Effects of Immigration and Assimilation
An important concern in immigration research involves the effects of immigration and assimilation on health, education, and social programs, particularly in areas of high immigration concentration. Much folk wisdom has viewed assimilation as a linear process of progressive improvement and adjustment to American society. The general assumption is guided by an implicit deficit model: to advance socially and economically in the United States, immigrants need to "become American" in order to overcome their deficits in the new language and culture. As they shed the old and acquire the new, they acquire skills for working positively and effectively—a process that may not be completed until the second or third generation after entry.
Today's immigration is overwhelmingly composed of newcomers from Asia and Latin America, areas with significantly different languages and cultures than those of previous European immigrants in the late 1800s and earlier decades of the 1900s. Concerns have been raised about the speed and degree to which these immigrants can assimilate—and hence about the social "costs" of these new immigrants—before they begin to produce net benefits to their new society. The traditional assumption is that immigrants have costs to U.S. society in the initial period after arrival, but that the costs decrease and the benefits to society increase as duration of residence increases. It is further assumed that the benefits to society also increase with greater assimilation to American culture. Recent research findings, however, especially in the areas of perinatal health, mental health, and education, raise significant questions about such assumptions. Indeed, some of the findings run precisely opposite to what might be expected from traditional notions and theories of assimilation.
This chapter captures the workshop discussions of the effects of immigration and assimilation on social policies and programs, health, and education.
Social Policy and Welfare1
Immigration researchers disagree about many major issues that are essential for revising social policy, including the criteria used to admit immigrants and the extent of social supports required to ensure their successful integration. More specific areas of disagreement include: whether recent arrivals are less skilled than earlier arrivals; whether the pace of socioeconomic assimilation has slowed in recent years and, if so, why; whether the net social and economic impacts of immigration are positive or negative; which social groups and communities are the net beneficiaries (or losers) from the influx of new immigrants; whether legal immigrants, illegal aliens, and refugees face dissimilar prospects for integration in the United States and, if so, why; and whether the criteria currently used to admit immigrants are optimal for achieving social, political, humanitarian, and economic objectives. All analysts agree that reliable answers to all of these questions are necessary for future policy initiatives concerned with employment, schooling, and income maintenance.
Despite the many areas of disagreement among immigration experts, there is widespread consensus on three issues: (1) the volume of immigration is likely to increase over the next decade, (2) the demographic and socioeconomic diversity of the flows has increased in recent decades, and (3) currently available data are ill-suited to address adequately many policy-relevant questions about how immigration contributes to contemporary patterns of stratification.
Employment and Income Dynamics
One of the most serious deficiencies in the area of immigration and economic inequality is the absence of information about income and employment dynamics among various segments of the foreign-born population. Virtually all national estimates of immigrant employment, poverty, and welfare participation are based on data from the decennial census or the Current Population Survey. Although static measures of poverty status and welfare participation are useful for portraying aggregate trends and differentials in the prevalence of poverty in a given year, they do not illustrate the dynamics of income stratification processes. These tasks cannot be accomplished with currently existing data because administrative records on program participation seldom include nativity identifiers, and because nationally representative longitudinal or cross-sectional surveys seldom provide sufficient detail on type of program participation, much less duration of
episodes. Although the Survey of Income and Program Participation is suitable to address these questions and others about income and employment dynamics, items about immigrant status are now available only on the topical modules (i.e., questions on selected special attachments to the main questionnaire).
A review of the Survey of Income and Program Participation by a panel of the Committee on National Statistics (Citro and Kalton, 1993) recommended a number of changes for improving the survey. Because it is the preeminent source of survey data on the use of public services, information from the survey has great potential for contributing to current debates about the use of welfare, medical care, and other social services by immigrants. But to serve current policy analysis requirements, information is needed on potentially illegal statuses—a difficult challenge for any survey research. Workshop discussion did not address problems of such data collection, but such enhancement of the survey is worth further consideration. A further limitation of these data are the relatively small sample sizes of the Asian and Hispanic populations, which preclude detailed analyses of specific nationality groups.
Further advances toward understanding the process of socioeconomic integration of immigrants require a longitudinal analysis of employment and income dynamics. This is essential to determine if rising inequality among various groups of immigrants and their native-born counterparts results from greater numbers experiencing transitory or chronic episodes of joblessness, poverty, and welfare dependence. Studies of employment and income dynamics among immigrants should also help to clarify inconsistencies in current research regarding the relationship between length of U.S. residence and economic well-being. Longitudinal analyses of income and program participation among the foreign-born population are a necessary adjunct to policy because the program implications of transitory episodes of poverty and welfare participation differ appreciably from chronic dependence.
The Context of Immigration
Contextual analyses of immigrants' integration experiences are an important area of needed information. In practical terms, this means that future national surveys of immigrants should not only permit subgroup analysis, but should also represent the social and economic spectrum of communities in which immigrants reside. Whereas assessments of economic well-being based on national samples are worthwhile for broad generalizations about income inequality among nationality groups, they are inadequate for portraying the contexts within which economic integration processes unfold. Widely discrepant conclusions about the extent and nature of labor market competition between native-born and immigrant people illustrate the need to reconcile findings based on specific labor markets and those based on nationally representative analyses. In fact, the high
concentration of immigrants residing in a handful of large cities raises questions about the usefulness of analyses based on national populations.
The context for immigration involves the entrance and exit of immigrants. It is relatively easy to see the excellent opportunity for contextual studies presented by a case in which migration takes place and immigrants settle within an ethnic community. Contextual studies are also important, however, when what is called the "quality" of immigrants is being studied. George Borjas compared recent immigration flows with those prior to 1965 and found a declining quality of immigrants in terms of assimilation and productivity. But the quality of an immigrant should be related to more than wages. Immigrants who came before 1965, many of whom were Europeans, came during a period of lower rates of immigration. Recent flows are different. Education levels of immigrants vary, and the averages need to be used in context for good analyses to be done. An illustration of the importance of context is the case of Haitian children enrolled in poorer schools in Miami's inner city. The education and assimilation experiences of these children might have been more positive if they had not settled in Miami. In summary, the context of immigration is important in research.
Comparing Political and Economic Immigration
Because systematic comparisons of political and economic migrants have not been undertaken, a third important area is improving understanding about whether and how the integration experiences of refugees and legal immigrants differ. Refugees undertake politically motivated migration, whereas immigrants have economic motivation, according to a perspective taken by some. Although the distinction between political and economic migrants has been greatly overstated, there is little disagreement that the reception experienced by these two classes of immigrants is dramatically different. Existing research is inconclusive about the effects of resettlement assistance; it is not clear if such assistance facilitates or retards economic assimilation. A useful experiment to resolve this key policy question would compare two similar cohorts of immigrants who arrived at the same time from the same country. The data needed to conduct even this simple exercise, which is fundamental for assessing the effects of resettlement assistance, are not available. Yet this exercise is particularly critical in the current climate of fiscal retrenchment. Between 1980 and 1991, the federal government appropriated over $5 billion to various forms of resettlement assistance, but during the past five years the appropriations for refugee programs have been slashed. The reasoning behind sharply curtailing appropriations for resettlement assistance for refugees, as opposed to extending some form of resettlement assistance to all economic migrants, rests on a thin research base.
Effects of Amnesty
In light of the increased attention during the 1970s and 1980s to legal status among the foreign born, it is imperative to investigate how immigrants granted amnesty under the provisions of the 1986 Immigration Reform and Control Act are faring relative to other legal immigrants and native citizens. Nearly 3 million illegal aliens were granted legal status between 1987 and 1988, of whom more than 85 percent were from Mexico. Despite great interest within the policy research community in the effectiveness of employer sanctions and tighter border controls, there have been no comparable research initiatives to investigate the experiences of legalized immigrants. How well is the legalized population faring in the labor market relative to other groups of immigrants? Did the change in legal status influence employment and welfare behavior? Although there has been much speculation about likely changes following the amnesty program, research initiatives have not matched the speculative curiosity.
Until recently, no data were available to investigate research questions about the behavior of new immigrants under the legalization program. However, the Legalized Population Survey, conducted in 1986 by Westat, Inc., under contract to the Immigration and Naturalization Service, provides a unique opportunity to analyze the income and employment dynamics of recently legalized immigrants. It is a nationally representative survey of immigrants granted amnesty under the 1986 Immigration Reform and Control Act. A second part of the survey was in the field in 1992 and should provide additional valuable data. This survey should provide essential information about changes in employment and program participation, including the use of several in-kind programs (such as food stamps) that might be traced directly to a change in legal status. Analysis of these data is a high priority for evaluation of the behavioral consequences of legalization on welfare participation.
Workshop discussions identified four areas in which better information is needed for the improvement of studies of federal programs and immigrant adjustment. First, improved data are needed about income and employment dynamics. The Current Population Survey could benefit from special-purpose modules that include retrospective questions on changes in economic status. For the Survey on Income and Program Participation, it would be helpful to have a question on immigrant status included in an early wave of the interviewing and to include contextual variables in the survey data.
Second, comparative studies are needed on poverty and economic change for immigrants in different areas and cities. Workshop discussion suggested that it would be useful to have a set of comparative studies on immigrant adjustment, conducted with common variables, for a variety of metropolitan areas.
Third, more studies are needed of the different types of immigrants in order to note the comparative effect of assistance on economic adjustment. Refugees (political immigrants) are eligible for different federal assistance programs than economic migrants, who enter the United States based on scrutiny of their ability to gain successful employment.
A final area that warrants attention is the effect of the legal status, especially legalization, on immigrant adjustment. A substantial proportion (probably one-fifth or more) of the current foreign-born population entered the United States illegally during recent decades, and many of these illegal aliens are seeking legal status under the general and special agricultural workers provisions of the Immigration Reform and Control Act. Comprehensive studies are needed of the adjustment of this newly legalized population, compared with immigrants who entered legally.
Research is needed to improve our understanding of an important, contemporary public health enigma: the apparently better-than-average pregnancy outcomes among immigrant groups, regardless of socioeconomic status. Current health data on specific immigrant groups are limited (national-level vital statistics data lack information on immigration status), and immigrants' ethnic groups are often reported only for pan-ethnic categories (Asians and Hispanics). Still, pregnancy outcomes as measured either by birthweight or mortality are better among babies born to immigrant than to native-born mothers (Eberstein, 1991). Similar results have been reported for Spanish-surname mothers in California (Williams et al., 1986). Guendelman et al. (1990), using data from the Hispanic-HANES survey, found that low-birthweight rates were significantly higher for second-generation native-born women of Mexican descent compared with first-generation Mexico-born women, despite the fact that the latter population had a lower socioeconomic status, a higher percentage of mothers over 35 years of age, and less adequate prenatal care. The risk of low birthweight was about four times higher for second-generation compared with first-generation primiparous women, and two times higher for second-generation compared with first-generation multiparous women. Earlier, Yu (1982) reported that Chinese-American women have lower fetal, neonatal, and postneonatal mortality rates than women of European origin and those in other major ethnic and racial groups in the United States. Yu also reported that the superior health profile of Chinese-origin infants was observed at every level of maternal education and for all maternal ages.
Research in California over the past decade has found that infant mortality
rates for recently resettled Southeast Asian refugees (especially Vietnamese and Cambodians) were significantly lower than those for the non-Hispanic white population (Rumbaut and Weeks, 1989; Weeks and Rumbaut, 1991). The results are noteworthy because the Southeast Asians had the highest rates of poverty and fertility in the state, had experienced very high infant death rates prior to their arrival in the Unites States, lacked English proficiency, and had the latest onset of prenatal care of all ethnic groups. Other Asian groups (Japanese, Chinese, and Filipinos) and Hispanics (mostly of Mexican origin) also had lower infant death rates than whites, and much lower rates than those observed for Native Americans and blacks. The groups with below-average infant mortality rates consist largely of immigrants.
The evidence indicates that positive perinatal health outcomes among immigrant groups are a real phenomenon, worthy of further investigation. Are immigrant women superior health achievers, even when socioeconomic status is controlled and, if so, why? What are the effects on pregnancy outcomes of a wide variety of sociocultural and biomedical risk factors for foreign-born and native-born women of diverse ethnic and racial groups? Although there are significant differences by nativity and ethnicity in pregnant women's histories of smoking, alcohol, and drug abuse during pregnancy—behaviors that are deleterious to the infant's health at birth and that appear to be more prevalent among the native-born—such variables do not explain other independent effects of nativity and ethnicity on outcomes. There is considerable complexity to carrying out research in an area in which immigration, assimilation, and health interact. Existing vital statistics by themselves will not provide the research answers; alternative sources of data are needed and should include qualitative information as well as new studies based on comparative longitudinal designs (e.g., identifying immigrant and native-born women of different socioeconomic and ethnic groups early in their pregnancy and following them through the first year of the newborn's life). If we are to add significantly to the store of knowledge and to develop a larger set of intervention options, such research and data are essential.
Intriguing questions have been raised by research on the mental health of ethnic minorities in the United States, including immigrants. In a review of mental health prevalence rates reported in research over the past two decades (Vega and Rumbaut, 1991), studies suggest that rapid acculturation does not necessarily lead to conventionally anticipated outcomes, i.e., that improved adjustment to American society and a decrease in the mental health problems are associated with immigration. Instead, mental health studies suggest that assimmilation—in the various forms it can take—can itself be a traumatic process rather than a simple solution to the traumas of immigration.
For example, results from the Hispanic-HANES study (Moscicki et al., 1989),
with an exceptionally large regional sample, indicate low symptom levels of mental health disorders for Mexican-Americans in the southwestern United States and significantly lower rates of depressive symptoms and major depression for Cubans in Miami, compared with all other Hispanic groups. The Los Angeles Epidemiological Catchment Areas study also reported lower rates of major depression among Mexican-Americans than among non-Hispanic whites (Karno et al., 1987). Significantly, among Mexican-Americans, immigrants had lower rates of lifetime major depression than native-born people of Mexican descent; and among Mexican immigrants, the higher the level of acculturation, the higher was the prevalence of various types of psychiatric disorder (Burnam et al., 1987). Furthermore, the native-born Mexican-Americans and non-Hispanic whites were much more likely than immigrants to be drug abusers.
Other suggestions for future research emerged from the workshop discussions. Research should take the social and historical contexts of immigrants fully into account, in terms of entries, exits, and assimilation. And among nonimmigrant ethnic and racial groups, studies need to distinguish between different American-born generations (how many generations have passed since the immigration?) and conceptualize categories of race and ethnicity as social processes, rather than fixed, purely ascriptive categories. Moreover, research is needed to identify protective factors that appear to reduce mental health problems within diverse ethnic minority groups; recent findings show that certain immigrant groups exhibit lower symptom levels of psychiatric disorders than do majority group natives. Longitudinal studies are especially needed to characterize and investigate stress and its temporal patterning among immigrant groups, including patterns of immigrant adaptation to specific conditions of life change and their psychological or emotional sequence. And, given the unprecedented racial and ethnic diversification of the U.S. population as a result of sharply increased immigration from Asia and Latin America, research is needed to investigate the mental health consequences of racial and ethnic discrimination, and how different groups (especially first and second immigrant generations) perceive and react to discrimination.
The rapid surge of recent immigration has been accompanied by a rapid growth in the research literature on the educational attainment of immigrants; the research has concentrated predominantly on the educational levels of adult immigrants of working ages. Relatively little study has been given to the educational achievements of the U.S.-born second generation—the sons and daughters of immigrants—despite the fact that immigrant children are a highly visible presence in the schools now and they will represent a sizable component of the next generation of U.S. residents. The patterns of their educational attainment, language shift, and psychological adaptation cannot be predicted on the basis of their
parents' performance, nor from the experience of earlier waves of large-scale immigration. Research on the children of immigrants poses significant but so far unanswered theoretical and empirical questions. What factors account for variations in successful English-language acquisition for the children of immigrants? What is the role of family factors (encouragement of regular study and the setting of education and occupation goals, for example) for educational attainment?
Available results from the limited studies available are suggestive. In a study of students in the San Diego high schools, lower grade point averages were noted for Hispanics, Pacific Islanders, and blacks than for all other students. With the exception of Hispanics, immigrant minority students from non-English-speaking families had higher grade averages than either majority native-born students or immigrant minority students from English-speaking families. The highest grade point averages were those of students in immigrant families from China, Korea, Japan, Vietnam, and the Philippines. More remarkably, Hmong students (whose parents are largely illiterate peasants from the Laotian high-lands) and Cambodian students (whose parents are mostly from poorly educated rural areas) were outperforming the average native-born student. Other research has reported similar findings among immigrants of lower socioeconomic status from Central America, Southeast Asia, and the Punjab, and similar studies have found that Mexican-born immigrant students do better in school and are less likely to drop out than U.S.-born students of Mexican descent, despite the comparatively greater socioeconomic disadvantages of less assimilated foreign-born people (Rumbaut, 1990).
Studying the adaptation process of immigrant children—patterns of language acquisition, educational attainment, cultural and psychological adjustment, ethnic identity, and acculturation strategies—can best be approached through comparative longitudinal research designs in a variety of community contexts, supplemented by intensive ethnographic field work. Parental socioeconomic status and individual human capital can certainly be expected to have a strong effect on every aspect of the adaptation process, but those characteristics and related demographic variables cannot by themselves provide a completely satisfactory explanation. For that purpose, existing data sets are not adequate to the research tasks. Data are needed on the different contexts of reception and incorporation facing different immigrant groups, including the presence or absence of discrimination and the character of the ethnic communities in which immigrant children are raised. Such data would be further enhanced if information on schools and the school environment were collected.
In studying the effects of immigrants on American society, as well as the effects of American society on immigrants, better information is needed in five areas:
- The predominant portion of immigration studies has focused on the problems arising from immigration. Studies are needed that examine the overall effects of immigration, not just the negative impacts.
- There is a difference between cultural assimilation (e.g., learning English and feeling at home in American society) and structural assimilation (e.g., achieving economic success). In addition, assimilation is a ''segmented'' process, depending on the subculture of American society in which different immigrant groups reside (e.g., ethnic enclaves, segregated inner cities, white middle-class suburbs). Several aspects of assimilation are essential to study: taking on aspects of the destination community, adaptation to new social and economic characteristics (compared with those of the country of origin), and integration into the destination community. Cultural assimilation does not necessarily lead to structural assimilation. There is a need to study the relationship of cultural and structural adjustment in more detailed studies of nationality groups than has been done to date.
- Available studies have examined changes by age groups of immigrants, but data have been missing on the temporal and local-area contexts of individual assimilation. Further studies (similar to Tienda, 1992) are needed on immigrants and labor markets, with data on contextual aspects, temporal shifts, and labor market differentials.
- Available studies suggest that immigrants have lower mortality and morbidity compared with the native-born U.S. population. Fuller explanation of mortality and morbidity adjustment requires improvement of data on multiple causes of death, duration of residence of immigrants in the United States, and the residential context. These data, however, may be expensive to collect if they begin as new data collection systems; more study is called for on the benefits of such studies, relative to their costs.3
- For studies of the well-being of immigrants and their children, it is critical to have data on two items. First, country of origin is important because some immigrants originate in conditions with high mortality: survivors of high mortality are quite selective and may be seen as healthier in their years after arrival in the United States. Second, the local context of their destination community can influence health outcomes. Information on conditions for both the originating and destination communities are needed for interpreting health data.
Workshop participants observed that there is a need for further research on immigrant adjustment and the policies necessary for improved adjustment by
immigrants. Policy analysis requires improved information on, for example, the speed of adjustment to jobs, English language abilities, fertility changes, and individual endowments and community context.
Workshop discussion suggested that additional research is needed on immigrant assimilation and federal programs. First, workshop participants emphasized the need for contextual analysis, work that takes into account situations in which immigrants differ by type of entry and type of environment. Second, a moderate proportion of immigrants return to their countries of origin. Studies of return migration could provide useful insights into assimilation (or, in some cases, lack of assimilation). Third, different types of immigrants face different eligibility rules for welfare participation. Useful comparative studies of recent immigrants could be conducted that take advantage of the natural variation in welfare eligibility. Fourth, the visa category of immigrant entry is important for policy studies on the effects of immigration because the characteristics of legal immigration are affected by the number of visas issued. The decennial census and the Current Population Survey are not appropriate for collecting immigration status, however, because they are self-administered (respondents often do not know their specific immigrant status) and questions on immigration status on the Current Population Survey could affect the collection of employment data. Expanded data on immigration status could be collected better on special surveys or in conjunction with linked Immigration and Naturalization Service administration records. Fifth, most new immigrants in recent decades are members of racial and ethnic minorities. This introduces a new and complicated context for immigration studies, with the requirement for information on racial identity in conjunction with the analysis of other immigrant characteristics.
In the absence of a longitudinal survey of immigrants that would permit the estimation of duration models, the decennial census and the Current Population Survey are the primary instruments for analyzing the impacts of immigration. Modest revisions to both instruments with data on place of birth, citizenship, and year of arrival would greatly enhance the range of possible analyses.
It would also be worthwhile to add questions on immigration status to the core questionnaire for the Survey of Income and Program Participation, rather than limiting these questions to the topical modules. However, simply distinguishing immigrants from natives will not further the understanding of integration processes unless additional questions about immigration histories (especially the first and most recent arrival) are included as well. The Survey of Income and Program Participation is uniquely suited to examine employment and income dynamics over short durations, but it would be less successful in portraying long-term experiences of successive cohorts of immigrants, even if sample sizes were
sufficient for subgroup analyses. Furthermore, contextually based analyses are virtually impossible with the Survey of Income and Program Participation.
To aid in monitoring the self-sufficiency of refugees, the Office of Refugee Resettlement in the U.S. Department of Health and Human Services has conducted a national survey of Southeast Asian refugees, which has proven invaluable for monitoring the economic progress of political immigrants. The Annual Survey of Refugees was converted to a longitudinal survey beginning with the 1984 interviews, tracking a randomly sampled group of refugees over their initial five years in the country. The survey permits comparisons of refugees arriving in different years and hence allows an evaluation of the relative influence of changing conditions of the period on the process of economic and social integration.
The survey would be strengthened if two changes were made. First, the length of time refugee families are followed should be extended from the current 5 to at least 10 years. This is necessary because, at least in California, a significant share of the refugee population had not exited welfare after five years of U.S. residence. Because many refugees remained dependent on welfare at the end of the study period, the data analysis is limited by the small number of refugees who have made the transition to work and adequate income. Second, it would be useful to include other entrants (such as Haitian and Cuban [Mariel] "entrants," even though they did not enter the United States as refugees) in the Annual Survey of Refugees so that their adjustment experiences can be compared more systematically with those of Southeast Asian refugees.
Federal programs to assist immigrants economically began in the early 1960s with efforts to aid refugees from Cuba. These programs have continued, with an emphasis on providing economic support to refugees. Given the national interest in programs to deal with the economic situation of immigrants, the lack of data on the incidence and prevalence of poverty among the foreign-born population is a serious deficiency.
It is important to note that the adjustment of immigrants differs for legal immigrants, illegal aliens, and refugees—each of whom has different social and economic characteristics and different eligibility for federal and state welfare programs. Refugees in California, for instance, seem to remain on welfare longer than other immigrants. In contrast, the welfare participation of aliens legalized under the provisions of the Immigration Reform and Control Act seems to be comparatively low.