The process used to select sites was designed to yield a sample that is diverse in terms of the policy environment as well as the immigrant population's total size, language diversity, and mix of citizenship and immigration status. More specifically, the criteria considered when selecting the six sites included: (1) demographics of the immigrant population; (2) trends in public benefit caseloads; and (3) availability of safety net assistance for legal immigrants.
Immigrant-Related Demographic Data. Data on metropolitan area characteristics from the U.S. Current Population Survey (CPS) 1 were used to rank sites according to the following demographic factors: total population; foreign-born share of the total population; share of the foreign-born population entering within last ten years; and share of the foreign-born population from the top three countries of birth (as a measure of immigrant diversity). Sites were selected with the intent of achieving a mixture of sites with large and diverse immigrant populations, as well sites that are new settlement areas (i.e., those with small but rapidly growing immigrant populations).
The final six sites chosen for the study vary considerably along these selection criteria. For example, the sites range in total population size from New York, the largest city in the country, to Sedalia (MO), a small Midwestern town in a county of about 39,000 people. The foreign-born share of the population is also highest in New York (33 percent in the metropolitan area) and lowest in Raleigh (5 percent) and Sedalia (estimated at 5 percent by respondents). The foreign-born populations of Raleigh and Sedalia grew very rapidly during the 1990s, making these two localities new immigrant settlement areas.
Public Benefit Caseload Trends. TANF, FSP and Medicaid caseload participation between 1996 and 1999 were examined to identify general caseload trends in the four focal programs. As shown in Exhibit A-1, AFDC/TANF declines were greatest in Texas, followed by North Carolina and Missouri. FSP declines were also greatest in Texas, followed by Washington and Virginia. Although Medicaid coverage dropped nationally and in some of the states included in our study between 1996 and 1997, by 1999 coverage had increased in all six states: by single digit percentages in New York, North Carolina and Texas, and by double-digit shares in Missouri and Virginia. In all five states except Texas, the growth in the new SCHIP program brought increases in public health insurance coverage into the double digits, but by 2000 Texas's SCHIP program had also enrolled very large numbers of children.
|* Washington State Medicaid figures are for Fiscal Year 1998, because figures for 1999 are incomplete on agency websites.
Source: Program trend data were obtained from the following agency websites: TANF — U.S. Department of Health and Human Services, Agency for Children and Families (http://www.dhhs.gov/new/stats/newstat2.html); Food Stamp Program — U.S. Department of Agriculture, Food and Nutrition Services, (http://www.fns.usda.gov/pd/fspmain.htm); Medicaid: U.S. Department of Health and Human Services, Centers for Medicare and Medicaid Services, (http://www.hcfa.gov/medicaid/msis/mstats.htm); SCHIP — U.S. Department of Health and Human Services, Centers for Medicare and Medicaid Services, (http://www.hcfa.gov/init/children.htm)
Availability of Safety Net Assistance for Legal Immigrants. In order to capture differences in state choices about providing benefits to non-citizens, the availability of state substitute programs for legal immigrants losing eligibility for TANF, food stamps and/or Medicaid under PRWORA was examined in combination with state rankings previously developed by the Urban Institute based on the availability of several different benefit programs, using data from 1998 (Zimmermann and Tumlin 1999). Of the final sites selected, the states of Washington and Missouri were ranked "most available" (i.e., state substitute benefit programs for legal immigrants are most available in these states). New York was ranked "somewhat available," North Carolina and Virginia "less available," and Texas "least available." It should be noted that "less available" and "least available" were the most common rankings when all 50 states are considered, with a total of 21 states ranked "less available," and another 12 states ranked "least available."