white ethnics without undergoing alterations in phenotype, today, when the language of color is so prominent in racial discourse, intermarriage and the blurring of pigmentation and physical differences among mixed-race offspring are often predicted to be key agents of change. And though the kind of economic prosperity that occurred in the postwar years is not on the horizon today, the changing demography of the country—with fewer native whites in the working-age population and the labor market as the large cohort of baby boomers retires—is bound to create opportunities for some of the descendants of the post-1965 immigrants to move up the occupational ladder (Alba 2009; Alba and Foner 2015).

As we look ahead, additional research questions that can enhance our understanding of the links between immigration and racial and ethnic identities emerge. One broad question concerns how characteristics of different cities and other localities affect patterns of identity formation. Place matters, and a range of contextual features of cities—including the racial, ethnic, and class composition, and size of immigrant-origin as well as native white and minority populations and their relative political and economic standing—can influence the development of ethnic and racial identities. Another issue is how much it matters that one national-origin group may overwhelmingly dominate the immigrant population, as Mexicans do in many cities, and that a large proportion of them are also undocumented. Does being Mexican mean something different in Los Angeles, where Mexicans are around 40 percent of the immigrant population, versus New York City, where they are only about 6 percent (Foner 2005; on views of Latinos in the U.S. South, see Brown et al. 2018). To take another example, the Haitian case suggests how other factors in an urban area—historical events shaping immigrant inflows and residential patterns—can affect the way an immigrant population is perceived. Haitians are a more highly stigmatized and visible group in Miami than in their other major area of settlement, New York City, in part because of distinctive features of Miami: the significant number of "boat people" who arrived there in the 1970s and 1980s and Haitians' association with their own distinctive neighborhood, Little Haiti, which has been home to many poorer Haitians as compared to New York City, where Haitians live in the same neighborhoods with English-speaking West Indians (Foner 2016).

To come back to a study mentioned earlier, how widespread are communities like those in Silicon Valley that Tomás Jiménez and Adam Horowitz studied, where traditional status rankings of ethnic and racial groups no longer hold (2013)? In their case study of a community that included only whites and large numbers of highly skilled Asians, whites are associated with mediocrity and Asians with the highest achievement, flipping the relative status of long-established native-born and immigrant groups. In general, what is the impact of living in ethnoburbs, advantaged suburbs where concentrations of well-off Asian families reside, for example, for the identities of those Asian immigrants and their children? And do other non-white immigrant-origin groups also occasionally find themselves near the top rather than near the bottom of their community hierarchy?

As immigrants have dispersed across the country in recent years and become less concentrated in traditional gateway cities, questions have arisen about the way so-called new destinations are distinctive contexts for immigrant incorporation (Singer 2014). These questions need to include ethnic and racial identities. Do whites, to raise just one issue, experience more group threat in new destinations where large numbers of Latino immigrants have recently moved into what had been virtually all-white communities, in contrast to traditional gateways that have long been accustomed to immigration and ethnoracial diversity? Do we need to consider the rate of change in a community as well as absolute numbers to fully understand the ways in which threat might operate? Also of relevance in the present political climate are the effects of local government policies impinging on unauthorized immigrants. Whereas some cities have developed migrant-inclusive policies, such as establishing sanctuary cities or developing municipal identification cards for use as identification by the unauthorized (de Graauw 2014), other communities have passed restrictive legislation designed to limit unauthorized immigrants' employment and residence or have actively worked with federal immigration enforcement to deport the unauthorized (Armenta 2017; Donato and Rodriguez 2014; Flores 2014; Menjivar and Kanstroon 2013). These responses, whether inclusive or exclusive, are likely to influence the ways that people define and understand their race and ethnicity and that of others.

Whatever the location, another set of questions relates to the range of social statuses and characteristics of those of immigrant origin as they affect the construction of racial and ethnic identities, that is, an intersectional perspective that emphasizes how race and ethnic identities are experienced in interaction with other identities such as gender and occupation (see, for example, Collins 2015). Legal status looms large given that one in four immigrants in the United States is undocumented. The common assumption that most Mexican immigrants are undocumented (only about half are and the number is declining) has contributed, as noted earlier, to negative views of all Mexican immigrants and questions about their right to be in the country. Just as we need to know more about the links between legal status and identities, education, social class, and rising inequality are also potentially important in influencing identities. More than a quarter of immigrants have a college degree, and the figure is higher for their U.S.-born children. How do educational and occupational achievements shape how Asian, Latino, and black immigrants identify themselves and how others identify them? These same issues of class and education can be raised for white identity as well, given emerging research showing links between class position and the importance of white identity (Stets and Phares 2016). A further intersection of potential significance is that of gender (Donato and Gabaccia 2015). Are there critical interactions among some of these demographic categories, such that, for example, gender might be more influential when considering some ethnic groups than others?

Finally, returning to the idea of change, we need to ask how identities might vary over the life course and differ by age and life stage. Even within the limited format of the U.S. Census and within a ten-year span, people have been found to change their choice of ethnic label, a finding particularly true for those who use a combination of categories at one of the two time points (Liebler et al. 2017). In this issue, Cynthia Feliciano and Rubén Rumbaut's research indicates that for some adult children of immigrants ethnic attachments become less central over the life course (2018). Other research shows how ethnic and racial identities vary between the first and second generation—for example, members of the U.S.-born second generation are more likely to adopt a hyphenated American identity than their immigrant parents and to be less attached to the national identity associated with their parents' country of origin (Portes and Rumbaut 2001, 165)—but we still have much more to learn about generational differences.

These are just some of the many questions that call for further study. As immigrants continue to enter and settle in the United States; as a huge second, and now third, generation descended from post-1965 immigrants grow up and take their place as adults in American society; and as political developments pose unpredictable challenges, understanding how immigrants and their descendants affect, and are affected by, the meanings attached to ethnoracial differences and their place in ethnoracial hierarchies are topics that should be high on our research agenda.