Conclusions and Recommendations
In assessing the health, well-being, and adjustment of children in immigrant families, the committee faced from the outset a severe lack of research and data, which impaired our ability to answer many questions related to our charge. The federal government funds a rich and diverse set of data collection systems to monitor the health and well-being of the U.S. population, but few of these efforts seek information about country of birth, parents' country of birth, or citizenship—information that is needed to ascertain the circumstances of first-, second-, and third- and later-generation children. Although available data are thin, we were able to commission original research with existing data to draw preliminary conclusions on critical issues, and to provide the foundation for important recommendations for improved data collection and new research.
Recent and continuing changes in public policy posed additional constraints. First, there are no currently available data on the effects of the recent major reduction in benefits to immigrant children and families brought about by enactment of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. Second, the political environment that will influence the implementation of these new policies is in flux. Most states are still sorting through various decisions that need to be made regarding welfare reform and children's health insurance. Third, the wide
discretion that states have in making decisions on the eligibility of foreign-born children and adults for benefits is likely to produce much variation in the circumstances of these children among—and even within—states.
For these reasons, we recognized that the health of children in immigrant families, their exposure to potential risk and protective factors, their access to publicly funded benefits and programs, and their use of health care services may change markedly in the years ahead. Such changes may result both from the effects of policy reforms for children in immigrant families now residing in the United States and from the effects the policies may have on future immigration. An understanding of previous practices and policies that have provided for foreign-born children and their parents therefore represents critical baseline information on which an accurate assessment of the effects of the recent policy reforms will necessarily depend. The findings and conclusions of the committee are preliminary indications of how children in immigrant families are faring and what this implies about their future prospects. The committee's recommendations focus on data needs that are essential both to a more comprehensive understanding of the health and development of children in immigrant families and to efforts to assess the effects of recent policy reforms on them.
FINDINGS AND CONCLUSIONS
Whenever possible, we compare children in immigrant families with U.S.-born children in U.S.-born families. This provides a means of deciphering generational patterns of health and adjustment and examining immigrant families in the context of the range of life chances experienced by U.S.-born children in U.S.-born families, including black children who, like immigrant children, are often considered to be at risk. The vast diversity that characterizes children in immigrant families can be considered the most salient and most important theme that surfaces from the committee's review and synthesis of this fledgling area of inquiry.
Risk and Protective Factors
Research on risk factors and, to an even greater extent, protective factors specific to children in immigrant families is extremely sparse. Little is known about the conditions or circumstances that are associated with negative outcomes for these children, and less is known about conditions that reduce the likelihood of negative outcomes, whether conceived as the absence of risk factors or the presence of influences that act in their own right to protect development. Nevertheless, a few conclusions that point to interesting questions for future research may be drawn from the existing literature.
In 1990, first-generation immigrant children were substantially more likely to be poor than was the case in 1960 (32.9 percent versus 22.7 percent). In 1960, first- and second-generation children were less likely to be poor than third- and later-generation children. By 1990, first-generation children, and to a much lesser extent, second-generation children, were more likely than third- and later-generation children to live in poverty.
Children in immigrant families and those with U.S.-born parents are about equally likely to have fathers and mothers who are college graduates. At the other end of the spectrum, however, first- and second-generation children are more likely than third- and later-generation children to have fathers with very low educational attainment. Between 1960 and 1990, however, the share of children living with fathers who had completed fewer than eight years of schooling declined by almost two-fifths for the second generation and by one-sixth for the first generation.
• Throughout the century, the vast majority of children in immigrant families have had fathers in the labor force, making them comparable to U.S.-born children with U.S.-born parents. Accordingly, poverty and low parental education are not closely associated with lower rates of father's employment in immigrant families. Children across immigrant generations have also historically not differed in their likelihood of having a mother in the labor force, although enormous increases in maternal employment have occurred for all children over the course of the century.
To the extent that differences in parental employment are found for children in immigrant families, they emerge in fathers' rates of full-time, year-round work, which are lower for the earlier generations, particularly for the first generation.
In 1990, children in immigrant families, especially the second generation, were more likely to live in two-parent families than were U.S.-born children with U.S.-born parents. This was not the case at the turn of the century or in 1960, when first-, second-, and third-generation children were about equally likely to live with two parents.
In both 1910 and 1990, the vast majority of children in immigrant families lived in a household in which their parents did not speak English at home. The proportion of children who speak English, however, increases rapidly from the first to the second generation, such that, in 1990, 81 percent of second-generation children spoke English "exclusively or very well." Yet linguistic isolation remains a concern. In 1990, a sizeable number of immigrant children from several of the major sending countries, including over 40 percent of children from Laos, Cambodia, Vietnam, China, and the former Soviet Union, lived in households in which no person age 14 or older spoke English either "exclusively" or "very well."
These general patterns, however, camouflage the diversity in socioeconomic circumstances that characterizes children in immigrant families from different countries of origin. Children in immigrant families from 12 countries of origin, which account for close to half of all such children, now experience poverty rates that are comparable to those experienced by Hispanic and black children who are U.S. born with U.S.-born parents. These countries include those that are the sources of many officially recognized refugees (the former Soviet Union, Cambodia, Laos, Thailand, and Vietnam) and that provide most legal and illegal unskilled migrant laborers (Mexico, Honduras, Haiti, and the Dominican Republic). Children from most of these countries had working fathers, and most lived in two-parent families. These children were also distinguished, however, by their relatively high rates of living in linguistically isolated households and of not speaking English exclusively or very well. Children with ori-
gins in most of these countries are likely to be classified as minority—Hispanic, Asian, or black—a factor with profound implications for their exposure to both risk and protective factors as they grow up.
Many measures of physical health and risk behaviors that have been reported for children in immigrant families indicate that they are healthier than their U.S.-born counterparts in U.S.-born families—a finding that is counterintuitive in light of the minority status, overall lower socioeconomic status, and higher poverty rates that characterize many of the immigrant children and families that have been studied. Evidence on this issue is patchy, however, focusing on some immigrant groups and some age groups and frequently relying on parental reports rather than direct medical examinations; the research that exists, however, is quite consistent. Of concern are indications that the relatively good health of children in immigrant families appears to decline with length of time in the United States and from one generation to the next. Moreover, some children in immigrant families appear to be at risk for certain health problems (e.g., drug resistant tuberculosis) that, if left untreated, could have adverse implications for the health of the entire U.S. population.
Specifically, for many of the groups of immigrant mothers that have been studied, children born in the United States to foreign-born mothers are less likely to have low birthweight or to die in the first year of life than are children born to U.S.-born mothers from the same ethnic group, despite the generally poorer socioeconomic circumstances of the immigrant mothers for many specific countries of origin. Immigrant parents report that their children experience fewer acute and chronic health problems compared with third- and later-generation families. And adolescents in immigrant families report lower levels of neurological impairment, obesity, asthma, and health risk behaviors such as early sexual activity; use of cigarettes, alcohol, marijuana, or hard drugs; delinquency; and use of violence compared with their counterparts with U.S.-born parents.
The neonatal and adolescent health advantages of immigrants
appear to deteriorate over time as environmental conditions impinge on development and as youth become increasingly removed from their immigrant origins and assimilated into the youth culture—and often the minority youth culture—in the United States. These data raise the intriguing possibility that children in immigrant families are somewhat protected, albeit temporarily, from many of the deleterious health consequences that typically accompany poverty, minority status, and other indicators of disadvantage in the United States. For example, the comparatively low levels of cigarette smoking, alcohol consumption, and drug use during pregnancy; a more healthful diet; and highly supportive family networks may all play a role in the positive neonatal outcomes of babies born to immigrant women in the United States. A central challenge involves identifying the factors—genetic, familial, behavioral, environmental—tied to the countries of origin that play into these protective functions. Although the limited available evidence suggests that protective factors specific to immigrants may diminish with increasing duration of residence in the United States and across generations, future research should assess the possibility that the increasing size and the geographic concentration of the immigrant population, especially from Mexico, may help to shield against the loss of the protective factors they brought with them from their home countries.
Despite this generally positive portrait, not all conclusions that can be drawn about the health of immigrant children are favorable. Children in immigrant families from Mexico, for example, are more likely than third-generation white children to be reported by parents to be in poor general health, to have teeth in only fair to poor condition, and to exhibit elevated blood lead levels. In addition, epidemiological evidence as well as physician reports indicate that children of recently arrived immigrants, and particularly those from certain high-risk countries of origin, are at elevated risk of harboring or acquiring tuberculosis, hepatitis B, and parasitic infections and of having unsafe levels of lead in the blood. Exposure to pesticides is an additional health risk of great concern for children of migrant farmworkers, in light of its documented links to specific ailments and chronic health conditions.
Adolescents in immigrant families appear as likely as third- and later-generation adolescents to experience feelings of psychological well-being and positive self-concept and to avoid serious psychological distress that can, in the extreme, contribute to adolescent suicide rates. These positive signs of adjustment are maintained despite perceptions among adolescents in immigrant families—particularly those of Hispanic and Asian origin—that they have less control over their own lives and are less popular with classmates, compared with their third- and later-generation peers. Factors of particular importance to adjustment are living in a family that is mutually supportive and cohesive and in a community that is supportive of the family's resettlement, speaking English well, and being comfortable bridging one's immigrant origins and American culture.
Children in immigrant families also, on average, perform just as well if not better in school than their third- cand later-generation peers. They also have somewhat higher middle school grade point averages and math test scores, although reading test scores in the first generation are lower than for later generations, probably as a result of their poorer English proficiency. Children in immigrant families from different countries of origin differ greatly in how well they perform in school, however. Children from some Asian countries, for example China and Korea, tend to outperform students from European countries, who, in turn, receive higher grades and test scores than those from Mexico. Similarly, Chinese adolescents in immigrant families have grades and math test scores that are much higher than those of third- and later-generation Chinese and white children. But Asian groups, such as the Lao and Hmong, have scores that are well below national norms.
Use of Public Benefits
First-generation immigrant children are more likely than later-generation children to live in families receiving public assistance. This is particularly the case for Mexican, Cuban, other
Hispanic, and Asian children. This pattern derives from the disadvantaged socioeconomic and demographic circumstances of their families, rather than from their immigrant status per se. At the same socioeconomic levels, both first- and second-generation children are less likely than third- and later-generation children to live in families receiving most forms of public assistance.
Second-generation children were nearly identical to third and later-generation children in their likelihood of living in families receiving benefits from major welfare programs. When generational differences in family socioeconomic levels were controlled, second-generation children were actually less likely than third-and later-generation children to live in families receiving public assistance.
Mexican-origin children were no exception to these patterns. Among those at the same socioeconomic levels, the first and second generations were usually less likely than the third and later generations to live in families that relied on public assistance. Among West European-origin children at the same socioeconomic level, the first and second generations were usually about as likely, or less likely, than the later generations to rely on public assistance. Only among Asian, Cuban, and East European children at given socioeconomic levels was the first generation generally more likely than later generations to use public assistance, probably because of the access available in the past to these populations as refugees from Southeast Asia and the former Soviet Union.
This portrait of reliance on public assistance reflects eligibility rules prior to welfare reform, when access to public benefits and programs for children in immigrant families were essentially identical for legal immigrants and citizens. Today, however, many legal immigrants are barred from a range of federal means-tested benefits, including income assistance (TANF) and Medicaid for their first five years in the country. As a result, new analyses examining immigrant families' reliance on public benefits are urgently needed.
Health Care Coverage and Access to Care
Noncitizen children in immigrant families are about three times more likely than third- and later-generation children to lack health insurance, and second-generation children with at least one immigrant parent are about two times more likely than third- and later-generation children to lack health insurance. Both noncitizen children and, to a lesser extent, citizen children with an immigrant parent are also more likely to lack a usual provider or source of health care and, accordingly, are substantially more likely to have gone without a doctor's visit in the previous year.
All children require access to preventive and acute health care for their own well-being, as well as for the protection of the nation's public health. Yet first-generation immigrant children and, to a lesser extent, second-generation children are less likely to receive this care than their counterparts who are U.S.-born children of U.S.-born parents. The measure of ''care" used in this literature is modest, namely whether the child had seen a physician in the past year.
Research now indicates that access to health care is affected in identical ways for children who are immigrants or live with an immigrant parent and those who are U.S. born with U.S.-born parents. All children are more likely to receive recommended health care when they have health insurance and an ongoing connection to the health care system through a usual provider or source of care. Of particular importance is evidence linking expansions in Medicaid coverage that occurred between 1984 and 1992 to substantially reduced odds that a child went without a doctor's visit in the past year for both children who are immigrants and those who are not, but to a larger extent for immigrant children.
Medicaid, in fact, plays a vital role in health coverage for first- and second-generation children in immigrant families, as well as for third- and later-generation children. About 1 in 4 first-generation children, overall, receive coverage from Medicaid. Children in immigrant families who are Medicaid-eligible are more likely to have visited a physician in the previous year compared with uninsured children in immigrant families. And, in large part due
to the automatic eligibility of refugees for Medicaid, Southeast Asian children exhibit very low rates of uninsurance despite their very low socioeconomic status.
As with public assistance in general, it is critical to assess the effects of recent reductions in insurance coverage for children in immigrant families on their receipt of basic preventive health care and care for health problems.
RESEARCH AND DATA COLLECTION
To be effective, public policies designed to foster the adjustment of children in immigrant families to American society must be founded on rigorous scientific knowledge about their needs, the processes that generate these needs, and approaches to addressing them. Nearly a dozen federal agencies conduct or fund data collection and research efforts that constitute the core of the nation's system for monitoring and understanding the physical and mental health of children, their exposure to risk and protective factors, and their access to and use of public benefits (Box 5-1). Few of these major data collection efforts provide a scientifically sound basis for monitoring or studying the health status and resources available to children in immigrant families. This enormous gap in knowledge is of great concern, given the rapid growth of this population and the pressing need to assess how they are affected by the recent policy reforms that sharply restrict their eligibility for health care and social benefit programs.
Specifically, none of the existing federal surveys and surveillance systems collects the necessary data with a sample designed to allow nationally representative estimates for first-, second-, and third- and later-generation children by detailed country of origin or by immigration status. Similarly, none of the national education data sets, including the three critical ones in listed in Box 5-1, provides the basic data needed to identify children by immigrant generation and country of origin, and only a few research efforts have begun to make use of the capacity to approximate these measures.
Knowledge about the physical and mental health, school progress, access to and use of health and social program benefits and services, and the socioeconomic risk and protective factors
experienced by children in immigrant families could be expanded substantially by additional research using existing data. Small but critical improvements in data collection and in other components of major surveillance systems and surveys could enormously expand the potential for new research that would provide policy makers with information about the growing population of children in immigrant families that is now available only for U.S.-born children in U.S.-born families.
New data collection is also essential to address large gaps in knowledge that persist about the ways in which families, schools, neighborhoods, racial and ethnic stratification, and community and national policies interact with each other, and with different home country cultures and immigration experiences, to influence healthy development and successful adaptation. No single surveillance system or survey currently collects the necessary longitudinal data for children in immigrant (and other) families on physical and mental health, access to publicly funded benefits and programs, health insurance coverage and health service utilization, and exposure to potential risk and protective factors, including those factors unique to immigrants because of their culture and immigration experiences.
In light of these major limitations of existing data, the committee makes a number of recommendations for new research studies, data collection, and information dissemination. Interdisciplinary teams will be necessary to implement many aspects of these recommendations, because the required expertise spans several scientific, medical, and health disciplines.
New Research Studies
Recommendation 1. The federal government should fund a longitudinal survey of children and youth in immigrant families, measuring physical and psychosocial development and the range of contextual factors influencing their development.
Issues and questions pertaining to children in immigrant families currently in the United States, and to children born in future years to these families, will go unstudied without a new longitudinal survey. Information on country of birth for children
BOX 5-1 SELECTED DATA COLLECTION SYSTEMS AND SURVEYS
Health Status, Risk Behaviors, Insurance, and Expenditures
Two central issues that emerged during the committee's review call specifically for longitudinal research. The first pertains to some evidence that suggests that children in immigrant families from at least some countries of origin may, despite negative socioeconomic risk factors, experience physical health outcomes that are superior to those for U.S.-born children in U.S.-born families. The second issue concerns evidence that this relatively good health status declines over time, and engagement in behaviors that pose risk to health increases over time, especially from one generation to the next.
The evidence supporting these broad characterizations is quite limited, however, in four respects. First, this research has focused largely on cross-sectional samples of infants and adolescents, thus limiting efforts to understand developmental patterns over time. Second, issues of selection that have hindered the literature on immigration make it very difficult to ascertain the causes of patterns that have emerged in the research on immigrant children. In particular, possibilities are very limited for considering carefully the contribution of unmeasured differences between individuals who migrate and those who do not, and of differing histories of migration across groups. Immigrants may be self-selected from among the populations in their countries of origin for better health, greater ambition, or other unmeasured factors that account for their comparatively good outcomes in the United States. Third, only a small number of circumstances and outcomes have been measured for children and adolescents in immigrant families. Fourth, little is known about possible causal
or protective factors that account for either the unexpected positive or declining well-being of these children.
If first- or second-generation children experience better outcomes than would be expected from their status on risk factors known to affect the U.S. population, a diverse set of protective factors may be involved. These factors include the social capital or culturally sanctioned behaviors that immigrants bring with them, such as nuclear and intergenerational family cohesion; gender roles; norms about smoking cigarettes, drinking alcohol, sexual activity; and reciprocal responsibilities among members of the community. To the extent that more sophisticated data collection efforts replicate the intriguing patterns to date, it will also be critical to understand whether, why, and which of these protective processes decline over time and across generations and to identify other processes and practices associated with assimilation to American culture that may account for declining health among children in immigrant families.
Knowledge about these processes could, in turn, guide the development of public policies aimed at maintaining beneficial norms and behaviors among immigrant families and promoting similar practices among other U.S. families. Research to develop this knowledge would focus on understanding the dynamics of the processes by which protective factors operate, whether and why they decay or are sustained over time, and the role of the family, the community, and public policies.
This knowledge can be obtained only through new national research of the type recommended here. Essential elements of this research are listed below:
It is longitudinal, following a sample of children over a period of years, in order to map and analyze the causes of change in their development;
It measures a wide range of physical, psychosocial, educational, and other developmental outcomes to identify the domains in which children in immigrant families with various characteristics may be specially advantaged or experience special risks;
It measures pertinent family, community, and policy variables and processes to provide insights into the dynamic relationships among possible causal processes; and
It permits explicit comparisons, not only among the various immigrant groups, but also to U.S.-born black, Hispanic, Asian, and white children with U.S.-born parents with regard to policies and practices that either have worked or failed for different groups of children.
The vast majority of children in immigrant families who will require health, education, and other services during the next decade or more are the children now living in, or who will be born to, immigrant families that already reside in the United States. To ensure that scientific knowledge about these children will be available to policy makers and the American public at the earliest possible date, the new survey must include in its sample children currently in immigrant families when the survey is fielded, as well as children subsequently born into these families. Ideally, the sample would represent:
all children in immigrant families that currently live in the United States and children who are subsequently born into these families;
children in immigrant families in (at a minimum) the 6 states with the largest number of such children, or alternatively, those living in the 8 or 10 metropolitan areas with the largest number of children in immigrant families supplemented with a nonmetropolitan or rural sample;
specific countries in each of the major geographic regions that send substantial numbers of immigrants, including the Caribbean; Central and South America; East, Southeast, Central, and West Asia; East and West Europe; and Africa; and
children in populations that are difficult to enumerate, such as children of undocumented immigrants and of migrant farmworkers.
Because limited resources may preclude selection of a sample that fully meets all of these criteria, the committee urges the National Institute of Child Health and Human Development to convene an advisory group to develop design guidelines that balance these sample characteristics, especially national, state, and metropolitan sampling frames.
In this context, we reviewed the proposal for a major new survey, the New Immigrant Study (NIS), which has fielded a pilot data collection with federal support and is currently seeking funds for full implementation. The proposed study would use federal administrative records to draw a sample of foreign-born persons who were given visas in a particular year for the purpose of establishing lawful residence, either permanently (i.e., legal permanent residents, refugees) or temporarily (i.e., students, temporary workers). A second sample would be drawn of a second cohort of "new immigrants" entering three or four years after the first. Both cohorts would be resurveyed periodically.
The proposed New Immigrant Study could provide valuable information on immigration and adaptation processes among cohorts of future immigrants and their families. Our assessment of the current NIS plans, using as criteria the four essential elements and four sample characteristics described above, indicates the following.
The NIS would, or could easily, meet two of the criteria. It would be a longitudinal study that would follow first-generation children and that could follow second-generation children through time. It could, depending on sample size, represent many important countries of origin. It is uncertain, however, to what extent the NIS would meet two additional criteria, by measuring (1) a wide range of physical, psychosocial, educational, and other developmental outcomes for children, as well as (2) family, community, and policy variables and processes pertinent to child development, adaptation, and well-being.
The proposed NIS would not meet four of the eight criteria. It would not represent children in immigrant families that currently live in the United States, or in the states or metropolitan areas with the largest number of immigrants, because sample selection would include only persons who were given a visa during specified future years. It would not permit explicit comparisons between various immigrant groups and third- and later-generation white, black, Hispanic, and Asian children, because the sample would not include persons currently in the United States. Although other selected national data sources might provide a basis for simple comparisons to the NIS of child outcomes, the limits on available data would preclude explicit comparative
analyses of many child outcomes for specific racial and ethnic minorities and for important aspects of the family, community, and public policy. Finally, the NIS would not include undocumented children or the children of undocumented parents, unless they were given a visa during the sample year or lived with a person who was given a visa during a sample year.
In view of these limitations, and the fact that one of every five children in the United States today is an immigrant or the child of an immigrant, we recommend a new longitudinal survey with a sample selected from the population of children who now live in the United States, as well as children subsequently born into their families. Information should be collected longitudinally on the various children and adults within specific families. Data from this study should be made publicly available at the earliest possible time to promote their widespread use and, accordingly, the rapid accumulation of new knowledge about children currently in immigrant families.
Recommendation 2. A series of ethnographic studies on the physical and mental health of children and youth in diverse immigrant families should, insofar as possible, be embedded in the proposed longitudinal survey of children in immigrant families or in other national surveys.
The proposed longitudinal survey of children in immigrant families can provide statistically reliable estimates of major outcomes and processes for this population as a whole and for important social and cultural subgroups. But survey methodology is limited in its ability to study the meaning and interpretation that individuals give their situations. Ethnographic studies have small samples that may not statistically represent the larger population, but they are well-suited to providing rich interpretations of the processes that can only be highlighted, not probed in depth, with survey methodology.
The origins and effects of health-promoting behaviors, manifestations and effects of biculturalism, and the processes associated with the migration experience itself are examples of issues that lend themselves to this methodology, as does the study of notions about health and illness that immigrants bring and that
shape their help seeking and service use in the United States. Priority should also be given to ethnographic research, as well as to research that uses other methodologies, that addresses how aspects of growing up in the sending country prior to migration interact with the experiences of children after arrival in the United States.
Recent efforts to combine survey methodology with ethnographic methodology have been quite successful, by using ethnographic techniques to study a subsample of the larger survey sample. A single study combining the two approaches will yield much more information, both scientifically and for developing public policy, than would distinct studies each of which use only one approach. Because social and cultural processes may vary enormously across immigrant communities, it is critical that ethnographic studies of multiple diverse communities be conducted to fully understand and interpret results from broader surveys. Funding sources for this ethnographic research could include the federal government as well as private foundations.
Recommendation 3. Both quantitative and qualitative research should be conducted on the effects of welfare and health care reform for children and youth in immigrant families, and on how access to and effectiveness of health care and other services are affected by the provision of culturally competent care.
The welfare and health policies that constitute the major sources of public assistance for legal foreign-born (first-generation) children and citizen children with immigrant parents (second-generation) are undergoing dramatic change and reassessment. Welfare reform, in particular, represents the most significant shift in the modern era in the treatment of immigrant children and youth. Yet these policy changes have been forged without the benefit of an informed discussion of their likely effects on these children's development, adaptation, and future prospects. The combination of current policy change and the absence of a major investment in research to ascertain the effects of reforms in welfare and health care specifically on the lives and future prospects of children in immigrant families is troubling.
The consequences of recent and continuing changes in wel-
fare and health policies may be more immediate and profound for immigrants than for others because eligibility for noncitizen immigrants has been either cut more deeply or made a state option. Yet none of the major evaluations of welfare and health reform is focused particularly on consequences for children in immigrant families, nor have they been identified as a major subgroup for study. The effects of these changes in the provision of services and benefits to children should be studied in detail, particularly to ascertain the nature and extent of any differential effects for first-, second-, and third- and later-generation children.
Important efforts already under way to study the effects of welfare reform on children and families include the Bureau of the Census's Survey of Program Dynamics, funded under the welfare reform legislation; the 12-state study sponsored by the U.S. Department of Health and Human Services; the 50-State Database Set and the National Survey of American Families of the Assessing New Federalism Project of the Urban Institute; and the study called Welfare Reform and Children: A Three-City Study. At the present time, none of these efforts includes a specific focus on children in immigrant families, although a new federally supported effort to examine the effects of welfare reform at selected sites on the economic and health status of immigrants generally is being conducted by the Urban Institute.1
Efforts to assess the consequences of health and welfare reforms should include substantial subsamples of children and youth in immigrant families. They should also pay attention to factors uniquely relevant to them, such as their circumstances of immigration, the duration of child and parental residence in the United States (which affects eligibility for public benefits), the immigration status of siblings and parents, the likelihood that
The new Urban Institute project will explore the impacts of welfare reform on immigrants and the organizations that serve them, with a particular but not exclusive focus on effects associated with the loss of food stamps. This project will focus on two cities (Los Angeles and New York) and on immigrants in the United States at the time welfare reform was enacted (versus those arriving afterward). The project includes a somewhat restricted assessment of child health given its focus on a range of outcomes for the broader immigrant populations at the study sites.
parents seek citizenship, and their access to resources through other family members and sponsors. These efforts also need to measure both direct effects on immigrant children of their elimination from eligibility for basic benefit programs and indirect effects that ensue from family members' loss of benefits and from the possibility that some parents—particularly those who are undocumented—may be reluctant to sign up for benefits for which their children are eligible.
Children in immigrant families who remain eligible for welfare and health benefits may experience detrimental effects associated with their parents' and grandparents' loss of benefits. Such families may need to compensate for lost resources, and some strategies (e.g., shifts in goods purchased and trade-offs made in either the timing or intensity of services, including health services, that are sought) may have adverse consequences for children. Evidence on this issue does not exist.
Since the early 1970s, undocumented children have been ineligible for federal health and social benefits, with important exceptions that include emergency medical care, supplemental nutrition benefits (WIC), basic public health services such as immunizations, and public education. The recent welfare reform law continues these restrictions. Little is known about the medical and health needs of undocumented children or children with undocumented parents, or their use of services, compared with legal immigrant and U.S.-born citizen children. Little more is known about citizen children in families with undocumented immigrant parents. The committee's efforts to identify research on undocumented children revealed a glaring and significant gap in the scientific literature and greatly constrained our capacity to draw any conclusions regarding these children.
Despite the inherent difficulties that face efforts to obtain information on undocumented children and citizen children with undocumented parents, it is important that the continuing reexamination of public policy be informed by a dedicated effort to gather through available public records and other sources all available knowledge about their health and nutritional status, and their access to and utilization of pertinent benefits and programs, including those for which they are not legally eligible.
The need for health care to be provided in a culturally compe-
tent manner, including immigrant involvement in programs for their own care, has being widely recognized by federal and international health agencies and professional associations of physicians and social workers. Expert professional judgment and plausible examples suggest that culturally competent care by service providers may be key to achieving positive outcomes, and that the active participation of immigrants in programs for their own care may be important. Potentially important factors include what health outcomes immigrant parents and their children view as important. Little systematic research addresses these issues.
Because immigrants from various countries are quite heterogeneous socioeconomically and culturally, these studies should encompass a wide range of countries of origin, community settings in the United States, and types of service providers. Such studies might be effectively integrated with individual ethnographic studies or with ethno-surveys focused on individual, family, and community processes that influence child outcomes.
The committee urges the Health Resources and Services Administration of the U.S. Department of Health and Human Services to establish a clearinghouse to assemble and serve as a repository for knowledge about the nature and efficacy of various programs and approaches designed to foster the development of cultural competence among service providers and to directly involve immigrants in the delivery of health care and mental health services. Efforts supported by federal and state governments, professional organizations, and health care institutions should be systematically assessed to provide the basis for implementing and evaluating community intervention programs.
Data Collection and Dissemination
In addition to these recommended new research studies, the committee recommends measures to improve existing data resources and highlight immigrant children in existing data reports, each of which would substantially improve the available knowledge about these children.
Recommendation 4. The federal government should collect and code information on country of birth, citizenship status, and
parents' country of birth in key national data collection systems. This information should be made available through public use microdata samples and other vehicles for public distribution of data.
The federal government funds public health information and surveillance systems and national surveys to monitor physical and mental health, and circumstances and risk factors related to family, education, employment, income, housing, crime, and participation in public programs (see Box 5-1). It has been recognized for decades that accurate interpretation of these data requires analysis using the basic demographic measures of age, sex, race, and ethnicity. Because children in immigrant families constitute a large and increasing proportion of the population; because their healthy development is in the national interest; and because country of origin, citizenship status, and parents' country of birth may have important implications for healthy development, the committee urges that these three data elements be added to the list of basic demographic variables obtained in the data collection systems discussed below.
The Centers for Disease Control and Prevention (CDC), which has primary responsibility for health surveillance systems and public health programs, recently proposed standards for ''citizenship" data to be used in health information and surveillance systems (Centers for Disease Control and Prevention, 1997). The committee commends and urges the adoption of this proposal.2
Recent experience suggests that the inclusion of questions about citizenship will not produce a disincentive to participation. Specifically, the Adolescent Health Survey asked foreign-born adolescents about their citizenship status, and of the 1,900 who were asked the question, only one refused. There were no reports from the survey field staff indicating sensitivity to the question. Since 1994 in the Current Population Survey, the Bureau of the Census has asked country of birth, citizenship, year of entry, and parents' country of birth for every household member. It has encountered no negative reaction in the field to these questions, and the nonresponse rate is quite low. For example, the nonresponse rates were about 1 percent each for respondent's country of birth, mother's country of birth, and father's country of birth, and 4.6 percent among the foreign-born for the question "Are you a citizen?"
The committee also urges that the standard be expanded to include data elements on country of birth and parents' country of birth. Country of birth data and parents' country of birth data are essential to identify the origins of immigrant citizens as well as noncitizens. The CDC surveillance systems and surveys that should collect these data include those focused on behavioral risk factors, birth defects, immunizations, cancer registration, HIV/ AIDS, sexually transmitted diseases, tuberculosis, youth risk behavior, and chronic diseases.
The National Center for Health Statistics within the CDC has immediate responsibility for compiling data from the birth and death registration systems, and for conducting the National Health Interview Survey, the National Health and Nutrition Examination Survey, the National Survey of Family Growth, and the Medical Expenditure Panel Survey. Taken together, these data collections constitute the core of the nation's system for periodic monitoring of health status and expenditures, fertility behavior, and mortality. Also directly relevant to health is information on the experience of adolescents as victims of crime, including exposure to violence and injury, that is collected annually in the National Victimization Survey under the auspices of the Bureau of Justice Statistics.
Two additional surveys provide critical information on adolescent risk behavior. The Monitoring the Future survey, funded by the National Institute on Drug Abuse, provides annual estimates for high school seniors on the prevalence and incidence of illicit drug use and on important values, behaviors, and lifestyle orientations. The National Longitudinal Study of Adolescent Health, funded by the National Institute of Child Health and Human Development (NICHD), provides an extremely rich source of longitudinal information for a national sample of youth who were in grades 7 through 12 in 1995 regarding physical and mental health, health risk behaviors, and the family, school, peer, and neighborhood environments.
The educational progress of children and the role of various influences are monitored by the National Center for Education Statistics with the National Educational Longitudinal Surveys, the National Household Education Survey, and the Early Childhood Longitudinal Survey.
Family and economic risk factors for children are measured with the greatest geographic specificity for the largest number of population subgroups by the Bureau of the Census in its Decennial Census of the Population and Housing. The Current Population Survey, sponsored by the Bureau of the Census and the Bureau of Labor Statistics, provides an annual update of decennial census information, with much greater detail on labor force participation, income, and school enrollment, but with less geographic specificity and for fewer subgroups.
Participation in welfare programs as related to family economic circumstances is also assessed in greatest detail by the Bureau of the Census through its Survey of Income and Program Participation and its Survey of Program Dynamics. The Bureau of Labor Statistics is the primary sponsor of the National Longitudinal Surveys of Youth, which focus on the transition from school to work and include supplements with extremely rich data on child development funded by the National Institute on Child Health and Human Development.
Among these data collection systems and surveys, only the Current Population Survey collects the full set of immigration data recommended here beginning in 1994, with funding from the Immigration and Naturalization Service and the National Institute of Child Health and Human Development. As the immigrant population grows and these data elements become essential to a proper understanding of trends in the social, economic, and health status health of the U.S. population, the cost of adding these elements to any specific survey is judged by the committee to be quite modest. But the returns on this expenditure would be quite substantial improvements in our ability to understand the health and social needs of immigrant children.
Recommendation 5. As the federal government develops new surveys or draws new samples to supplement or extend existing surveys, it should select and include subsamples that are large enough to reliably monitor the circumstances of children and youth in immigrant families as a whole and, where feasible, for specific countries of origin.
New samples are drawn periodically for continuing surveys and will be drawn for new national surveys in the future. Despite growth of the immigrant population, samples in most national surveys are too small to sustain statistically reliable estimates for the foreign-born population as a whole. This difficulty can be resolved by oversampling—drawing samples in which the foreign-born and their families represent a larger proportion than they are in the population as a whole.
Immigrants from different countries vary enormously in socioeconomic resources, language, and culture, but samples in most surveys are too small to provide estimates by specific country of origin. This problem can be mitigated in particular surveys by oversampling immigrants from a few specific countries of origin, where criteria relevant to the primary content of the survey provide the basis for selecting specific countries for study.
For any given survey, the trade-offs involved in adding a nationally representative refresher sample of immigrants or subsamples of immigrants from specific countries of origin should be explicitly debated. At the present time, it is rare for either strategy to be seriously considered. One exception is the Panel Study of Income Dynamics (PSID) conducted at the University of Michigan,3 which recently supplemented its core sample with a representative sample of immigrants.
Recommendation 6. Key indicators of child well-being published in the annual report of the Federal Interagency Forum on Child and Family Statistics should, insofar as possible, distinguish among foreign-born immigrant children (first generation), U.S.-born children in immigrant families (second generation), and U.S.-born children in U.S.-born families (third and later generation).
A recent presidential executive order mandates the Federal Interagency Forum on Child and Family Statistics to publish an
annual report on children (U.S. Department of Health and Human Services, 1996, 1997), but as yet there is virtually no public dissemination of information on even the most basic indicators on the lives and well-being of children in immigrant families—the fastest-growing segment of the child population. Precedents for federal publication of annual reports on important minority groups include the Current Population Reports of the Census Bureau on the black and Hispanic populations. We recommend that key indicators of child well-being published in this annual report should distinguish, insofar as possible, between foreign-born children in immigrant families (first-generation), U.S.-born children in immigrant families (second-generation), and U.S.-born children in U.S.-born families (third- and later-generation).