NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Research Council (US) and Institute of Medicine (US) Committee on the Health and Adjustment of Immigrant Children and Families; Hernandez DJ, editor. Children of Immigrants: Health, Adjustment, and Public Assistance. Washington (DC): National Academies Press (US); 1999.
Children of Immigrants: Health, Adjustment, and Public Assistance.
Show detailsPeter David Brandon
National debate over immigration policy is not new in the United States. The immigration debate in the United States dates back to the colonies, which argued over who was ultimately responsible for destitute newcomers. In recent times, though, the debate has focused mainly on immigrants' adjustment to American society and their alleged displacement of native-born workers in the job market.
With passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which greatly affects safety net provisions for immigrants, it is clear that the relationship between immigration and social policy is again a critical part of the national debate over immigration policy. The 1996 legislation reflects a public perception1 that immigrants should pay their own way and that it is wrong for immigrants to depend on welfare.2
There is evidence that immigrant households have higher rates of welfare receipt and that the rates have increased over time (Borjas and Hilton, 1996). Yet because the U.S. welfare system contains such a patchwork of programs addressing so many different needs and populations, much more research is needed to explore immigrants' utilization of public assistance programs. And although immigrant children comprise a large part of the immigrant population as well as representing a sizable fraction of the welfare caseload, no previous analyses apart from Hofferth (this volume) and Currie (1997) have focused on needy immigrant children. Using data from the Survey of Income and Program Participation, this chapter begins to fill this gap in our knowledge by providing a more complete picture of immigrant children and their families' participation in means-tested entitlement programs. In essence, the real picture differs from the one framed in the public's eye.
Context And Background Studies
Until the new welfare legislation became law in August 1996, legal immigrants could receive benefits from federal public assistance programs. If immigrants were needy or unemployed, they faced the same eligibility rules as citizens. Basically, administrative rules for public assistance drew no real distinctions between legal permanent residents and citizens. Even undocumented immigrants were eligible for some emergency relief, such as medical calve.3
In the context of the old welfare system, several important national-level studies compared public assistance receipt between immigrant and native groups. Blau (1984, 1986), Simon (1981), and Simon and Akbari (1996) used the 1976 Survey of Income and Education (SIE) to conclude that immigrant families received substantially less in annual transfer payments than native families. Tienda and Jensen (1986) updated Blau's 1984 analysis by using 1980 census data rather than the 1976 SIE. They demonstrated that rates of welfare use in 1979 were higher for natives than for immigrants when other factors such as age, education, and marital status were taken into account.
Five years later Borjas and Trejo (1991) reexamined issues that Blau (1984) and Tienda and Jensen (1986) had raised about immigrant participation in the American welfare system. Using both the 1970 and 1980 censuses, their work distinguished itself from previous works by investigating intracohort changes in welfare participation over time. The important findings from their study were that among households headed by males the gap in welfare receipt between natives and immigrants grew larger over the decade, but immigrant households headed by females were converging with native households headed by females.
Borjas and Hilton (1996) have suggested that studies should include all noncash benefits when attempting to understand welfare receipt by immigrants. Although these researchers found that immigrant-native differences in the receipt of cash benefits were small, those differences grew significantly after noncash benefits were included.
The studies cited above have helped us understand the propensity of immigrants to participate in the U.S. welfare system. To varying degrees, each study has implied that future researchers must recognize the differences in demographic profiles among immigrant groups; some researchers may even argue that there is no typical immigrant profile for welfare utilization. In any case, these studies indicated the need for more research (1) to identify differences in welfare use among immigrants according to the ethnic group to which they belong and (2) to learn about differences in welfare use according to the degree of assimilation immigrants have undergone. As the present study shows, successive generations of immigrants have different propensities to enter the U.S. welfare system. Indeed, Kao and Tienda (1995) have suggested that as new cohorts of immigrant children assimilate they do worse in the U.S. mainstream with respect to school performance, not better.
Moreover, past studies—like those cited above—have focused on welfare receipt among immigrant adults or households. Equally imperative, however, is learning about the lives and economic conditions of immigrant children. Despite large numbers of immigrant children comprising the next generation (Edmonston, 1996), only a scant amount of research documents their economic well-being, performance in school, health, and, as this study highlights, their exposure to the U.S. welfare system. But as Hofferth (this volume) states, a better understanding of how well immigrant children are faring in U.S. society is vital, particularly as schools become more racially and ethnically diverse, as the number of non-English-speaking children grows, and as the costs of supporting needy minority children rise.
Data Description And Empirical Approach
This study used data from the 1986, 1987, 1988, 1990, and 1991 panels of the Survey of Income and Program Participation (SIPP). The SIPP is a longitudinal survey of a random sample of the U.S. population. The five SIPP panels spanned the period from October 1985 through March 1992. Each wave of the survey was conducted every four months, so each participant was interviewed three times a year about his or her economic experiences over the past four months, including benefits received from many different means-tested entitlement programs. As the households were reinterviewed at four-month intervals for six, seven, or eight waves, depending on the particular panel, the survey provided 24, 28, or 32 consecutive months of data for each household.4 Besides providing monthly details about the use of cash and noncash transfer programs, it also collected monthly data on household composition, employment, and sources of private income.5
The SIPP was well suited to the present study because it also reports the race and ethnicity of each respondent; where each respondent was born; and, if born abroad, the year of arrival in the United States.6 With this information, persons were classified according to whether they were natives or immigrants and according to their race and ethnicity. Immigrants in the sample were those who were born abroad and who were not naturalized citizens of the United States or persons who were born abroad to citizens of the United States.
Possessing year of arrival as well as having birthdates of children permitted determination of whether children came to the United States with their parents or were born after their parents arrived here. If children were born before their parents immigrated to the United States, they were classified as foreign-born children; if they were born after at least one of their parents arrived here, they were classified as native born. Children born to native-born parents were classified as native born. In one set of analyses where the focus was the impact of immigrant generation on whether a child resided in a family that received public assistance, the latter group was the comparison group. To summarize, children were classified as (1) foreign-born children, (2) native-born children with at least one foreign-born parent, and (3) native-born children with native-born parents.
The SIPP collected data on all persons over age 15 in each household as well as data on all other persons, including children, who lived with or moved into any given household. A person in each household was called the "household reference person." This person is the one who either owns or rents the house. Analyses were restricted to the sample of persons who were household reference persons, who were at least 15 years old, and who reported that they were the parents (or guardians) of children under age 18.7
Once it had been established which children under the age of 18 lived in households that were and were not receiving some form of public assistance, a research file was generated in which the child became the unit of analysis while maintaining pertinent household information and data about program participation. Since the sample of children remained in the SIPP for about two years on average (see Table 11A-1), this feature of the survey was exploited to create a person-year file. In other words, a file was created wherein one record represented a year for each year that a child was in the SIPP and remained under age 18. The empirical models therefore calculated the probability that a child under age 18 was in a family that received some form of public assistance in a single year. Thus, the results presented in this study were based on the child person-year as the unit of analysis.8
Table 11A-1
Characteristics of Children's Years in the Survey of Income and Program Participation, 1986-1992.
Key demographic information on household heads was appended to each child in the sample as well as indicators of immigrant generation, race, and ethnicity. These factors were affixed because age of household heads, citizenship, marital status, education levels, health, family size, and ages of youngest dependents were expected to influence whether children were in a family that received some sort of public assistance in a single year of childhood. Finally, to control for trends in recipiency over time, several binary variables were created, representing the year in which each child was in the survey.
Table 11-1 lists the welfare programs studied and the benefits that each program provides. The programs were Aid to Families with Dependent Children (AFDC); Supplemental Security Income (SSI); other welfare; Medicaid; food stamps; heating assistance; housing assistance (rent subsidies or public housing); Women, Infants, and Children (WIC); general assistance (GA); and school lunches. Table 11-2 provides definitions of the variables representing the characteristics of families, and Table 11-3 presents the incidence of types of public assistance use for the child person-year sample. Other tables contained in Appendix 11A display summary statistics of the sample (Table 11A-1) and results of regressing participation in each public assistance program on ethnicity, recency of arrival, and poverty status, controlling for the demographic variables and the year in which program participation occurred (Tables 11A-2 through 11A-5).
Table 11-1
Description and Costs of Means-Tested Programs, 1991.
Table 11-2
Definitions of Variables.
Table 11-3
Annual Probability of Children Receiving Public Assistance in Children's Years in the Survey of Income and Program Participation.
Table 11A-5
Odds Ratios from Regression of Public Assistance on Ethnicity and Generational Groups.
Similar to Hofferth's (this volume) methodology, AFDC, SSI, GA, and other welfare were combined to produce an indicator of cash assistance, and Medicaid, food stamps, heating assistance, housing assistance, and WIC were combined to produce an indicator of noncash assistance. Then a final indicator for receipt of any assistance, cash or noncash, was created.
Findings
The data contain observations on 1,372 foreign-born children with foreign-born parents; 4,910 native-born children with at least one foreign-born parent; and 36,643 native-born children with native-born parents. There are 3,292 children of Mexican origin; 579 of Puerto Rican background; 85 of Cuban background; 1 of other Hispanic background; 835 children of Asian background; and 475 of Eastern and 6,791 of Western European backgrounds. There are 24,279 white children and 5,219 black children.
When ethnicity and generation were combined, the following numbers of foreign-born children with foreign-born parents were obtained: 399 Mexican children, 68 Puerto Rican children, 7 Cuban children,9 193 children of other Latino background, 64 Western European children, 26 Eastern European children, and 221 Asian children. Similarly, by combining ethnicity and generation, the following numbers of native-born children with at least one foreign-born parent were obtained: 1,509 Mexican children; 39 Puerto Rican children; 75 Cuban children; 402 children of other Latino background; 428 Western European children; 73 Eastern European children; and 431 Asian children. The numbers of native-born children with native-born parents were 1,384 Mexican children; 472 Puerto Rican children; 3 Cuban children; 589 children of other Latino background; 6,299 Western European children; 376 Eastern European children; and 183 Asian children.
Once separate person-year records were created and years in which each child was 18 or older were deleted, 125,822 person-years were observed)10 Breaking down the sample by ethnicity and generation in child person-years yielded 1,145 foreign-born Mexican children and 4,455 native-born Mexican children with at least one parent born in Mexico; 200 foreign-born Puerto Rican children and 116 native-born Puerto Rican children with at least one foreign-born parent; and 1,388 native-born Puerto Rican children with native-born Puerto Rican parents. For Cuban children there were 213 native-born children with at least one parent born in Cuba (see note 9). Asian children in person-years included 595 who were foreign born; 1,261 who were native born with at least one parent born in an Asian country; and 533 native-born children. For Eastern and Western European children combined, there were 265 foreign born; 1,472 native born with a least one parent foreign born; and 19,557 native born. Finally, there were 539 other Latino children person-years; 71,284 white person-years; and 15,255 black person-years.
The sample distributions on child and family characteristics for individual child-years by ethnicity and generation are shown in Table 11A-1. On average, children in the sample were about 9 years old, and the head of the family was age 36, married, working, at least a high school graduate, and caring for 2.4 children, with the youngest about age 6.5.
Foreign-born Mexican American families had the most children, 3.5 on average, with about 3 for native-born children with a foreign-born parent. Asians and Puerto Ricans have the next-largest family sizes, about 3 children. Cubans have the smallest family size, about 2 children each.
Mexicans also possessed the lowest levels of education. Among foreign-born Mexican children with foreign-born parents, 76 percent were in a family whose head had less than a high school education. Among native-born Mexican children with native-born Mexican parents, however, there were much higher levels of high school completion; 51 percent of these Mexican children were in a family whose head had at least a high school education. Puerto Rican children were in families in which about half the heads had less than a high school education. Foreign-born Eastern and Western European children were in families in which very few heads had less than a high school education. Children born in Puerto Rico were in families in which 25 percent of heads had health problems that stopped them from working. That was nearly four times higher than the number of foreign-born Asian children and over double that of foreign-born Mexican children.
Puerto Rican and black children were more likely than any other group of children to live with only one parent. Cuban and Asian children were most likely to live with two married parents. Though less apparent, among Mexican children, those who were native born were more likely to live with one parent than their foreign-born counterparts. Foreign-born Asian children with foreign-born parents and native-born Asian children with at least one foreign-born parent lived with heads who were more likely to be citizens than other groups of children, while Western European and Mexican children experienced the opposite.
Incidence of Public Assistance Receipt by Ethnicity and Generation
Table 11-3 shows the incidence of participation in public assistance programs by combining ethnicity and generation for per-son-years in which each child was under 18. In the table, foreign-born Mexican children are more likely than white children in native-born families to receive AFDC, food stamps, Medicaid, other welfare, school lunches, and housing and heating assistance. They are at least two times more likely than white children to live in families that receive either AFDC (17 percent compared to 7 percent) or Medicaid (24 percent in contrast to 9 percent). Use of SSI and the WIC program is equal to that of white children in native-born families. Except for other welfare, native-born Mexican children with a foreign-born parent are more likely than white children in native-born families to use all forms of public assistance. Mexican children in native-born families are also more likely than white children in native-born families to use all forms of public assistance.
Cuban children in native-born families are actually less likely than white children in native-born families to receive AFDC, other welfare, housing assistance, heating assistance, and WIC benefits. In contrast, all other Latino children, except Puerto Rican children (discussed below), are more likely to use all forms of public assistance (see Table 11-3).
According to Table 11-3, foreign-born Asian children are more likely than white children in native-born families to receive AFDC, SSI, food stamps, Medicaid, WIC, school lunches, and housing and heating assistance. They are nearly four times more likely than white children to live in families that receive either AFDC (27 percent compared to 7 percent) or Medicaid (31 percent in contrast to 9 percent). Use of other welfare is less than that of white children in native-born families. Except for WIC, other welfare, and heating assistance, native-born Asian children with a foreign-born parent are more likely than white children in native-born families to use the major public transfer programs. Asian children in native-born families also are more likely than white children in native-born families to use AFDC, SSI, other welfare, food stamps, housing, and school lunches but are less likely to use Medicaid, heating assistance, and WIC.
Among Eastern Europeans, Table 11-3 indicates that foreign-born Eastern European children are less likely than white children in native-born families to use AFDC, SSI, other welfare, and Medicaid but are more likely to receive WIC, school lunches, and housing and heating assistance. Except for WIC, Eastern European children with native-born parents are less likely than white children in native-born families to use the major public transfer programs. Eastern European children with at least one foreign-born parent are less likely than white children in native-born families to receive AFDC, SSI, other welfare, food stamps, housing assistance, heating assistance, WIC, and school lunches but are more likely to receive Medicaid benefits.
Foreign-born Western European children are less likely than white children in native-born families to receive AFDC, SSI, other welfare, assistance for housing and heating, and school lunches but are more likely to use food stamps, Medicaid, and WIC. Western European children with native-born parents and Western European children with at least one parent who is foreign born have essentially the same chances of using various forms of public assistance as white children in native-born families.
Among all groups of children, Table 11-3 shows that black children and Puerto Rican children have the highest rates of public transfer receipt, with 35 and 42 percent, respectively, receiving AFDC. Whereas 9 percent of white children received Medicaid, 40 percent of black children and 52 percent of Puerto Rican children lived in families that received Medicaid. With respect to food stamps, 51 percent of black children and 52 percent of Puerto Rican children lived in families that received them, while only 14 percent of white children lived in families that received food stamps.
Multivariate Analyses of the Determinants of Public Assistance Receipt
Comparing the prevalence of welfare receipt among foreign-born children with that of native-born children is informative but, without adjusting for the extent of disadvantage across groups, knowing whether the higher rates of receipt are due primarily to socioeconomic differences, such as education levels, or to recency of arrival in the United States remains uncertain. To resolve this uncertainty, multivariate analyses were conducted in which each type of public assistance receipt was regressed on ethnicity, generation, the combined measure, poverty status, and other demographic factors.
Tables 11A-2 through 11A-5 show the coefficients from such regressions. The coefficients suggest that differences in receipt decline once poverty status and demographic factors are controlled and lead to the conclusion that socioeconomic characteristics of families of immigrant children, rather than recency of immigration per se, explain the higher rates of public assistance use.
The first set of multivariate results (see Table 11A-2) suggest that controlling for poverty and other socioeconomic factors, and except for SSI and school lunches, foreign-born children and native-born children with at least one foreign-born parent are less likely to be in families that receive public assistance than native-born children with native-born parents. Specifically, both groups of children are less likely to receive AFDC, food stamps, Medicaid, and housing and heating assistance but are more likely than native-born children in native-born families to receive SSI. Foreign-born children are also more likely than native-born children with native-born parents to use the federal school lunch program.
After controlling for recency of arrival, poverty status, and other factors, it is evident that Mexican children are less likely than white children with native-born parents to use SSI and other welfare, are no more likely to use AFDC or WIC, but are more likely to use food stamps, Medicaid, and housing and heat assistance. Compared to white children, only for food stamps and Medicaid were Cuban children more likely to use these programs.
Puerto Rican children, on the other hand, were more likely than white children to use AFDC, SSI, food stamps, Medicaid, school lunches, and housing and heating assistance but less likely to receive other welfare. Aside from heating assistance, Asian children exhibited exactly the same pattern as Puerto Rican children. And, both Eastern and Western European children were either less likely or no more likely than white children to use AFDC, SSI, food stamps, and the WIC program. Possibly because of refugee status, however, Eastern European children were more likely than white children to use Medicaid, housing assistance (as were Western Europeans also), and heating assistance. After controlling for poverty status, black children were much more likely than white children to use all public assistance programs.
Figure 11-1 displays the predicted probabilities of receiving AFDC, SSI, food stamps, and Medicaid. Except for SSI, where the predicted probabilities diverge less across the different generations, foreign-born children with foreign-born parents and native-born children with at least one foreign-born parent are much less likely than native-born children with native-born families to receive AFDC, food stamps, and Medicaid. Moreover, the predicted probabilities shown in Table 11A-2 and Figure 11-1 suggest that foreign-born children with foreign-born parents are less likely to receive food stamps. What becomes greatly apparent from subsequent figures and tables (Figures 11-2 through 11-5) once ethnicity and generation are combined is that predicted probabilities for receipt of public assistance diminish and that differences in predicted probabilities of receipt among foreign-born children and native-born children disappear.
Figure 11-1
Predicted probability of public assistance receipt by nativity.
Figure 11-2
Predicted probability of SSI receipt by ethnicity and nativity.
Figure 11-5
Predicted probability of AFDC receipt by ethnicity and nativity.
Table 11A-3 contains the multivariate results that combine ethnicity and generation. These results bolster the thesis that economic deprivation and other sociological factors rather than immigration status better explain the use of public assistance across ethnic groups. Foreign-born Mexican children, for example, are less likely than white children in native-born families to receive AFDC, SSI, and food stamps and are no more likely than white children in native-born families to receive Medicaid. Native-born Mexican children with a foreign-born parent are also less likely to receive AFDC and Medicaid and are no different in their use of food stamps than white children, although these children are possibly more likely than white children to receive SSI. Mexican children in native-born families are more likely to receive AFDC, food stamps, and Medicaid when controlling for poverty but less likely than white children in native-born families to receive SSI.
Figure 11-3
Predicted probability of Medicaid receipt by ethnicity and nativity.
For the other transfer programs, foreign-born Mexican children are less likely than white children in native-born families to receive heating assistance and WIC benefits but are more likely to receive school lunches and to live in families that receive housing assistance. Foreign-born Mexican children are no different than white children in their use of other welfare programs, again controlling for poverty status. Both native-born Mexican children with a foreign-born parent and Mexican children in native-born families are more likely than white children to receive school lunches and housing assistance. However, while the latter group of children is more likely than white children to receive heating assistance, the former group is no different than white children with respect to heating assistance. Mexican children in native-born families with at least one foreign-born parent are more likely than white children to use the WIC program when controlling for poverty, but again compared to white children in native-born families, Mexican children in native-born families are actually less likely to receive WIC benefits.
Figure 11-4
Predicted probability of food stamp receipt by ethnicity and generation.
Comparing Mexican children from various generations, again with the same controls in the multivariate analyses, foreign-born Mexican children are much less likely than Mexican children in native-born families to be living in a family that relies on AFDC and food stamps. The same pattern exists for heating assistance. Only for WIC, school lunches, and housing assistance are foreign-born Mexican children more likely than Mexican children in native-born families to receive government assistance.
Overall, foreign-born Mexican children are less likely than white children in native-born families to live in families that receive AFDC, SSI, food stamps, heating assistance, and WIC while being no more likely to receive Medicaid or other types of welfare. But they are more likely to receive school lunches and housing assistance. Among Mexican children in native-born families, however, the chances of receiving AFDC, food stamps, Medicaid, and heating assistance are higher than among foreign-born Mexican children and white children in native-born families. Only with respect to other welfare are Mexican children in native-born families less likely to receive heating assistance than are white children in native-born families.
Although the proportions in Table 11-3 of foreign-born Mexican children receiving the major public transfer programs are higher for nearly every program than white children (WIC and SSI are the exceptions), the multivariate analyses suggest that, after controlling for poverty, region, and family factors, foreign-born Mexican children are actually less likely than both white and Mexican children in native-born families to receive public assistance.
Cuban children are less likely than white children in native-born families to receive AFDC and are more likely to receive food stamps and Medicaid, but they are no more likely than white children in native-born families to receive SSI, housing assistance, or heating assistance. Regarding the WIC program and other welfare, no data were available for Cubans, but data were available for the school lunch program. As with other major transfer programs, Cuban children are less likely to use the school lunch program than are white children in native-born families.
Insufficient data prohibited cross-generational comparisons among Cubans as it did among the group of Hispanics categorized as ''other Latino." For the other group of Latino children, they are more likely than white children in native-born families to use every public assistance program, except for heating assistance and other welfare.
Poverty status, geographic region, and socioeconomic characteristics of families were, again, controlled for in the multivariate analyses for Asians. The results suggest that foreign-born Asian children are more likely than white children in native-born families to receive AFDC, SSI, food stamps, Medicaid, housing assistance, and school lunches. Receipt of heating assistance by foreign-born Asian children is no more likely than by white children in native-born families. Native-born Asian children with at least one foreign-born parent share the same pattern as foreign-born Asian children, albeit for a smaller set of transfer programs.
Compared to white children with native-born parents, native-born Asian children with at least one foreign-born Asian parent are more likely to receive AFDC, SSI, other welfare, Medicaid, housing assistance, and heating assistance. For three programs, though—food stamps, WIC, and school lunches—native-born Asian children with a foreign-born parent are no more likely than white children to be recipients of the three programs. Again, Asian children in native-born families are more likely than white children to receive AFDC, housing assistance, heating assistance, and school lunches, but after controlling for poverty their receipt of SSI, food stamps, other welfare, and Medicaid is no different than that of white children in native-born families.
Comparing Asian children from various generations, foreign-born Asian children are much more likely than Asian children in native-born families to be living in a family that relies on SSI, Medicaid, food stamps, and other welfare programs. On the other hand, Asian children in native-born families, like foreign-born Asian children, have higher rates of receipt of AFDC, heating assistance, housing assistance, WIC, and school lunches than white children in native-born families.
Results suggest that all Asian children, regardless of country of birth, are more likely than white children in native-born families to receive AFDC, housing assistance, and school lunches. However, only foreign-born Asian children and Asian children with at least one foreign-born parent are more likely than white children to receive Medicaid. The relatively lower use of heating assistance among native-born Asian children compared to other Asian children and white children with native-born families perhaps reflects interstate migration of subsequent generations. More meaningful comparisons would eventuate if the refugee status of Asian children's families was known and country of origin for the foreign born established.
In contrast to foreign-born Mexican children, the multivariate analyses suggest that after controlling for poverty, geographic region, and family factors, foreign-born Asian children are more likely than both white and Asian children in native-born families to receive SSI, food stamps, Medicaid, and other welfare assistance. The port of entry for foreign-born Asian children may have lowered the probability of them, compared with Asian children in native-born families, receiving heating assistance and made the chances of use no different than those of children in native-born families.
Results suggest that foreign-born Eastern European children are more likely than white children in native-born families to use some transfer programs. These particular children are more likely than white children to receive Medicaid, housing assistance, and other welfare, although they are no different in the likelihood of receiving AFDC, food stamps, heating assistance, and school lunches. Native-born Eastern European children with at least one foreign-born parent are less likely to receive AFDC and food stamps, as well as housing assistance and school lunches. Regarding other welfare, housing assistance, and the WIC program, native-born Eastern European children with a foreign-born parent are no more likely than white children to use these programs, but they are more likely to receive Medicaid. Eastern European children in native-born families are less likely than white children to receive SSI, WIC, and school lunches but are no more likely to receive Medicaid and housing assistance.
For some programs, comparisons are possible among Eastern European children. Foreign-born Eastern European children are no different in their receipt of Medicaid and housing assistance than Eastern European children in native-born families. And foreign-born Eastern European children appear to be more like white children in native-born families with respect to the federal school lunch program; both of these groups of children are less likely than Eastern European children in native-born families to participate in the school lunch program.
In contrast to foreign-born Eastern European children, foreign-born Western European children are less likely than white children in native-born families to receive AFDC, Medicaid, heating assistance, and school lunches. Moreover, these particular children are no different in their receipt of food stamps than white children in native-born families. The pattern is more distinguished for native-born Western European children with at least one foreign-born parent. They are less likely to receive food stamps, housing and heating assistance, WIC, or school lunches.
The pattern of either being less likely to receive or to receive at the same rate as white children in native-born families changes only when results for Western European children in native-born families are examined. Western European children are in fact more likely than white children to receive housing assistance and school lunches but are less likely than white children to receive SSI, food stamps, Medicaid, and heating assistance. Also, their receipt of AFDC is no different than that of white children.
For all Western European children, no great differences exist in the receipt of heating assistance. Foreign-born Western European children and those with native-born parents are actually less likely than native-born Western European children with at least one foreign-born parent to receive Medicaid, whose chances of receiving Medicaid are like that of white children. And while foreign-born Western European children are no different than white children with respect to food stamps, native-born Western European children are actually less likely than white children to receive food stamps.
Controlling for poverty and the other predictors of public assistance use, children born in Puerto Rico are more likely than white children in native-born families to participate in all of the major public assistance programs, except for WIC and other welfare. Similarly, Puerto Rican children born in the United States with at least one parent born in Puerto Rico are less likely than white children with native-born families to receive WIC benefits but are more likely to receive Medicaid and housing assistance. For this group of children, results suggest that the use of food stamps is similar to that of white children in native-born families. Puerto Rican children with both parents born in the United States are more likely than white children with native-born families to receive AFDC, food stamps, Medicaid, housing assistance, heating assistance, and school lunches. Only for use of the WIC program are Puerto Rican children with native-born families comparable to white children with native-born families.
Comparisons among groups of Puerto Rican children are less meaningful since Puerto Ricans are U.S. citizens and thus eligible for welfare. The results reflect this by showing that there are high rates of public assistance receipt across the three groups of Puerto Ricans. The one subset of results which indicate that U.S.-born Puerto Rican children with at least one parent born on the island are no different than white children in their use of AFDC, food stamps, and heating assistance may be due to smaller sample sizes that cannot properly distinguish differences.
Lastly, black children in native-born families are more likely than white children in native-born families to use all forms of public assistance, after controlling for poverty status, geographic region, and other factors. The greatest differences in the increased likelihood of their receiving some form of assistance compared to white children in native-born families occurs with food stamps, Medicaid, and housing assistance.
Overall, the analyses suggest that, after controlling for key factors, several foreign-born groups of children are typically less likely than white children in native-born families to use public assistance, rather than more likely as implied by Table 11-3. The results also mirror the diversity of the American welfare experience across immigrant/ethnic groupings of children. Asians display a decline, rather than an increase, in program use across generations, perhaps reflecting changes from refugee status for the first generation to comparative successes among later generations. For Eastern Europeans a similar though less clear pattern emerges as well. For them use of welfare programs declines until it is essentially no different than that of native-born whites. Mexicans, on the other hand, experience the opposite pattern. Successive generations of Mexicans—here, native-born children with native-born parents—actually become more likely than white children to use several of the major welfare programs. Their use of public assistance is still less than that of native-born Puerto Ricans and blacks but rises above that of foreign-born Mexican children. Thus, different stories arise for different generations of ethnic groups with respect to the use of public assistance.
Time Trends in Public Assistance Receipt and Other Results
The aim here was to examine the effects of immigrant status, poverty, ethnicity, race, and time of arrival in the United States on the receipt of public assistance. To keep that focus, the discussion has refrained from elaborating on other factors determining welfare receipt. The important point to note is that the other findings fully complement those findings reported by Hofferth (this volume), Tienda and Jensen (1986), and Borjas and Hilton (1996). In other words, female headship, poor health, low education, and more children raise the probability of receiving various forms of public assistance.
Further, no discernible trends were found in the use of public assistance. Nevertheless, the binary dummy variables that controlled for the five successive years predicting the probability of participating in a given program at times suggested significant differences across years. Specifically, yearly dummy variables indicate lesser use of AFDC, food stamps, and school lunches in the earlier years of SIPP (mid-to late 1980s).
To complete the analyses, separate regression analyses predicting participation for all types of public assistance programs were conducted on the 1990 and 1991 panels only. These final additional analyses assessed the sensitivity of results to smaller sample sizes and analytical sampling weights. The outcome from the analyses showed that the directions and magnitudes of the estimates remained unchanged. Occasionally, magnitudes of estimates decreased, making effects statistically insignificant, but this was the exception not the rule.
Conclusions
Findings from multivariate analyses support the economic disadvantage hypothesis rather than the immigrant status hypothesis. From the results there is evidence that foreign-born children with foreign-born parents and native-born children with foreign-born parents were less likely than white native-born children with native-born parents to receive public assistance. This result contradicts assertions that immigrant families overuse the welfare system. In fact, foreign-born immigrant children were neither more nor less likely than native-born white children with native-born parents to receive cash assistance, and native-born immigrant children with foreign-born parents were less likely to receive cash assistance. In addition, foreign-born children with foreign-born parents and native-born children with foreign-born parents were less likely than native-born white children to receive noncash forms of public assistance. After controlling for poverty and other sociodemographic factors, the association between lower use of public assistance and recency of arrival remained.
Another conclusion from the study was that, although children living in families of minority status were more likely to use public assistance programs more heavily than white children, the higher likelihood was not tied to immigration status. Indeed, receipt of public assistance significantly differed by recency of arrival for each ethnic and racial group. Among Mexican Americans, native-born Mexican American children with at least one foreign-born parent were less likely than white children to receive AFDC. In contrast, among Asian Americans, native-born Asian American children with foreign-born parents and native-born Asian American children with native-born parents were more likely than white children to live in families that received AFDC, and native-born Asian children with at least one foreign-born parent also were more likely to receive SSI, housing assistance, and Medicaid. Among Eastern European children, both native-born children with foreign-born parents and those with native-born parents were less likely than native-born white children to live in families that received AFDC, food stamps, WIC, or school lunches. Generally, there were large differences among different immigrant groups in the receipt of public assistance.
Puerto Rican children and other Latino children were more likely than white children to receive all forms of public assistance. This conclusion was not reached for Cuban children, however. Use of public assistance among Cuban children was essentially program specific. Thus, an accurate portrayal of program use according to these data is that most minority children who are beneficiaries of the U.S. welfare system live in families that are native born and therefore eligible for public assistance.
Recency of arrival in the United States is not the engine driving high receipt of public assistance among ethnic and racial minorities. Rather, poverty, truncated educational opportunities, meager job opportunities, and assimilation difficulties are the problems that need to be confronted and remedied.
Table 11A-1 Follows On Next Page.
Table 11A-2
Odds Ratios from Logistic Regressions of Program Participation on Generation.
Table 11A-3
Odds Ratios from Logistic Regressions of Program Participation on Ethnic Groups within Generation.
Table 11A-4Odds Ratios from Regression of Public Assistance on Ethnicity and Generation
| Cash | Noncash | Total | |
|---|---|---|---|
| Gen1 | 0.87** | 0.77*** | 0.74*** |
| Gen2 | 0.68*** | 0.75*** | 0.76*** |
| Mexican | 0.99 | 1.40*** | 1.43*** |
| Puerto Rican | 2.98*** | 3.31*** | 3.31*** |
| Cuban | 0.82 | 3.99*** | 3.58*** |
| Latino | 1.82*** | 2.12*** | 2.18*** |
| Black | 2.16*** | 2.49*** | 2.56*** |
| Other | 1.33* | 2.95*** | 3.20*** |
| West Eur | 1.00 | 0.94*** | 0.96* |
| East Eur | 0.70*** | 1.16* | 1.13 |
| Asian | 3.35*** | 1.45*** | 1.71*** |
| Year1 | 0.67*** | 1.02 | 1.00 |
| Year2 | 0.78*** | 0.95 | 0.92** |
| Year3 | 0.77*** | 0.88*** | 0.85*** |
| Year4 | 1.04 | 0.89*** | 0.90*** |
| Agehd | 0.98*** | 0.96*** | 0.97*** |
| Health | 2.63*** | 2.59*** | 2.59*** |
| Married | 0.18*** | 0.22*** | 0.22*** |
| Lths | 2.30*** | 2.23*** | 2.24*** |
| Smec | 0.84*** | 0.70*** | 0.70*** |
| Colp | 0 29*** | 0.29*** | 0.29*** |
| Yngage | 0.97*** | 0.98*** | 0.99*** |
| Nkids | 1.08*** | 1.19*** | 1.17*** |
| South | 1.61*** | 1.49*** | 1.63*** |
| West | 2.67*** | 1.43*** | 1.63*** |
| North | 1.42*** | 1.23*** | 1.26*** |
| Belowpov | 4.93*** | 5.02*** | 4.84*** |
NOTES: Omitted categories include whites and native-born children with native-born parents. Scenario one: ''Cash" is participation in either AFDC, SSI, or other welfare; "Noncash" is participation in either food stamps, Medicaid, or housing or energy assistance; "Total" is participation in any of these programs. Scenario two: "Cash" is participation in either AFDC, SSI, other welfare, unemployment assistance, or general assistance; "Noncash" is participation in either food stamps, Medicaid, housing or energy assistance, or WIC; "Total" is participation in any of these programs.
*** p<.01; **p <.05; * p <. 10.
SOURCE: SIPP (1986-1991 panels).
References
- Blau, F. 1984. The use of transfer payments by immigrants. Industrial and Labor Relations Review 37(2):222-239.
- 1986. Immigration and the U.S. taxpayer. Pp. 89-110 in Essays on Legal and Illegal Immigration, S. Pozo, editor. , ed. Kalamazoo, Mich.: W.E. Upjohn Institute for Employment Research.
- Borjas, G.J., and L. Hilton 1996. Immigration and the welfare state: Immigrant participation and meanstested entitlement programs. Quarterly Journal of Economics CVXI: 575-604.
- Borjas, G.J., and S.J. Trejo 1991. Immigrant participation in the welfare system. Industrial and Labor Relations Review 44(2):195-211.
- Bureau of the Census 1991. SIPP Users Guide. Various issues. Washington, D.C.: U.S. Government Printing Office.
- 1993. Statistical Abstracts of the United States. Various issues. Washington, D.C.: U.S. Government Printing Office.
- Currie, J. 1997. Medicaid use by children of immigrants. Focus 18(2):54-57.
- Edmonston, B., editor. (ed.) 1996. Statistics on U.S. Immigration: An Assessment of Data Needs for Future Research. Committee on National Statistics, National Research Council. Washington, D.C.: National Academy Press.
- Kao, G., and M. Tienda 1995. Optimism and achievement: The educational performance of immigrant youth. Social Science Quarterly 76(1):1-19.
- Primus, W. 1996. Immigration provisions in the new welfare law. Focus 18(2).
- Simon, J. 1981. What immigrants take from and give to the public coffers. U.S. Immigration Policy and the National Interest: Appendix D to Staff Report of the Select Commission on Immigration and Refugee Policy .
- Simon, J., and A. Akbari 1996. Determinants of welfare payment use by immigrants and natives in the United States and Canada. In Immigrants and Immigration Policy: Individual Skills, Family Ties, and Group Identities, H. Duleep, editor; and P. Wunnava, editor. , eds. Greenwich, Conn.: JAI Press.
- Tienda, M., and L. Jensen 1986. Immigration and public assistance participation: Dispelling the myth of dependency. Social Science Research 15(4):372-400.
- U.S. House of Representatives 1993. Background Material and Data on Programs Within the Jurisdiction of the Committee on Ways and Means (Green Book). Various issues. Washington, D.C.: U.S. Government Printing Office.
- U.S. House of Representatives, Ways and Means Committee 1996. Overview of the Entitlement Programs: 1996 Green Book. Washington, D.C.: U.S. Government Printing Office.
Footnotes
- 1
See Primus (1996).
- 2
Refugees are treated differently from other immigrants because of the circumstances under which they arrived in this country. In the early 1990s refugees comprised only about 10 percent of all immigrants, however.
- 3
Title IV of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 changed who gets welfare and how they get it. The crucial point is that the new provisions affect legal and illegal immigrants. For more background, see U.S. House of Representatives (1996).
- 4
Rotation group 1 of the 1986 panel was followed for only 24 months instead of 28 months.
- 5
The empirical approach developed in this study parallels Hofferth's (this volume) study of Mexican, Puerto Rican, and Cuban immigrants. It distinguishes itself from her study by capitalizing on the larger sample size and scope of the SIPP to study all immigrants, including immigrants from specific places such as Asia, Eastern Europe, and Western Europe.
- 6
The SIPP does not report if a household entered the United States with a refugee visa, but it does identify some countries that send refugees: Cuba, Czechoslovakia, Hungry, Poland, the former Soviet Union, and Vietnam.
- 7
The SIPP does not identify children under age 15 who lived independently, although the number undercounted would probably be small and leave results unchanged.
- 8
Results that are based on the child, or the child person-year, may differ from those in which the family is the unit of analysis. For instance, immigrants have larger families, making children in these families underrepresented in family-based analyses.
- 9
Because sample sizes for foreign-born Cuban children and native-born Cuban children with foreign-born parents were too small, these groups were recombined in the regression models.
- 10
Owing to initial data construction of the 1986, 1987, and 1988 SIPP panels, data for SSI and GA were unavailable. Other SIPP panels still possessed large samples, however, numbering 66,465 child person-years. Models were estimated for these programs.
- Receipt Of Public Assistance By Immigrant Children And Their Families: Evidence ...Receipt Of Public Assistance By Immigrant Children And Their Families: Evidence From The Survey Of Income And Program Participation - Children of Immigrants
Your browsing activity is empty.
Activity recording is turned off.
See more...