[ Main Page of Report | Contents of Report ]
This chapter provides general and site-specific contextual background information on demographic characteristics and policy developments relevant to this study. Background information on the four focal programs — TANF, FSP, Medicaid, and SCHIP — is provided in Appendix C.
By 2000 the foreign-born share of the U.S. population had more than doubled to almost 11 percent, from a low of 5 percent three decades earlier in 1970 (Fix and Passel 2001). And growth in immigrant and LEP populations has been accompanied by an increasing dispersal of immigrants across the country. This trend marks a significant shift away from the traditional clustering of the overwhelming majority of foreign-born residents in just a handful of states that characterized immigration patterns in the past.
Nationally, about one fifth of all children and one quarter of children in low-income families (i.e., those with incomes under 200 percent of the federal poverty level) have immigrant parents (The Urban Institute 2001). Data from the Census 2000 Supplementary Survey indicate that, nationally, the number of Limited English Proficient (LEP) adults increased from 12 to 16.5 million between 1990 and 2000. During that time, the LEP share of the total adult population rose from 6 to 8 percent.
The composition of these immigrant families, especially in terms of their mixed immigration status and relatively large family and household size, is of particular relevance to this study. Almost 10 percent of all families in the U.S. are "mixed-status" families, meaning that at least one child is a U.S. citizen but at least one parent is not a citizen. According to 1998 figures from the U.S. Current Population Survey (CPS), 85 percent of families with at least one non-citizen parent have at least one citizen child. The vast majority of immigrant families who apply for benefits have both citizens and non-citizens in them and while non-citizen adults are often ineligible for benefits, their children — who are usually citizens — are generally eligible (Fix and Zimmermann 1999). In addition, immigrant parents have more children on average than native-born parents and they are more likely to live in extended families and multi-family households, making for even larger households.
Citizenship and Immigration Categories Used in This Report
|
The six sites examined in this study are: New York City (five counties/NY), Dallas (Dallas and Tarrant Counties/TX), Seattle (King County/WA), Raleigh (Wake County/NC), Arlington (Arlington County/VA), and Sedalia (Pettis County/MO). These sites vary widely in population size — from New York, the nation's largest city, to Sedalia, a primarily agricultural town in a county of 39,000 (Exhibit 2-1). The sites represent a mix of new immigrant settlement areas and cities with larger, more established immigrant communities. The size and diversity of the immigrant population varies considerably across the sites, as does the share of limited English speakers and diversity of languages represented (for a full description of site selection criteria, see Appendix A).
Characteristics of Immigrant Populations in the Six Study Sites. As shown in Exhibit 2-1, New York City and Dallas both have Hispanic population shares over 25 percent, compared to near 5 percent in Seattle, Raleigh and Sedalia. Asian population shares are highest — about 10 percent — in Seattle, New York and Arlington, while Sedalia's Asian population is under one percent of the county's total population.
| Site | Total Population | Share Hispanic (%) | Share Asian*(%) |
|---|---|---|---|
| Arlington County, VA | 189,453 | 19 | 9 |
| Dallas County, TX | 2,218,899 | 30 | 4 |
| New York City (5 counties), NY | 8,008,278 | 27 | 10 |
| Raleigh (Wake County), NC | 627,846 | 5 | 3 |
| Seattle (King County), WA | 1,737,034 | 6 | 11 |
| Sedalia (Pettis County), MO | 39,403 | 4 | 0.4 |
| * Represents share one race, Asian.
Some other Asians may have reported more than one race. Source: U.S. Census Bureau. 2002. "Census 2000 Summary File 1 (SF-1) 100-Percent Data Table DP-1." American FactFinder. Available at http://factfinder.census.gov/servlet/BasicFactsServlet . |
|||
Three study sites have foreign-born population shares of 19 percent or over (Exhibit 2-2). New York City has the highest share of foreign-born and non-citizen populations, followed by Arlington, then Dallas County.(1) Only Sedalia (Pettis County) has a foreign-born population below 10 percent. In all sites except New York City and Seattle (King County), more than half of the foreign-born population arrived during the 1990s, and in every site the majority of immigrants were non-citizens rather than naturalized citizens. Sedalia and Raleigh (Wake County) represent "new settlement" sites, with the highest shares of immigrants entering during the 1990s — 70 and 64 percent respectively. In New York City, by contrast, only 43 percent of immigrants entered during the 1990s.
| Site | Foreign-Born Population | Non-Citizen Population | Entered During 1990's | |||
|---|---|---|---|---|---|---|
| Number | Percent of Total Population | Number | Percent of Total Population | Number | Percent of Total Population | |
| Arlington Co., VA | 52,693 | 28% | 38,300 | 20% | 30,543 | 16% |
| Dallas Co., TX | 463,574 | 21 | 361,373 | 16 | 264,942 | 12 |
| King Co. WA | 268,285 | 15 | 149,849 | 9 | 131,848 | 8 |
| New York, NY | 2,871,032 | 36 | 1,592,345 | 20 | 1,224,524 | 15 |
| Pettis Co., MO | 1,158 | 3 | 885 | 2 | 806 | 2 |
| Wake Co., NC | 60,602 | 10 | 44,240 | 7 | 38,994 | 6 |
| Source: U.S. Census Bureau. 2002. "Census 2000 Table DP-2. Profile of Selected Social Characteristics: 2000." American FactFinder. Available at http://factfinder.census.gov/servlet/BasicFactsServlet. | ||||||
According to program staff and advocates, the composition of the immigrant population in terms of immigration status also varies greatly across and even within local sites. Some sites are more likely than others to encounter refugees applying for benefits, because refugee resettlement is generally concentrated in a few major cities. Undocumented populations tend to be larger (at least as a share of all immigrants) in new settlement areas, while the share of naturalized citizens among the foreign-born is higher in locations with more established immigrant communities.
The contrast between newer settlement areas versus more established immigration destinations reflects the increasing dispersal of immigrants across new areas of the country, a pattern that has important implications for language access issues considered in this study. Welfare and other human service agencies in localities with large, established immigrant communities — such as New York, Dallas, Seattle and Arlington — have more experience serving non-citizens and greater community resources to draw upon for interpretation and translation services. Human service agencies in the newer settlement areas, represented in our study by Raleigh and Sedalia, have much less experience with and far fewer resources to address language and other issues experienced by these populations.
Even communities accustomed to receiving large numbers of immigrants are now home to an increasingly diverse number of immigrant groups who speak dozens of different languages. This diversity presents its own set of challenges for human service agencies accustomed to dealing with fewer immigrant and language groups. For example, according to study respondents, Arlington, Seattle and Dallas received substantial numbers of new refugees from Southeast Asia during the 1980s, but by 2001 most refugees living there were from a broader range of nations including Africa, South Asia, the Middle East, and Eastern Europe. According to a 1999-2000 Urban Institute survey, New York City has immigrants from over 100 countries (Capps et. al. 2002).
Characteristics of LEP Populations in the Six Study Sites. The largest cities — New York, Dallas and Seattle — have large and diverse immigrant populations speaking a wide variety of languages other than English. In New York City, Spanish speakers predominate, numbering nearly 2 million people, but there are five other languages with about 100,000 or more speakers, and over a dozen with more than 50,000 speakers. In Seattle, no single language is in the majority among non-English speakers and LEP persons speaking Asian languages outnumber Spanish speakers by more than two-to-one. In Dallas and Arlington, Spanish speakers predominate but there are also substantial numbers who speak Vietnamese and other languages. In Raleigh (Wake County) and Sedalia (Pettis County), the overall LEP population is relatively small and the most prevalent language spoken is Spanish. (See Appendix B for additional site-specific data on languages spoken in the home.)
| Site | LEP Population** Spanish Language |
LEP Population* Asian Languages |
LEP Population* Other Languages |
|||
|---|---|---|---|---|---|---|
| Number of LEP Adults | Percent of Total Adult Population | Number of LEP Adults | Percent of Total Adult Population | Number of LEP Adults | Percent of Total Adult Population | |
| Arlington Co., VA | 19,277 | 11 | 4,623 | 3% | 5,893 | 3% |
| Dallas Co., TX | 316,065 | 16 | 31,347 | 2 | 19,831 | 1 |
| King Co., WA | 32,382 | 2 | 72,219 | 4 | 33,070 | 2 |
| New York City, NY | 921,324 | 12 | 325,321 | 4 | 522,332 | 7 |
| Pettis Co., MO | 693 | 2 | 29 | 0.1 | 247 | 0.7 |
| Wake Co., NC | 20,542 | 4 | 5,697 | 1 | 6,228 | 1 |
| * Limited English Proficient persons speak English
"well", "not well" or "not at all" (i.e., not "very well") on the Census
Bureau's standard four-part question. The Census reports these figures for
the population ages 5 and over. Source: U.S. Census Bureau. 2002. "Census 2000 Table DP-2. Profile of Selected Social Characteristics: 2000." American FactFinder. Available at http://factfinder.census.gov/servlet/BasicFactsServlet. |
||||||
With the exception of New York City, the welfare agencies in our study sites do not track data on the English proficiency or immigration status of applicants or clients. The language composition picture that emerged from discussions with study participants closely matches that painted by the Census data, however. New York City and Seattle have the most language diversity in their applicant populations. TANF and Food Stamp recipients in New York speak at least 54 different languages, with the most common languages being Spanish, Russian, Chinese, Korean, Creole, French, Arabic, Yiddish and Vietnamese. One office in Seattle reported that speakers of about 50 languages came through their door during a single year. Languages commonly spoken by applicants and recipients other than Spanish in Seattle are Russian, Ukrainian, Somali and languages spoken in other African nations. Seattle agency staff reported that Cambodian, Vietnamese and Laotian speakers used to make up a larger share of applicants and recipients than is currently the case.
Eligibility workers and application assistants in Dallas reported encountering substantial numbers of Asians (including Vietnamese, Cambodian, Laotian, Korean and Chinese speakers) and, to a lesser extent, Africans and Middle Easterners (who speak a wide variety of languages including Arabic, Farsi, Somali and Swahili). In Arlington, most LEP applicants speak Spanish but a very small share speak a variety of other languages and, like Seattle, the share of applicants and clients who speak Cambodian, Vietnamese or Laotian has declined significantly over time, as refugee flows from Southeast Asia have subsided.
[Go to Contents]
After PRWORA, most post-enactment legal immigrants (those who arrived in this country after August 22, 1996) are not eligible for federally-funded Temporary Assistance for Needy Families (TANF), Medicaid, and SCHIP(2) during their first five years in the country. States have discretion to decide whether to extend the bar to include pre-enactment legal immigrants.(3) PRWORA also dropped food stamp eligibility for legal immigrant adults, regardless of date of entry, until they naturalize or prove that they (or their spouse or parents) worked in the country for a combined total of at least ten years. Finally, PRWORA exempted refugees during their first five years in the country and a few other small classes of immigrants from these legal immigrant eligibility bars.(4)
The federal government subsequently enacted partial restorations of legal immigrants' eligibility for food stamps in a piecemeal fashion. In 1998 Congress restored food stamp eligibility to children and disabled adults who entered the country before August 22, 1996, as well as to immigrants who had their 65th birthday before that date.(5) At the time of our site visits, working-age legal immigrants entering before August 22, 1996 and adult legal immigrants entering after enactment remained ineligible, unless they could demonstrate 10 years of work history or meet other federal exemptions. In May 2002, Congress replaced the 10-year work requirement for legal immigrant adults with a five-year bar consistent with that for TANF, Medicaid and SCHIP, although these changes do not take effect until April 2003. Eligibility was also restored for all legal immigrant children, regardless of date of entry, effective October 2003.(6)
PRWORA's immigrant eligibility provisions apply to the expenditure of federal funds for these programs, but not to state funding. Some states have opted to use their own state dollars to extend substitute TANF, food stamp, Medicaid and/or SCHIP benefits to post-enactment legal immigrant families. For example, 16 states provide some form of food assistance for post-enactment legal permanent residents (Schwartz 2001), and 19 states provide TANF replacement programs. Twenty-three states fund Medicaid for post-enactment legal immigrants, and another three partially restore these health benefits (Zimmermann and Tumlin 1999).
Of the six states included in this study, only Washington fully provides eligibility for all four programs to post-enactment legal immigrants. In contrast, North Carolina, Texas and Virginia do not provide state-only funding for any of these benefits for families with children. (7) New York and Missouri provide eligibility for some benefits, but not others (Exhibit 2-4). Missouri uses state-only funding to provide for post-enactment legal immigrants who are not eligible for TANF and extends food stamp eligibility to those non-citizens who receive cash assistance through this TANF replacement program. New York provides a state-funded replacement program for Medicaid and SCHIP as well as a variation of TANF cash assistance called Safety Net Assistance (SNA).(8) In New York City, the SNA program provides vouchers for rent to legal non-citizens barred from the TANF program (which accounts for most of the monthly grant), and the remainder of the SNA benefit in cash.
| Site | TANF | Food Stamps | Medicaid | SCHIP |
|---|---|---|---|---|
| Arlington, VA | ||||
| Dallas, TX | ||||
| New York, NY | ü | ü | ü | |
| Raleigh, NC | ||||
| Seattle, WA | ü | ü | ü | ü |
| Sedalia, MO | ü | * | ||
| * In Missouri, post-enactment legal immigrants who receive state replacement TANF benefits are also eligible to receive state-funded food stamps. Post-enactment legal immigrantswho do not receive this state-funded cash assistance, however, are not eligible for the state-funded food stamp benefits. | ||||
As mentioned earlier, there were significant caseload declines in public benefit programs in the years immediately following the enactment of PRWORA.(9) Legal immigrants' use of public assistance declined as well and while some of this decline may have been due to changes in eligibility, there was some speculation that some of the decline may have been the result of caseworkers and clients having difficulty distinguishing between eligible and ineligibility immigrants.(10) As of 1999, when only low-income families are considered, legal immigrants with children had lower participation rates for TANF and food stamps than their low-income citizen counterparts, but Medicaid participation was on par with participation among citizen families (Fix and Passel 2002). Another possible explanation for the comparatively low benefit program participation among legal immigrants — including those with citizen children — is that some may not apply for benefits due to fears and misconceptions about eligibility rules and the potential for benefit participation to have negative consequences for their immigration status and applications for citizenship.
More broadly, the caseload declines associated with welfare reform have led to increased interest in how program policies and administrative practices affected access to benefits. Various initiatives have since been undertaken to increase low income families' access to medical assistance and, to a lesser extent, food stamps by reducing barriers such as complex application processes, stigma or general lack of awareness concerning eligibility. The enactment of SCHIP in 1997 heightened federal, state, and local efforts to increase access to medical assistance for children through outreach and simplified application processes. To a large degree, these outreach and simplification efforts are credited for bringing about dramatic increases in SCHIP participation since the program's initial implementation and the turnaround in Medicaid caseload declines.
Finally, access issues related to limited English proficiency have also begun to receive greater attention. At the federal level, there have been a series of developments beginning in 2000 with Executive Order 13166, which required each federal agency to issue guidance for improving access to programs and activities funded by that agency for individuals with limited English proficiency. The U.S. Department of Justice subsequently issued complementary guidance to public agencies on how to ensure "meaningful access" for limited English speakers to public programs in 2000, and undertook a formal review of Executive Order 13166 and language access policies issued by all federal agencies during 2001-2002. This process led to issuance of a report on the costs and benefits of the Executive Order (Office of Management and Budget 2002) and final guidance to federal agencies in June 2002 (Department of Justice 2002).
[Go to Contents]
1. Naturalized citizens account for the difference in population size between the non-citizen and total foreign-born populations.
2. SCHIP was authorized after welfare reform was enacted, but it is similar to Medicaid in terms of benefit restrictions for non-citizens.
3. Only two states decided to exclude pre-enactment legal immigrants from public benefit eligibility: Alabama excluded them from TANF, and Wyoming excluded them from Medicaid. (Zimmermann and Tumlin 1999).
4. Other groups exempted from the bar on eligibility include asylees, Amerasians and Cuban/Haitians (for five years), as well as active-duty military, veterans, and their dependents.
5. Food stamp eligibility was also extended from five to seven years after entry for refugees and asylees. Agriculture, Research, Extension, and Education Reform Act, P.L. 105-185 (1998).
6. Farm Security Act, P.L. 107-171 (2002).
7. New York State's Food Assistance Program and the State Immigrant Food Assistance Program in Texas also cover some post-enactment elderly immigrants.
8. Additionally, New York provides state-funded health insurance for undocumented and other immigrant children considered not "qualified" under PRWORA in the Child Health Plus program.
9. According to federal administrative data, from 1996 to 1999, TANF caseloads dropped by 42 percent, while food stamp caseloads dropped by 29 percent nationally. Medicaid participation also fell slightly between 1996 and 1997, but then began to rise in 1998 and showed an overall increase of 13 percent by 1999. If low-income children insured under SCHIP are taken into account, this figure is 18 percent (see Appendix A).
10. Between 1994 and 1999, TANF non-citizen participation declined 60 percent, Food Stamp Program participation declined 48 percent and Medicaid participation declined 15 percent. (Fix and Passel 2002).
Main Page of Report | Contents of Report
Home Pages:
Human Services Policy
Assistant Secretary for Planning and Evaluation
(ASPE)
U.S. Department of Health and Human Services
(HHS)