Presented to:
Office of the Assistant Secretary for Planning and Evaluation
Department of Health and Human Services

This work was conducted under Subtask 2.2.12 of HHS Contract HHS-100-94-1009. Many constructive comments were provided by staff of the Department of Health and Human Services, including Linda Sanches, David Nielsen, Penelope Pine and Bob Tomlinson. We gratefully acknowledge data and advice made available by Ron North and Roger Buchanan of the Health Care Financing Administration and Charles Scott of the Social Security Administration. Many colleagues at the Urban Institute offered useful advice or data, including Brian Bruen, Rebecca Clark, Teresa Coughlin, Linda Giannarelli, Jeff Passel, Karen Tumlin and Wendy Zimmerman. All opinions expressed are the authors' and should not be interpreted as opinions of the Urban Institute or the Department of Health and Human Services.

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Assessing the impact of this change has been problematic because there are few data about the number and costs of immigrants on Medicaid. This report provides national and state-by-state estimates of the number and cost of noncitizen Medicaid beneficiaries.

Methods and Data. The analyses are based on the Medicaid Quality Control (QC) data base for the first half of 1994, with additional information about Supplemental Security Income (SSI) participants from the Social Security Administration (SSA). The QC data base includes verified data about 93,000 sampled Medicaid enrollees, roughly 2,000 per state, making it the largest known national sample of Medicaid beneficiaries. QC data include relatively detailed information about immigration status, as well as medical expenditures paid by Medicaid in the sample month. The QC sample represents the great majority of Medicaid beneficiaries, but in 31 states the QC sample excludes SSI recipients. Thus, SSA data are used to supplement information about elderly and disabled immigrants, although these data are more limited. Estimates of the number and costs of aged, blind and disabled Medicaid beneficiaries who are noncitizens were generated by combining QC and SSA data.

National Participation Estimates. Overall national estimates of the number of noncitizen immigrants on Medicaid, based on the combination of QC and SSA data, are presented in Table ES-1. About 2.4 million of the 32 million Medicaid enrollees in an average month in 1994 were noncitizens (including undocumented aliens with emergency coverage only). This is 7.5 percent of the total caseload. Measured another way, about 3.2 million immigrants were enrolled in Medicaid over the course of a year. Insofar as noncitizens are 12.6 percent of the population under poverty, according to the 1996 Current Population Survey, the number of immigrants on Medicaid is less than might be expected given their poverty. Adult and aged beneficiaries were more likely to be immigrants than were children or the disabled. A major reason for the low percentage among children is that immigrants' children are often native citizens born in the U.S. The total Medicaid expenditures for noncitizens (excluding Disproportionate Share Hospital or DSH payments) were $8.1 billion, or 6.9 percent of total expenditures.

PRWORA and later legislation create a very complex set of rules for determining the Medicaid eligibility of immigrants. While legal permanent residents are barred from full Medicaid coverage for their first five years in the U.S., refugees and asylees may participate in Medicaid for their first seven years in the country. Undocumented aliens were already barred from full participation in Medicaid. More detailed data on the citizenship/immigration status of Medicaid beneficiaries are only available in the QC data base, which lacks complete information about aged, blind and disabled beneficiaries. In the QC data base, 7.4 percent of Medicaid enrollees are noncitizens, which is slightly different than the combined estimate of 7.5 percent. Table ES-2 provides national estimates of more detailed immigration status, based on the QC data in 1994.

Only a small share of the pre-enactment (i.e., arrived in the U.S. before August 1996) Medicaid enrollees has lost eligibility, the "not qualified" group, including PRUCOL immigrants. The main effects are for post-enactment (i.e., arrive after August 1996) regular immigrants, who will be ineligible for Medicaid for five years. Because of changes in the rules for "deeming" income, a large fraction (perhaps most) of the immigrants will still be ineligible after five years expires. In determining eligibility, the income of those who sponsored immigrants will be deemed available to them, making it more difficult for immigrants to qualify as sufficiently poor. The Balanced Budget Act of 1997 modified PRWORA, so that most immigrants who were on SSI in August 1996 will retain both SSI and Medicaid benefits. Those in the U.S. before that date who later become disabled can also qualify for SSI. However, the elderly and disabled immigrants arriving after August 1996 will not qualify for SSI or Medicaid unless they become citizens or meet other exemptions. In general, the immigrant-related rules no longer apply for those immigrants who naturalize into U.S. citizenship.

The discussion above describes the basic federal rules under welfare reform. States have leeway to customize rules. Wyoming, for example, has elected to extend the bar on Medicaid participation to immigrants who entered the U.S. before August 1996; immigrants who were already on Medicaid will lose benefits. Texas has announced that it intends to continue to extend the bar for post-enactment immigrants beyond five years, until they become citizens. Other states, such as Washington and Hawaii, have chosen to partially protect immigrants, using state funds to cover some benefits lost under the federal law.

At this time, California is continuing to provide full Medicaid coverage to post-enactment and PRUCOL immigrants. Although California's governor proposed changing the rules to match the basic federal guidelines, the legislature did not enact these changes. Thus, prior state eligibility rules still apply. The governor will probably propose these changes again in 1998, so that the policies may change in the future. It seems likely that the federal government will not match state expenditures for non-emergency care provided to post-enactment immigrants in California, but there are a number of technical issues that still need to be resolved in determining state and federal shares.

Future Scope. The impact of PRWORA will grow over time as the number of post-enactment immigrants accumulates. The QC data for 1994 are useful if we assume that the distribution of immigrant categories in post-enactment years mirrors that in 1994 and that most states adopt policies close to the basic federal guidelines. With this assumption, about 5 percent of Medicaid enrollees would not be eligible for full benefits in their first five years because they are "regular" legal immigrants or are in the "not qualified" group.² Deeming rules will keep many off benefits after that time.

State Estimates. The distribution of immigrants varies widely across the states. California had the most noncitizens, about one quarter (25 percent) of its caseload. Indeed, the 1.3 million immigrants on Medicaid in that state were more than half of the national total of Medicaid immigrants. Table ES-3 summarizes combined data for the ten states with the highest number of immigrants: California, New York, Texas, Florida, Illinois, Massachusetts, Washington, New Jersey, Arizona and Pennsylvania. Immigrants are highly concentrated: these 10 states include 91 percent of the nation's total number of Medicaid immigrants and account for 90 percent of the nation's expenditures for immigrants. In 21 states, less than 1 percent of the Medicaid enrollees were immigrants.

Medicaid Expenditures for Immigrants. Are Medicaid expenditures for immigrants different than for citizens? We address this issue in two parts, first looking at expenditures for legally admitted immigrants, who have been eligible for the full range of Medicaid benefits, then for undocumented aliens, who were eligible only for emergency services.

A general challenge in analyzing Medicaid expenditures per capita is that medical expenditures are unevenly distributed: children have low average expenditures, while the aged and disabled have high expenditures. As shown in Table ES-1, immigrants on Medicaid include a smaller share of children and the disabled than for the overall caseload. After standardizing for differences in caseload composition, legally admitted immigrants had Medicaid expenditures that were quite close to the overall average, 116 percent of the overall expenditure per enrollee, based on the QC data. Legally admitted immigrants were a little more expensive in California (122 percent of the average for California) and a little less expensive in the rest of the nation (90 percent of average).

Emergency Services. The loss of Medicaid coverage does not leave immigrants completely unprotected; they remain eligible for emergency medical services under Medicaid.³ To understand more about emergency benefits, we examined data about the undocumented aliens who were eligible only for emergency coverage. After standardizing for differences in caseload composition, the overall average cost per capita for an undocumented alien was almost exactly the same as for an average beneficiary, 103 percent of the average cost. However, the per person costs were 60 percent of the average in California and 223 percent of the average in the rest of the nation.

A likely reason for these peculiar findings was the limited access of undocumented aliens to emergency Medicaid coverage in most of the nation. Official estimates of the number of undocumented aliens suggest that undocumented aliens in most of the nation were far less likely to enroll in Medicaid than in California (based on Immigration and Naturalization Service 1997). There might have been greater access in California because: (1) undocumented aliens in California could enroll in Medicaid at the welfare office, but received Medicaid cards indicating they had limited benefits; in most other states, undocumented aliens were enrolled only after they had an emergency and (2) the immigrant and health care provider communities may have been more aware of benefits in California. The post-emergency system of enrollment used elsewhere meant that few undocumented aliens were ever granted access to Medicaid. An uninsured alien might have avoided medical care as long as possible, thinking there was no way to pay. Even if the alien received care at an emergency room, the hospital might not bother to enroll the person in Medicaid unless there was a high medical bill. Thus, most states had a pattern in which very few undocumented aliens were ever enrolled in Medicaid and those admitted were quite ill and had very high expenditures.

Another complication regarding emergency care is that most state Medicaid programs are shifting to managed care. There will be tensions between emergency services, which are almost inherently fee-for-service, and the expansions of managed care. Immigrants trying to get access to emergency care under Medicaid may encounter additional problems because of these changes in the health care delivery system.

Implications. A substantial share of Medicaid beneficiaries - about one in thirteen or 2.4 million people - were noncitizen immigrants, in a typical month in 1994. Because the welfare reform legislation does not affect the eligibility of most of those who were present in the U.S. in August 1996, very few of the immigrants now on Medicaid are affected. However, profound changes will occur over time, as the number of more recently arrived immigrants accumulates. If high immigrant states curtail Medicaid eligibility in line with the basic federal policies, the number of people with Medicaid coverage, particularly adults and the elderly, will gradually fall as immigrants drop off the caseload. If high immigrant states instead opt to use state funds to supplement federal Medicaid coverage, they may forestall these problems, but the non-federally matched portion of their Medicaid expenditures will gradually rise. The impact will be much smaller in low immigrant states under either policy.

Immigrants who lose full Medicaid coverage are still eligible for emergency medical services. However, given the historical experience with emergency coverage for undocumented aliens, it is possible that legal immigrants will have great difficulty obtaining access to emergency coverage in the Medicaid program. Administrative barriers, such as the inability to get a Medicaid card, problems obtaining the required documentation, and interactions with Medicaid managed care programs, may make it very hard for immigrants to get Medicaid coverage, even when they are eligible. States could reduce barriers to entry by developing administrative systems to ease access to emergency coverage for immigrants, such as by determining eligibility in advance and issuing Medicaid cards that indicate eligibility only for emergency services. There is little information at this time about how many states have adopted policies like these.

The data presented in this paper are for 1994. It is difficult to predict how these distributions will apply in the future. There have been, and will continue to be, changes in the rate of immigration to the U.S. and in the composition of immigrants.⁴ Further, the rate of naturalization among immigrants has risen sharply, at least partly because of concern about the welfare reform changes. The first issue, changes in the rate and type of immigrants, mean there may be fewer immigrants eligible for Medicaid in the future, if there are fewer total immigrants or if they are less poor. The increase in naturalization might mean that the immigrants who lose Medicaid eligibility could gain it after becoming citizens, increasing Medicaid caseloads again. It is not possible to predict how these factors will balance out.

What will happen to low-income immigrants without Medicaid coverage and what does this mean for the health care system? It seems probable that some will get private insurance coverage, whether they get it through employment or through relatives. The rest will join the ranks of the nation's uninsured. Even before these changes, almost half (43 percent) of the noncitizen immigrants in the U.S. were uninsured, a rate of uninsurance three times the national average (Employee Benefits Research Institute 1996). Many uninsured immigrants will avoid or delay medical care because of financial problems. Others will seek charity or uncompensated care from safety net providers, such as public hospitals, community health centers and such. Health care providers in high immigration areas may find that the loss of Medicaid and increase in demand for uncompensated care create substantial financial hardship.

NOTES

1. In general, the terms "noncitizen" and "immigrant" are used interchangeably in this report, unless otherwise specified. At the broadest definition, immigrants are foreign-born people who reside in the U.S. with the intent to stay, regardless of whether they were legally admitted or came in on an undocumented (illegal) basis. Legally admitted immigrants may eventually become citizens through the naturalization process. There are many categories of immigrants (e.g., refugees, asylees), who are classified based on how they were admitted to the U.S., in accordance with immigration law.

2. Exemptions are available for those with 40 or more quarters of Social Security earnings and for soldiers, veterans or their dependents. However, almost no regular immigrants will meet one of those two exemptions in their first five years.

3. In addition, the Balanced Budget Act of 1977 provided $25 million to be distributed to twelve states to help pay for the costs of emergency medical care provided to undocumented aliens.

4. For example, new rules require that those who sponsor immigrants' entry to the U.S. have incomes of at least 125 percent of poverty, which will restrict the ability of low-income citizens to bring family members into the country.

* The full report is available by contacting the Publications Sales Office

Usage, posting and reprint of materials on the UI web site:

Most publications may be downloaded free of charge from the web site in PDF format. This information may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact paffairs@ui.urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.