One of the more contentious issues in the 1996 welfare reform debate was whether the federal government should provide welfare benefits to non-citizens who are legal residents of the United States. The sometimes bitter debate revealed a fundamental divide in how advocates, analysts, and policymakers think about welfare for non-citizens. On the one hand, those who support welfare benefits for non-citizens point out that at the time of the 1996 reforms, legal non-citizens enjoyed access to a wide range of welfare benefits. This access was based on the principle that non-citizens come to America to participate in the full range of American social, economic, and political life and that, with modest exceptions, they should be treated like other Americans. The children of non-citizens, the vast majority of whom are American citizens, are especially deserving of the safety net provided by welfare programs. Supporters believe that to deny non-citizens and their children welfare benefits is to leave them outside the protective sphere of social welfare guaranteed to the disabled and destitute by federal and state government policy. Moreover, like other Americans, non-citizens pay taxes, and unlike non-citizens in many other countries, can be drafted in time of war.
By contrast, those who oppose benefits for non-citizens argue that restricting their access to welfare has been a principle of American domestic policy since colonial times. The federal law, established in 1882 and strengthened in the early twentieth century, was that immigration officials should refuse entry to any non-citizen who appeared likely to become a "public charge" and should deport those who did, although actual deportations of public charges have been rare. Those supporting the 1996 reforms also point out that, both in statutory law and in the minds of American citizens as revealed in polls, welfare for non-citizens has always been suspect. The congressional prohibitions from the late nineteenth century have already been mentioned, but more recent restrictions are also notable. Specifically, in 1993 the Democrati-cally controlled Congress extended from three to five years the period that immigrants entering the country had to wait before they could qualify for the Supplemental Security Income (SSI) program. Thus, the principle that non-citizens do not have the right to benefits on the same basis as citizens has long been established and maintained by the Congress. Finally, those in favor of benefit cuts for non-citizens have been strongly motivated to save taxpayer dollars and balance the budget
In 1996, those who supported reduced welfare for non-citizens, primarily Republicans, had the upper hand in both houses of Congress. The reforms they enacted were sweeping in both their intent and effect. As the 2002 reauthorization debate begins, those opposed to the 1996 reforms and those in favor of them are preparing for another lively debate on both the fundamental principles and the specific provisions of federal law on welfare for non-citizens. There is a good chance that Congress will enact some benefit expansions.
Overview of 1996 Provisions
The 1996 reforms changed almost every aspect of non-citizen eligibility for welfare benefits. Although the provisions are exceedingly complex, a rough general principle provides useful guidance. With some exceptions, non-citizens entering the United States after August 22, 1996, the date of enactment of the welfare reform legislation, are not eligible for most welfare benefits, including Temporary Assistance for Needy Families (TANF), SSI, Medicaid, and the State Children's Health Insurance Program (SCHIP), until they have been in the U.S. for at least five years. Keeping this general principle in mind, a more thorough review of the 1996 provisions reveals both their complexity and breadth.
In determining welfare eligibility for non-citizens, two broad criteria are taken into account. The most general screen is the concept of "qualified" and "not qualified" alien. Generally, qualified aliens are non-citizens who have been permitted to reside permanently in the United States. They include legal permanent residents (LPRs), refugees, asylees, Cuban/Haitian entrants, and a few other categories. Qualified immigrants may be eligible for federal and state benefit programs that aim to help families with limited income and resources. Not qualified aliens (mostly illegal and temporary immigrants), by contrast, are ineligible for all except emergency benefits?a policy that was largely in place even before the 1996 reforms.
The second factor is date of entry into the United States. Specifically, qualified immigrants who entered after August 22, 1996 are barred from SSI and food stamps until they become citizens and from TANF, Medicaid, and SCHIP for five years after entry. Qualified immigrants, who entered before August 22, 1996, have wider eligibility for these benefits, in part because states have elected to pay for the benefits and in part because federal legislation enacted since 1996 restored some benefits for those already in the United States. As a result, many immigrants who entered before 1996 are eligible for TANF, Medicaid, SCHIP, and SSI. Eligibility for food stamps is more restricted.
In addition to the two broad criteria, there are several other factors that condition non-citizens' eligibility for welfare benefits. Children born to either qualified or not qualified non-citizens after they have entered the United States are citizens and are therefore eligible for benefits on the same basis as native citizens. Refugees and asylees are eligible for all welfare benefits for the first seven years they reside in the U.S., after which their eligibility is greatly reduced. Non-citizens who have worked for ten years and armed forces personnel and their dependents are eligible for all benefits. Finally, some emergency benefits?especially Medicaid?are provided to all non-citizens.
A major provision of the 1996 legislation required that most legal immigrants have sponsors with incomes over 125 percent of the federal poverty line. The provision also required the income of sponsors to be deemed as available to the immigrants when calculating eligibility for welfare benefits, which usually has the effect of disqualifying them from benefits. The sponsor was also held liable for the costs of any welfare benefit used by the immigrant. The purpose of these requirements was to ensure that sponsors assume financial responsibility for newcomers, thereby reducing the potential burden on taxpayers.
These provisions constitute a revolution in welfare policy for non-citizens. But there is widespread disagreement about their fairness. The fairness issue will be a major part of the reauthorization debate over welfare for non-citizens. But the debate will also be informed by the actual effects of the 1996 changes, a topic to which we now turn our attention.
Effects of the 1996 Reforms
Declining Use of Public Benefits by Type of Family
The intent of the 1996 reforms was to reduce the use of welfare by non-citizens. Studies by Michael Fix, Jeffrey Passel, and Wendy Zimmermann at the Urban Institute show that use of public benefits by LPRs not only declined sharply, but did so at a faster rate than use of benefits by citizens (the 1996 reforms also reduced eligibility for many benefit programs among citizens).
Data for 1999 from the Current Population Survey conducted by the U.S. Census Bureau show that there has been a sustained decline in legal non-citizens' use of TANF, SSI, food stamps, and Medicaid or SCHIP from 1994 through 1999. As shown in Figure 1, the sharpest decreases occurred in TANF (60 percent) and food stamps (48 percent) and the lowest in Medicaid (15 percent). The declines in participation rates for LPR families exceeded the declines experienced by citizen families for TANF, SSI, and food stamps, but not Medicaid (not shown).
Fix and Passel have tried to disentangle the effects of the law's immigrant restrictions on different types of low-income families with children: those headed by LPRs, those headed by refugees, and those headed by citizens. The trends in benefit receipt by low-income LPR families with children reveal that they experienced large declines in use of TANF (53 percent) and food stamps (38 percent) between 1994 and 1999. But their participation in Medicaid and SCHIP remained essentially unchanged. Over the period, poor LPRs and poor citizens' use of most benefits declined at about the same rate. By 1999, poor LPR families were less likely to use TANF and food stamps than citizens were, while they were slightly more likely to use Medicaid.
The benefit cuts not only led to reduced benefit use among the LPRs who had been the principal targets of reform, but also spilled over to U.S.-born citizen children who live in families in which one or both parents is a non-citizen. These children, who are fully eligible for benefits, constitute a surprisingly large group. Roughly three quarters of all children living in immigrant families with non-citizen parents are citizens. Moreover, 10 percent of all children and 15 percent of poor children in the United States live in mixed status families. These families have substantially lower participation rates in benefit programs than citizen families for TANF (7.8 versus 11.6 percent) and food stamps (19.8 versus 27.9 percent). For both programs, mixed status families experienced significant declines in participation from 1994 to 1999.
There have also been striking declines in benefit use among refugees. Refugees are a protected population under welfare reform, retaining their eligibility for federal benefits for seven years after entry. Nonetheless, between 1994 and 1999 steep declines occurred in use of TANF (78 percent), food stamps (53 percent), and Medicaid (36 percent) by low-income refugee families with children. Before welfare reform, refugee families had use rates that were more than double those for either citizen or LPR families. However, by 1999 the rates for refugee families had fallen to levels roughly equal to those of citizens.
As noted, there was no change in receipt of Medicaid (including SCHIP) among low-income LPR families with children between 1994 and 1999. There are a number of explanations for these stable Medicaid use rates. These include the introduction of expanded health care coverage under SCHIP, stepped up state and local outreach for child health insurance, and the impact of new federal guidelines clarifying that use of health benefits would not be a bar to obtaining a work permit ("green card") or citizenship. In addition, the doctors, hospitals, and clinics that provide health care have a strong incentive to keep both immigrants and natives enrolled in government health programs to ensure the payment of medical bills. Other welfare programs do not have third parties who have such direct incentives to make sure low-income families are signed up for welfare benefits. Another possible explanation for the relatively modest decline among LPR families with children may be increased use of emergency Medicaid by legal immigrants.
Despite the lack of decline in Medicaid use between 1994 and 1999, the percentage of LPR children who are uninsured remained higher than that of other children. Less than 20 percent of poor citizens' children were uninsured in 1999, while the rate among citizen children of LPRs was 27 percent and the rate for non-citizen children of LPRs was 39 percent.
Declines in immigrant use of benefits are evident nationwide, but they are especially pronounced among the states that offer the least generous safety nets such as Texas and Florida. As a group, the thirty least generous states experienced rapid growth in their immigrant populations between 1995 and 2000, with the number of foreign born families with children rising by 31 percent - more than four times the rate of the remaining twenty states. TANF participation among low-income LPR families with children dropped 73 percent in the least generous states between 1994 and 1999 as compared with 45 percent among the more generous states. As might have been predicted, the 1996 reforms, which gave more discretion to states in determining non-citizens' eligibility for welfare benefits, have led to substantial differences across states in benefit eligibility and benefit use.
Taken together, these migration and policy trends cast doubt on the welfare magnet theory, which holds that immigrants are drawn to places with generous welfare benefits. These trends also raise the concern that low-income LPR families will find themselves outside increasingly localized safety nets as the labor market tightens.
Explaining the Trends
How do we explain these declines in benefit use among non-citizens? Increased naturalizations could reduce the number of non-citizens receiving benefits if the immigrants who become citizens do so to continue receiving welfare benefits. Between 1994 and 1999, the number of non-citizen families that naturalized did increase rapidly. These increases were the product of a number of demographic and policy shifts that include, but go beyond, welfare reform. For example, increased naturalizations also resulted from the fact that 2.7 million immigrants acquired legal status around 1990 under the 1986 Immigration Reform and Control Act and thus became eligible to become citizens in the mid-1990s.
This rise in the number of naturalizations, however, can account for only a small proportion of the drop in benefit receipt by non-citizens. While the number of families containing a naturalized citizen grew by 1.5 million between 1994 and 1999, the number of such families participating in welfare programs rose by only 170,000. At the same time, the number of LPR families on welfare dropped by 480,000. Thus, increased naturalizations fall well short of offsetting decreases in benefit use among non-citizens.
If naturalizations account for only a fraction of the decline in LPR use of welfare, perhaps increased income?which would lower benefit eligibility?can explain the decline. However, Urban Institute analyses reveal that only about one-quarter of the reduction in TANF and food stamp participation rates for both citizens and LPR families is explained by changes in income. Thus, rising incomes also explain only a fraction of the decline in welfare use.
Research by Wendy Zimmermann and Karen Tumlin at the Urban Institute shows that, while fewer immigrants are joining TANF, non-citizens are having a difficult time leaving. They find that legal immigrants make up a significant share of the remaining caseload in several major cities and that, as compared with native American TANF recipients, immigrants in these cities face a number of barriers to entering the workforce, including limited English proficiency, less education, and less recent work experience than native workers. The researchers also found that the programs needed to help potential non-citizen workers overcome these barriers were often not available.
To sum up, welfare reform's immigrant restrictions have led to a rapid decline in TANF and food stamp use among LPR families with children, citizen children in mixed status families, and refugees. There have also been declines in the use of Medicaid among legal, working-age adult immigrants, but not among poor non-citizen families with children or children themselves. Increased naturalization rates and higher incomes contributed to the declines in benefit receipt by non-citizens, but they fall well short of accounting for the entire decline. We are left to conclude that the benefit cuts of 1996 directly contributed to the decline in welfare use by non-citizens.
Hardship Among Non-citizens
Whatever the cause, there is some evidence that declining welfare use by non-citizen families may be associated with increased hardship. Randy Capps at the Urban Institute found that roughly one half of all immigrant families with children had incomes below 200 percent of poverty and that the children of immigrants were more likely than the children of natives to have no health insurance (22 versus 10 percent) and to have some difficulty getting enough food to eat (37 versus 27 percent). Similarly, George Borjas of Harvard has produced strong evidence that the exclusion of immigrants from food stamps is leading to rising food insecurity among non-citizen households.
The Policy and Politics of Reauthorization
The reauthorization debate will feature proposals by both the Bush administration and by members of Congress to expand welfare benefits for non-citizens. The most important proposals will be those that expand benefits for non-citizens who arrived after 1996 because the distinction between pre-1996 and post-1996 entrants constitutes something like a line in the sand for policymakers. Although there have already been several changes in the 1996 reforms, no legislation has been enacted that moved benefits for non-citizens across the 1996 line of demarcation. Given what is known about human longevity, if benefits are not extended to post-1996 entrants, there will eventually be very few non-citizens receiving welfare benefits.
In penetrating the 1996 line in the sand, it is unlikely that there will be serious attention to proposals that attempt a complete restoration of benefits for non-citizens. The major reason is that a complete restoration of food stamp, TANF, SSI, and Medicaid benefits would be very expensive. According to the Congressional Budget Office, just restoring SSI and Medicaid to non-citizens who qualify for SSI would cost nearly $25 billion over five years. Under congressional budget rules, increased spending must be offset either by equivalent cuts in other programs or by tax increases, neither of which is feasible in this case. Thus, proposals are likely to be more modest than complete restoration of benefits.
In recent months, the president and members of the House and Senate have offered several proposals that would expand the benefits of several groups of non-citizens including children, the elderly, and the disabled. The Senate Farm Bill included several provisions expanding food stamp benefits for children, working non-citizens, refugees, and the disabled that would cost around $1.1 billion over ten years. President Bush recently proposed extending food stamps to post-1996 immigrants who have been in the country for five years, essentially applying the same five-year ban on food stamps now imposed on TANF and Medicaid. Proposals to expand post-1996 immigrants' eligibility for Medicaid, TANF, and, on a more restricted basis, SSI for the disabled are also likely. One legislative proposal already advanced would expand Medicaid coverage for immigrant children and pregnant women. When fully implemented, Medicaid proposals of this type would cost around $2 billion over five years.
The politics of these proposals are fascinating. It is useful to recall that the original sweeping changes in welfare policy for non-citizens were enacted as part of the most comprehensive welfare reform legislation since the New Deal of 1935. Thus, many individual provisions of the immense bill, including the cuts in benefits for non-citizens, were swept along with the overall package and did not receive extensive scrutiny from Congress. Nor were there separate votes on many of these provisions on the floor of either house of Congress. It is not clear, then, whether the non-citizen reforms would have passed if a separate vote had been required. Moreover, in 1995 and 1996, when Congress shaped and then approved the final welfare reform package, the overriding goal of Congress and President Clinton was saving money so that the federal government could reach a balanced budget. Over half the savings in the 1996 welfare reform legislation were in the cuts for non-citizens (although Clinton did not support them).
Conditions in 2002 are sharply different than those in 1996. There is less pressure to achieve budget cuts and there are no prominent proposals to deepen the cuts in welfare for non-citizens. Democrats now control the Senate, and a Republican president has proposed to partially reverse the cuts in food stamps for non-citizens. These shifts respond, in part, to a changing electorate. The Census Bureau reported last year that Hispanics had moved ahead of African Americans as the nation's largest minority population. Cuts in non-citizens' benefits are opposed by ethnic groups that count substantial numbers of immigrants among their number, making many politicians reluctant to support cuts. This is especially the case for Republicans, the party that proposed and enacted the 1996 cuts in welfare benefits for non-citizens. Because many Republicans are trying to improve their image among minority groups?especially Hispanics?it may not be surprising that the president has taken the lead in proposing an expansion in non-citizen benefits in his 2003 budget. Finally, many states believe they have been forced to spend more money on non-citizens due to their loss of federal benefits. Especially given the pressure on state budgets caused by the recession, states can be expected to support all proposals that restore federal benefits to non-citizens.
The bottom line is that at least some modest restoration of benefits, including some that provide eligibility to post-1996 entrants, seems destined to have major support from Congress and the Bush administration. Thus, if Congress finds a way to finance a moderate expansion of non-citizen welfare benefits, the legislation has a good chance of becoming law. A modest restoration would then leave the door open to future expansions.