United States General Accounting Office GAO Report to Congressional Requesters September 1999 MEDICAID ENROLLMENT Amid Declines, State Efforts to Ensure Coverage After Welfare Reform Vary GAO/HEHS-99-163 United States GAO General Accounting Office Washington, D.C. 20548 Health, Education, and Human Services Division B-281152 September 10, 1999 The Honorable William J. Coyne Ranking Minority Member Subcommittee on Oversight Committee on Ways and Means House of Representatives The Honorable Sander M. Levin Ranking Minority Member Subcommittee on Trade Committee on Ways and Means House of Representatives Medicaid, a joint federal-state program, spent about $160 billion in federal fiscal year 1997 to finance health coverage for more than 40 million low-income individuals, including adults and children in families and aged, blind, and disabled people.1 States administer Medicaid within broad federal guidelines that specify which categories of individuals states must cover and which groups states have the option of covering. Before federal welfare reform, states were required to automatically provide Medicaid coverage to families enrolled in the Aid to Families With Dependent Children (AFDC) cash assistance program. When federal welfare reform was enacted in August 1996, automatic eligibility for Medicaid was uncoupled from eligibility for cash assistance, and states implemented a variety of initiatives intended to move families from welfare to the workforce.2 Some experts were concerned that, despite congressionally enacted protections for continued Medicaid coverage, about a third of the over 40 million low-income people who had been automatically eligible for Medicaid could lose coverage. Of particular concern was the possibility that children might unnecessarily lose coverage because, before welfare reform, more children gained access to Medicaid on the basis of family receipt of cash assistance than via other avenues of eligibility, such as disability or other special medical needs. Moreover, recent reports of welfare and Medicaid enrollment declines have raised questions about the unintended consequences of welfare reform for Medicaid and the viability of the federal and state protections to ensure continued Medicaid eligibility for low-income families and children. 1 Fiscal year 1998 expenditures were $177 billion; data on the number of beneficiaries for 1998 are not yet available. 2 The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (P.L. 104-193). Page 1 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Given your concerns about apparent declines in Medicaid enrollment after welfare reform, you asked us to (1) analyze Medicaid enrollment changes for families and children following welfare reform, as well as associated key federal protections established for Medicaid, and (2) assess states’ welfare-related policies and practices that can influence Medicaid enrollment. In conducting our work, we analyzed enrollment data from the 50 states and from the Health Care Financing Administration (HCFA) for 1995 and 1997.3 To assess the procedures and protections that states have used to enroll Medicaid-eligible individuals since welfare reform, we also contacted 21 states to review state policies and practices that influenced enrollment. We performed our work between July 1998 and July 1999 in accordance with generally accepted government auditing standards. (For more detailed information on our study scope and methodology, see apps. I and II.) Between 1995 and 1997, Medicaid enrollment declined nationwide, but Results in Brief substantially less than welfare participation. Overall, Medicaid enrollment among the nonelderly and nondisabled adults and children declined by about 1.7 million, or 7 percent, compared with a 3.1 million, or 23-percent, decline in welfare participation. Shifts in individual states’ Medicaid enrollment for these adults and children during this period ranged from a 19-percent decline in Wisconsin to a 26-percent increase in Delaware. While most states experienced declines in Medicaid enrollment, enrollment increased in some states, in part as a result of individual state program expansions. On the other hand, Medicaid and welfare enrollment declines have been attributed to strong state economies, low unemployment rates, and new state welfare-to-work initiatives. The smaller declines in Medicaid enrollment may also be due to federal eligibility protections built into welfare reform and ongoing expansions of Medicaid coverage for low-income children that predate welfare reform. One eligibility protection that predates welfare reform—transitional Medicaid assistance—provides an additional year of Medicaid coverage for individuals who lose Medicaid eligibility as a result of employment or increased income. The extent to which transitional Medicaid has affected national enrollment trends, however, is uncertain because of the lack of uniform reporting and tracking of this entitlement. 3 1995 provided a baseline for enrollment before the 1996 enactment of welfare reform, and 1997 was the most current year for which HCFA enrollment data were available when we initiated our work. Page 2 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Our analysis also shows that changes in state-level welfare policies and practices can both positively and negatively influence Medicaid enrollment, as seen in the following examples. • States we contacted are increasingly implementing welfare reform-related policies and programs designed to divert families from enrolling in cash assistance programs. While a diversion strategy such as requiring a job search before providing cash assistance can affect the timing of Medicaid eligibility determinations, such a requirement does not appear to exceed the maximum time allowed by the Medicaid statute. • The length of time that states we contacted provide transitional Medicaid for newly working families varied, ranging from 1 to 3 years. Participation in transitional Medicaid ranged from about 4 percent to 94 percent of those leaving cash assistance in the several states that were able to provide such data. Additionally, adherence to beneficiary income reporting requirements can affect the extent to which families initially receive and then retain transitional Medicaid coverage for the full period for which they may be entitled. States taking advantage of the statutory authority to obtain HCFA waivers of these income reporting requirements reported higher participation rates than states that did not. • Some states have initiated outreach and education campaigns to counter confusion among beneficiaries regarding Medicaid eligibility. In concert with the State Children’s Health Insurance Program (SCHIP) authorized in 1997, some states have simplified Medicaid application and eligibility determination processes to facilitate enrollment. Also, officials in some states have noted increased Medicaid enrollment stemming from eligibility screening for SCHIP. Recognizing that the income reporting requirements can limit beneficiaries’ access to the transitional Medicaid entitlement, HCFA has submitted a legislative proposal to eliminate these requirements for up to 1 year. Our work shows that increased state flexibility to ease reporting requirements could facilitate the transition from welfare to work and make Medicaid more available to eligible individuals. Therefore, we are recommending that the Congress consider allowing states, as a feature of their Medicaid programs, to guarantee a full year of transitional Medicaid coverage to eligible beneficiaries without quarterly income reporting requirements. State options in this area would be similar to the flexibility granted under section 4731 of the Balanced Budget Act of 1997 (BBA), which allowed states to guarantee children a longer period of Medicaid coverage, regardless of changes in a family’s financial status or size. This report also recommends that the Administrator of HCFA (1) analyze states’ Page 3 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 use of transitional Medicaid and (2) provide states with technical assistance regarding best approaches to implementing transitional Medicaid. In commenting on a draft of this report, HCFA concurred with these recommendations. The Personal Responsibility and Work Opportunity Reconciliation Act of Background 1996 changed the relationship between receipt of cash assistance and Medicaid eligibility by delinking the two programs involved, potentially affecting about a third of the Medicaid population. The act replaced AFDC with fixed block grants to the states to provide Temporary Assistance for Needy Families (TANF) and ended the entitlement of families to cash assistance. Under TANF, states have the flexibility to design their own cash assistance programs, which may include developing strategies that may divert potential applicants from cash assistance entirely. In our 1998 report on TANF implementation, we noted that many states had begun implementing diversion strategies, such as providing job search assistance or making one-time lump-sum payments, to divert potential applicants from cash assistance.4 The act also established a 5-year lifetime limit on receipt of TANF benefits.5 In an effort to safeguard access to health insurance for eligible low-income people, the welfare reform law also required that states implement a separate Medicaid eligibility category, which ensured that low-income families meeting a state’s July 16, 1996, AFDC eligibility criteria could qualify for Medicaid without also receiving cash assistance. The welfare reform law also provided states with new choices regarding how to administer Medicaid and determine applicants’ eligibility for coverage.6 Previously, states had been required to use a single state agency to administer both AFDC and Medicaid as well as a single application for use in determining eligibility for both programs. Now states have the option of using separate state agencies, applications, eligibility criteria, 4 TANF became effective on July 1, 1997, but states had the option of implementing their TANF programs as early as October 1996. TANF implementation dates varied among the 21 states we contacted from October 1996 in Connecticut, Florida, Utah, and Wisconsin to January 1998 in California. California’s TANF plan was approved in November 1996, and the state implemented its program in January 1998. For more information on TANF implementation, see Welfare Reform: States Are Restructuring Programs to Reduce Welfare Dependence (GAO/HEHS-98-109, June 17, 1998). 5 For information on the states’ efforts to track the impact of welfare reform on former cash assistance recipients, see Welfare Reform: Information on Former Recipients’ Status (GAO/HEHS-99-48, Apr. 28, 1999). 6 See Medicaid: Early Implications of Welfare Reform for Beneficiaries and States (GAO/HEHS-98-62, Feb. 24, 1998) for additional information on states’ Medicaid-related choices following federal welfare reform. Page 4 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 and application procedures for TANF and Medicaid. While welfare reform required that states use the July 16, 1996, AFDC eligibility income and resource (asset) criteria for determining Medicaid eligibility, states were free to apply different criteria for TANF eligibility, including work and preapplication requirements that had to be satisfied before TANF applications were processed. In addition, the welfare reform law extended the life of the transitional Medicaid assistance program through the year 2001. The transitional Medicaid assistance program, established in 1988 under section 1925 of the Social Security Act, entitles certain families who are losing Medicaid as a result of employment or increased income to an additional year of Medicaid coverage. Families moving from welfare to work are entitled to an initial 6 months of Medicaid coverage without regard to the amount of their earned income as well as an additional 6 months of coverage if family earnings, minus child care costs, do not exceed 185 percent of the federal poverty level.7 The welfare reform law did not, however, change the time limits that states must meet in processing Medicaid applications, nor did it change federal oversight responsibility for Medicaid and cash assistance. States are still required to provide Medicaid applications upon request and to determine applicant eligibility for Medicaid coverage within 45 days of the date of application.8 In addition, the Department of Health and Human Services (HHS) continues to have oversight responsibility for both HCFA—the federal agency that administers Medicaid—and the Administration for Children and Families (ACF), which administered AFDC and now oversees the TANF block grant program. The federal welfare reform law also left unaltered Medicaid coverage for the so-called “expansion population”—pregnant women as well as infants and children under age 19 born after September 30, 1983, whose family income falls below states’ poverty-level standards.9 The Medicaid statute requires that states annually expand Medicaid coverage to children living in low-income families until October 2002, when children through the age 7 In 1999, the federal poverty level for a family of three was $13,880, or about $1,157 per month. 8 States have 90 days to determine eligibility for disability-related coverage. 9 The Omnibus Budget Reconciliation Act of 1989 (P.L. 101-239) required states to provide Medicaid coverage for pregnant women and children up to age 6 with family incomes below 133 percent of the federal poverty level. The act also froze eligibility standards at December 19, 1989, levels for 17 states that had chosen to provide coverage for pregnant women and infants in families with incomes above 133 percent and up to 185 percent of the federal poverty level. Page 5 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 of 18 will be eligible.10 Currently, the law requires that children only up to age 15 be covered; however, over 20 states have accelerated the expansion schedule and are allowing older children to qualify for Medicaid coverage sooner than prescribed by the law. By August 1996 when the federal welfare reform legislation was passed, 16 states, including 6 in our sample, were already covering children up to age 19. 11 One year after passage of welfare reform, as part of the BBA, the Congress established SCHIP, an optional health insurance program for children in families with incomes up to 200 percent of the federal poverty level who do not qualify for Medicaid.12 Beginning in October 1997, the Congress authorized about $40 billion over 10 years in federal matching funds for states’ SCHIP programs to expand health care coverage to uninsured low-income children. States have the choice of (1) expanding Medicaid coverage; (2) establishing a separate, stand-alone health insurance program; or (3) combining these two approaches. Because the federal contribution percentage is higher for SCHIP than for Medicaid, the Congress was concerned that states would have some incentive to enroll Medicaid-eligible children in SCHIP rather than in Medicaid. To ensure that Medicaid-eligible children are not enrolled in SCHIP, HCFA requires that states first screen all SCHIP applicants for Medicaid eligibility. Recently, states have begun implementing welfare reform and SCHIP concurrently, which has potential implications for Medicaid enrollment. From 1995 to 1997, national enrollment in Medicaid among adults and Medicaid Enrollment children declined by about 1.7 million, or 7 percent, ranging from a Decline Has Been decrease of 19 percent in Wisconsin to an increase of 26 percent in Influenced by Various Delaware. In contrast, national welfare participation declined an average of about 23 percent, or 3.1 million recipients, from 1995 to 1997; declines Factors, Including Key ranged from 7 percent in Alaska to nearly 56 percent in Wisconsin.13 (See Federal Protections app. II for an analysis of states’ Medicaid enrollment and welfare 10 Sec. 4601 of the Omnibus Budget Reconciliation Act of 1990 (P.L. 101-508) mandated the annual expansion of coverage for children living in families with incomes below 100 percent of the federal poverty level. 11 In addition to expansions for older children, some states have continued to expand eligibility above the required poverty level; by September 1997, 35 states provided Medicaid coverage for pregnant women and infants with incomes of more than 133 percent of the federal poverty level, and 14 states exceeded the 133-percent level for children up to age 6 as well. 12 SCHIP allows states with Medicaid income levels that already approach or exceed 200 percent of the federal poverty level to expand eligibility up to 50 percentage points above their existing Medicaid eligibility standards. For additional information on SCHIP, see Children’s Health Insurance Program: State Implementation Approaches Are Evolving (GAO/HEHS-99-65, May 14, 1999). 13 Only one state—Hawaii—had an increase in welfare participation during this time. Page 6 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 participation data.) In several states for which data were available, child enrollment for Medicaid showed a much smaller decline than that of adults. Factors that states cited as affecting declines in Medicaid and welfare enrollment—a strong economy, low unemployment rates, and new welfare-to-work initiatives—may have had a more limited effect on Medicaid than on welfare enrollment. In particular, employment in lower-wage positions, many of which do not offer health insurance coverage, is not likely to cause families losing cash assistance to become ineligible for Medicaid. Families may continue to be eligible for Medicaid because of federal requirements to disregard certain types of income in calculating Medicaid eligibility. Additionally, differences between the rates of decline of welfare and Medicaid enrollment may also be due to federal health coverage protections for low-income families and Medicaid eligibility expansions for low-income children. However, data on transitional Medicaid—designed to preserve coverage for families on a temporary basis—are not available nationwide, and the effect of transitional Medicaid on Medicaid enrollment remains uncertain. Welfare Declines Have Nationally, welfare participation declined three times as much as Medicaid Outpaced Medicaid enrollment from 1995 to 1997: 23 percent compared with about 7 percent. Declines As table 1 shows, for our sample of 21 states, welfare participation consistently declined more than Medicaid enrollment. Page 7 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Table 1: Percentage Changes in Medicaid Enrollment and Welfare Percentage change Participation in Our State Sample, in Medicaid Percentage change in 1995-97 enrollment welfare participation Nationala –7.4 –23.4 b Our sample –8.9 –23.4 Individual statesc California –7.6 –13.5 Colorado –9.6 –32.2 Connecticut –1.0 –11.5 Delaware +26.2 –15.8 Florida –13.4 –35.2 Georgia –4.6 –30.4 Idaho –10.8 –49.0 Indiana –9.0 –36.8 Kansas –10.5 –37.0 Kentucky –10.1 –20.3 Maryland –7.2 –31.2 Michigan –8.9 –26.9 Nevada –11.5 –29.3 North Dakota –8.0 –27.0 Ohio –15.9 –23.5 Oklahoma –9.3 –34.0 Oregon –13.0 –42.2 South Carolina +2.4 –32.9 Utah –5.0 –26.2 Vermont +2.6 –16.8 Wisconsin –19.0 –55.6 a We were unable to obtain comparable enrollment data for the District of Columbia, Rhode Island, and West Virginia. b We over-sampled states with Medicaid declines in order to focus our analysis on state policy and practices that can contribute to declines in enrollment. c States had varying amounts of Medicaid enrollment data available for analysis. For a state-by-state discussion of available data, see app. II, table II.1. Source: GAO analysis of the states’ Medicaid monthly enrollment data and of welfare data from HHS’ ACF. On the national level, the declines in welfare participation did not explain the changes in Medicaid enrollment; moreover, the ratios of declines for both programs were not closely related. Differences in state policy and Page 8 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 practices may also help explain some of the variation. For example, South Carolina’s Medicaid enrollment increased 2.4 percent, in part because of state expansions of eligibility for Medicaid and significant outreach efforts; during this same period, welfare participation dropped 32.9 percent. Delaware’s nearly 94-percent increase in adult enrollment occurred after the state implemented a Medicaid waiver expanding eligibility to adults with incomes up to the federal poverty level. During the same period, welfare enrollment dropped 15.8 percent. Ohio and Wisconsin had the largest declines in Medicaid enrollment among the states we contacted—about 16 percent and 19 percent, respectively—but very different declines in welfare. Ohio’s decline in welfare participation was less than twice its decline in Medicaid enrollment, while Wisconsin’s decline in welfare participation was almost three times as great as its Medicaid enrollment decline. Both states reported that an improved economy and successful welfare-to-work strategies might have accounted for some of the declines in Medicaid enrollment. However, they also expressed concern that beneficiaries and state caseworkers were confused about the change in the relationship between welfare and Medicaid. Magnitude of Medicaid Within the overall 7-percent decline in Medicaid enrollment nationally, Declines Varied Across the there was considerable variation among the states. As illustrated by figure States 1, 12 states experienced declines of 10 percent or more, while 4 states saw Medicaid enrollment increase by 5 percent or more. Twelve states had relatively stable enrollment with changes of less than 3 percent. (See app. II for more detail.) Page 9 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Figure 1: Changes in States’ Medicaid Enrollment, 1995-97 Page 10 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Note: We were unable to obtain comparable Medicaid enrollment data for the District of Columbia, Rhode Island, and West Virginia. Source: GAO analysis of the states’ monthly Medicaid enrollment data. These data indicate the change in Medicaid enrollment at one given point in time. In commenting on this report, officials in five states—Connecticut, Florida, Indiana, Michigan, and Oklahoma—reported that since 1997, Medicaid enrollment has stabilized or begun to increase. For example, Connecticut officials noted that in contrast to their 1-percent decline between 1995 and 1997, the state has experienced a net increase of nearly 9,000 children enrolled in Medicaid between June 1998 and January 1999. While Florida officials reported similar increases in enrollment among children, officials in Indiana indicated that adult enrollment has also increased. According to Michigan officials, the state’s enrollment has stabilized as a result of efforts to identify and reenroll eligible individuals who lost Medicaid coverage between 1995 and 1997. Oklahoma officials reported that since December 1997, when the state expanded Medicaid eligibility for pregnant women, infants, and children, enrollment has risen to approximately 224,000—a 14-percent increase since federal fiscal year 1995, which contrasts with the 9-percent decline reflected by our analysis. Additionally, Maryland officials commented that the state has made efforts to ensure enrollment in Medicaid for eligible individuals after reviewing its post-welfare-reform Medicaid and SCHIP policies and practices. In particular, Maryland has implemented such improvements as enhancing automated systems, training caseworkers statewide, and conducting outreach to the general public. For Families Leaving Cash While a strong economy helps explain the decreases in welfare Assistance, Medicaid participation, it may have had a more limited effect on Medicaid Eligibility Remains Likely enrollment. Officials in most of the 21 states we contacted attributed their declines in welfare and Medicaid enrollment to successful welfare-to-work Despite Strong State strategies and strong state economies. These officials said that low Economies unemployment rates have aided their efforts to help welfare recipients quickly find jobs. Nationally, the unemployment rate averaged about 4.9 percent in 1997, down from about 5.6 percent in 1995, and almost all the states in our sample experienced declines in their unemployment rates between 1995 and 1997. In comparison, in 1994 when welfare participation was at its peak (14.2 million), the unemployment rate was 6.1 percent. The federal minimum wage increased from $4.25 in 1995 to $4.75 in 1997.14 14 The national unemployment rate for the first 4 months of 1999 was 4.5 percent, with a $5.15 minimum wage. Page 11 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Despite strong state economies, families may continue to be eligible for Medicaid while working because of federal requirements and state flexibility to disregard certain types of income in calculating Medicaid eligibility. For example, states are required to exclude or “disregard” the first $30 of monthly family earnings for 1 year as well as one-third of the remaining income for the first 4 months of employment.15 Additional income disregards include $90 per month for work-related expenses, such as clothing and transportation, and $175 to $200 per child, based on age, for monthly child care expenses. Eight of our sample states—California, Connecticut, Delaware, Maryland, Ohio, South Carolina, Utah, and Vermont—had federal waivers in place or made changes to state Medicaid policy that allowed them to continue to use more generous income disregards in determining eligibility for Medicaid.16 The option available under federal welfare reform to continue waiver criteria, as well as the income disregards noted above, enables many former cash recipients working in minimum wage jobs to remain eligible for Medicaid, at least for the first 4 months of employment.17 Data on former cash recipients in California, Indiana, Maryland, and Wisconsin showed former welfare recipients generally held low-wage jobs—such as in retail stores, hotels, restaurants, and health care establishments—and worked less than full-time. For example, out of nearly 1,600 former welfare recipients surveyed in Indiana, 514 current and former cash recipients were working in low-wage jobs. In addition, 43 percent of the heads of households worked fewer than 32 hours per week and did not have health insurance. Over half of those surveyed had been offered employer-provided health insurance; of these, about 60 percent had declined coverage because it was too expensive or for other reasons. Only 8 percent of those declining coverage were enrolled in Medicaid.18 Although state monthly income eligibility standards varied greatly—from 15 After 4 months, states must still disregard the first $30 of earnings for 1-year but may alter any other income disregards. 16 Before the 1996 enactment of the welfare reform law, many states had received waivers from the federal rules applicable to the AFDC program. These waivers allowed states to experiment with various welfare polices, including the use of more generous income disregards for working families. Federal welfare law allowed the states to continue applying welfare waiver provisions to Medicaid eligibility when such provisions involved income and resource methodologies and certain other criteria involving family composition. However, provisions such as time limits, sanctions (withholding coverage), and more restrictive eligibility criteria cannot now be continued beyond the expiration of the waiver. Under TANF, states determine their own eligibility criteria, including any amounts of income to be disregarded. 17 See app. I for additional information on our methodology for this analysis. 18 See Abt Associates, Inc., and The Urban Institute, The Indiana Welfare Reform Evaluation: Program Implementation and Economic Impacts After Two Years (Nov. 1998). Page 12 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 $200 in South Carolina to $663 in California—it appears that in every state in our sample except Colorado, heads of three-member households could have worked full-time at the 1997 minimum wage and remained Medicaid-eligible. In Colorado, heads of three-member households could have worked up to 36 hours per week at the 1997 minimum wage and retained Medicaid eligibility. Additionally, a 1999 analysis by the Center on Budget and Policy Priorities showed that in 6 of the 21 states in our sample—California, Connecticut, Delaware, Ohio, Oregon, and Vermont—heads of three-member households could have worked full-time, earning as much as $5.15 per hour, and still have continued their Medicaid eligibility.19 Among the remaining 15 states, the number of hours per week that heads of three-member households could have worked and still have continued to be income-eligible for Medicaid ranged from 17 in Indiana to 36 in Florida.20 Federal Protections for Federal health coverage protections for adults in low-income families, Low-Income Families May protections and expansions of coverage for low-income children, and the Have Tempered availability of transitional Medicaid coverage for families moving from welfare to work may have also prevented greater declines in Medicaid Welfare-Related Medicaid enrollment. These protections, however, often appeared to have different Declines outcomes for children and adults. Of the 11 states in our sample that were able to readily provide separate Medicaid enrollment data for children and for adults, declines in child enrollment were significantly less than among adults in 10 states, as shown in table 2. Delaware was an exception to this trend; its nearly 94-percent increase in adult enrollment occurred after the state implemented a Medicaid waiver expanding eligibility to adults with incomes up to the federal poverty level. 19 Center on Budget and Policy Priorities, Employed But Not Insured (Washington, D.C.: Mar. 1, 1999). 20 Our previous work found that in four states in our sample—Indiana, Oklahoma, South Carolina, and Wisconsin—former cash assistance recipients were likely to earn an average hourly wage above the federal minimum wage. See GAO/HEHS-99-48, Apr. 28, 1999. Page 13 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Table 2: Percentage Changes in Child and Adult Medicaid Enrollment in Our Percentage Percentage State Sample, 1995-97 change in total Percentage Percentage decline in Medicaid change for change for welfare Statea enrollment childrenb adults participation California –7.6 –6.5 –10.5 –13.5 Colorado –9.6 –5.0 –21.9 –32.2 Connecticut –1.0 –0.2 –2.8 –11.5 Delawarec +26.2 +3.7 +93.8d –15.8 Georgia –4.6 +0.1 –18.7 –30.4 Idaho –10.8 –6.5 –24.7 –49.0 Kentucky –10.1 –7.2 –17.2 –20.3 Maryland –7.2 –4.6 –14.2 –31.2 Nevada –11.5 –5.2 –30.2 –29.3 North Dakota –8.0 –5.0 –15.8 –27.0 Utah –5.0 –4.8 –5.3 –26.2 a We were unable to readily obtain separate enrollment data for children and adults from Florida, Indiana, Kansas, Michigan, Ohio, Oklahoma, Oregon, South Carolina, Vermont, and Wisconsin. b The upper age limit states use to define children varies from 18 to 21. c In calculating the percentage of change for children and adults in Delaware, we excluded less than 5 percent of total adult/child enrollment because state officials were unable to determine in which of the two categories the data belonged. d In 1995, Delaware expanded adult eligibility to those with incomes of up to 100 percent of the federal poverty level. The state reported that this expansion caused 16,000 more adults to become eligible for Medicaid. Source: GAO analysis of states’ Medicaid monthly enrollment data. Protections for Adults in The welfare reform act provided a mandatory Medicaid coverage Low-Income Families protection for adults in low-income families and also allowed states to apply additional protections on a voluntary basis. In the absence of these protections, even larger declines in adult Medicaid enrollment could have resulted as welfare recipients moved into the workforce. As part of welfare reform, section 1931 was added to the Social Security Act. Section 1931 established a separate Medicaid eligibility category to protect adults—primarily women and their older teenaged children—who were previously eligible under their states’ AFDC programs. Specifically, the law requires that states use standards no more stringent than the AFDC standards in effect on July 16, 1996, as the criteria for determining Medicaid eligibility. All of the 21 states we contacted had either Page 14 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 established a section 1931 eligibility category or had submitted a state plan amendment to do so.21 Welfare reform also included several exceptions to the July 16, 1996, standards—two that allowed states to voluntarily expand Medicaid eligibility and one that allowed states to impose more restrictive standards.22 Of the 21 states in our sample, 10—California, Colorado, Delaware, Florida, Kansas, Nevada, North Dakota, Ohio, South Carolina, and Vermont—expanded Medicaid eligibility by increasing their resource and income standards or liberalizing their determination methodologies. None of the 21 states applied more restrictive eligibility policies to Medicaid.23 Other health coverage protections for adults include provisions of the Medicaid statute that predated federal welfare reform and allow states to use higher income standards and more liberal methodologies for determining Medicaid eligibility than used for determining applicant eligibility for cash assistance. For example, section 1902(r)(2) of the Social Security Act permits states to reduce (or even eliminate) income and resource standards for many categories of prospective Medicaid beneficiaries, including, for example, pregnant women, children, and certain blind or disabled people.24 States such as California, Colorado, Connecticut, Delaware, Kansas, and Vermont expanded Medicaid eligibility for adults in low-income families by disregarding more than the standard amounts for resources.25 California, Colorado, Connecticut, and Kansas allow families to have more than the former AFDC standard amount 21 In commenting on a draft copy of this report, Maryland officials noted that they have submitted a state plan amendment and taken steps to delink their cash assistance and Medicaid programs. 22 Under sec. 1931(b)(2), states were permitted to expand Medicaid eligibility by (1) using less restrictive methodologies for calculating family income and resources than used on July 16, 1996, or (2) increasing their AFDC July 16, 1996, income and resource standards by as much as the year’s consumer price index. States were also allowed to restrict eligibility by lowering their AFDC July 16, 1996, income standards, but not below May 1, 1988, levels. 23 States could not impose additional restrictions on Medicaid eligibility without jeopardizing access to SCHIP funds. In particular, the BBA stipulated that states participating in SCHIP could not use more restrictive income or resource standards than the standards used for Medicaid on June 1, 1997. 24 Before welfare reform, when cash assistance and Medicaid eligibility were still linked, states disregarded additional amounts of earned income as an incentive to encourage cash assistance recipients to work. 25 Under the former AFDC program, cash recipients were limited to $1,000 in total resources (assets). However, in calculating family assets, states were required to disregard (not include in the calculation) certain assets, such as a personal residence and its contents, burial plots, and education grants and scholarships, and to discount the value of other assets. For example, in calculating total resources, states were required to discount vehicle equity by $1,500 and prepaid funeral or burial arrangements by $1,500 per person. Page 15 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 of $1,000 in liquid assets, while both Delaware and Vermont have eliminated asset tests for families applying only for Medicaid coverage. Protections and Mandated Smaller declines in Medicaid enrollment for children can also be attributed Program Expansions for to program expansions for children mandated by the Medicaid statute. As Children in Low-Income table 3 shows, as of September 30, 1997, 11 of the 21 states we contacted Families provided Medicaid coverage to children older than 14 years of age, with family incomes at or above the federal poverty level. Such coverage will soon become a legal requirement for all low-income children up to age 19.26 Table 3: Medicaid Coverage of Pregnant Women, Infants, and Poverty level percentages for Children in Our State Sample as of Medicaid coverage Upper age September 30, 1997 Pregnant Children limit for women under Children 6 defining and infants age 6 and older “child” Federal minimum mandatory coverage 133a 133 100 14 State California 200 133 100 14 Colorado 133 133 100 14 Connecticut 185 185 185 16b Delaware 185 133 100 18 Florida 185 133 100 14 Georgia 185 133 100 19 Idaho 133 133 100 14 Indiana 150 133 100 18 Kansas 150 133 100 17 Kentucky 185 133 100 14 Maryland 185 185 185 14 c Michigan 185 150 150 16 Nevada 133 133 100 14 North Dakota 133 133 100 18 Ohio 133 133 100 14 c Oklahoma 150 133 100 14 Oregon 133 133 100 19 South Carolina 185 150 150 18 (continued) 26 The Omnibus Budget Reconciliation Act of 1989 requires states to annually phase in Medicaid eligibility to children born after September 30, 1983, until all children up to age 19 in families with incomes below 100 percent of the federal poverty level are covered. By October 1, 1999, all children up to age 16 will be covered. Page 16 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Poverty level percentages for Medicaid coverage Upper age Pregnant Children limit for women under Children 6 defining and infants age 6 and older “child” Utah 133 133 100 18 d Vermont 200/225 225 225 17 Wisconsin 185 185 100 14 a The minimum mandatory income requirement for pregnant women and infants may be higher than 133 percent of the federal poverty level for states that, as of December 19, 1989, had opted to set eligibility for this category between 133 percent and 185 percent of the federal poverty level. b Beginning January 1, 1998, Connecticut expanded coverage to include children up to age 19. c After the BBA of 1997, Michigan expanded Medicaid eligibility to include children up to age 19 in families with incomes below 200 percent of the federal poverty level, and Oklahoma expanded program coverage to include children in families with incomes of up to 185 percent of the federal poverty level. d Vermont provides Medicaid coverage for pregnant women with family incomes up to 200 percent of the federal poverty level and coverage for infants with family incomes of up to 225 percent of the poverty level. Source: National Governors’ Association. Since 1990, children have been able to qualify for Medicaid when living in families with substantially higher incomes than those of cash assistance recipients. States’ cash assistance programs typically limited eligibility to families with incomes well below the federal poverty level—ranging from a high of about 81 percent of the federal poverty level in Connecticut to a low of 15 percent in Alabama. In 1997, five of the states in our study—Connecticut, Maryland, Michigan, South Carolina, and Vermont—covered children between ages 6 and 14 in families with incomes of 150 percent or more of the federal poverty level. Four of these five states—Connecticut, Michigan, South Carolina, and Vermont—also covered older children in families with incomes of 150 percent or more of the poverty level. The remaining states are still phasing in coverage for older children or are using SCHIP funds to accelerate coverage.27 Transitional Medicaid In addition to the protections in the welfare law and state efforts to Assistance Designed to Protect expand coverage, transitional Medicaid assistance is another avenue for Families Moving From Welfare preventing the immediate loss of Medicaid coverage for families who to Work transition from welfare to work. Section 1925 of the Social Security Act requires that states provide transitional Medicaid coverage to families 27 Florida and Wisconsin are using SCHIP funds to extend coverage to older children. Page 17 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 losing Medicaid eligibility as a result of employment or other financial circumstances.28 Under this provision, states are specifically required to provide two sequential 6-month periods of transitional Medicaid to families when certain conditions are met.29 Nationwide, the extent to which eligible families obtain and keep coverage under transitional Medicaid is unknown. HCFA does not require state reporting or otherwise monitor state compliance with the requirement to provide this program benefit. Further, states do not separately identify families receiving transitional Medicaid when reporting enrollment data to HCFA. HCFA proposed a regulation for transitional Medicaid on December 14, 1993, but did not finalize it because of staffing constraints.30 The proposed regulation essentially reiterated the statute but did not provide states with additional structure or guidance regarding implementation or ways to consistently monitor that beneficiaries receive and retain coverage under this benefit. On March 22, 1999, however, the agency sent guidance to TANF administrators and state Medicaid and SCHIP directors on expanding health coverage to families making the transition from cash assistance to work. While the March 1999 guidance provides that states must not deny or terminate Medicaid eligibility unless all possible avenues to such eligibility have been exhausted, HCFA does not address transitional Medicaid in any detail. Changes in state welfare policies and practices have had both positive and Changes in State negative influences on Medicaid enrollment. We identified four Welfare Policies and approaches of states’ welfare reform programs that may influence Practices Have Had Medicaid enrollment: (1) diversionary programs, which are intended to help families avoid the need to enroll in TANF; (2) eligibility policies and Mixed Influences on procedures, which states use to determine who is qualified for coverage; Medicaid Enrollment (3) transitional Medicaid assistance, which helps families moving from welfare to work; and (4) education and outreach efforts, which are aimed 28 Other circumstances include increased hours of work or changes in income disregards, which states can choose to disallow after 4 months of employment. Families who lose Medicaid eligibility on the basis of the former AFDC standards because of increased child or spousal support are entitled to 4 months of transitional coverage. 29 For instance, in order to qualify for transitional Medicaid, a family must have received Medicaid under the former AFDC standards in 3 of the 6 months immediately before becoming ineligible as a result of increased income. No limit on income is imposed during the initial 6-month period of transitional Medicaid. During the second 6-month period, however, a family’s gross monthly earnings, less child care expenses, cannot exceed 185 percent of the federal poverty level. 30 58 Fed. Reg. 65,312. Currently, HCFA officials are working on an updated version of the proposal, linking transitional Medicaid to enrollment in the new section 1931 Medicaid eligibility category instead of receipt of AFDC. The states did not strongly oppose the proposed regulation, and HCFA received only six comments on it. Page 18 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 at minimizing confusion about Medicaid eligibility following welfare reform. These program approaches vary by state and can affect Medicaid enrollment levels. Finally, our contacts with states showed that SCHIP outreach efforts have had a positive impact on Medicaid enrollment, particularly for children. States’ Welfare Diversion States are increasingly implementing policies and programs—such as Policies Can Influence mandatory, up-front job searches and offers of a one-time lump-sum Medicaid Enrollment payment—that are designed to divert families from enrolling in welfare. As table 4 shows, 18 of the 21 states in our sample (1) require that applicants search for employment before obtaining welfare assistance, (2) offer welfare-eligible applicants one-time payments in lieu of ongoing cash assistance, or (3) both. Although state officials told us that their diversion policies do not apply to Medicaid applicants, the imposition of these TANF requirements may confuse Medicaid applicants about eligibility requirements or dissuade them from completing a separate Medicaid application. In the states we contacted, up-front job searches may cause more confusion for Medicaid applicants than lump-sum payments, which, in any event, are not often chosen by beneficiaries. Page 19 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Table 4: Comparison of 21 States’ Welfare Diversion Policies as of Up-front job search Lump-sum payment April 1999 State required offered Californiaa X Coloradoa X Connecticut X Delawareb X Florida X Georgia X Idaho X X Indianab X X Kansas X Kentuckyc X a Maryland X X Michigan Nevadab X X North Dakota Ohioa X X Oklahoma X Oregon X South Carolina X Utah X Vermontd Wisconsine X Total 11 12 a Lump-sum payments and job search requirements may differ by county in these states. b States have not yet fully implemented mandatory job search or lump-sum payment programs. c Although Kentucky requires cash assistance applicants to join a job registry, it does not deny cash benefits to families that do not comply. As a result, we did not consider Kentucky to have a mandatory up-front job search requirement. d Vermont requires principal wage-earners in two-parent families—10 percent of the state’s June 1999 TANF caseload—to register with the state’s Department of Employment and Training and begin job search as a condition for cash assistance. e Although Wisconsin has a job access loan program for its cash assistance clients, we did not consider it a diversion strategy; unlike lump-sum payments offered in other states, clients must repay job access loans in cash or in-kind. Mandatory Up-Front Job To encourage work over welfare, 11 of the states in our sample have Search Policies established mandatory, up-front job search requirements that families must satisfy to be eligible for cash assistance. States’ requirements vary Page 20 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 greatly—from joining a job registry, as in Utah and Wisconsin, to spending time (ranging from 2 weeks in Maryland to 45 days in Oregon) pursuing various state-provided job leads. In states that have combined welfare and Medicaid applications—such as Maryland, Oklahoma, Oregon, and South Carolina—mandatory job search policies can delay determination of Medicaid eligibility until job search requirements are satisfied. For example, South Carolina officials told us that they hold combined applications until the job search requirements—10 verified employer contacts—are satisfied. After 30 days, if a job search is not completed, the welfare portion of the application is denied while the Medicaid portion is forwarded to a separate unit for an eligibility determination. Approved Medicaid applications are retroactive to the initial date of application. Maryland officials similarly explained that they have a 14-day hold on cash assistance applications; after that period, the state will determine family eligibility for cash assistance and Medicaid.31 Officials in the other two states did not indicate that the Medicaid portions of the combined applications are held pending a job search; however, we did not independently verify this. Lump-Sum Payments Officials in nine states reported offering welfare applicants one-time lump-sum diversion assistance, and officials in three other states indicated that they will soon implement such programs statewide.32 Of the nine states with operational programs, Utah’s and Maryland’s programs appear somewhat more active than those in the other seven states, where relatively few families have accepted lump-sum payments. According to Utah officials, between 190 and 200 families applying for cash assistance each month—less than 1 percent of the state’s 1997 average monthly welfare participation—accept lump-sum diversion payments in lieu of ongoing assistance. Utah began offering diversion assistance statewide in July 1996 under a welfare waiver that was approved before the federal reform law was enacted. State officials explained that families accepting diversion payments must be eligible for ongoing cash assistance; furthermore, the state enrolls these families in Medicaid and offers them child care and job placement assistance. Families can receive the 31 Since states are required to determine applicant eligibility for Medicaid coverage within 45 days of the date of application (90 days when a disability determination is involved), states do not appear to be exceeding the maximum amount of time allowed by the Medicaid statute. 32 For example, Delaware plans to implement its lump-sum payment program in October 1999. Indiana is piloting both a lump-sum payment program and up-front job search requirements in several of its counties. Indiana officials indicated that if the pilots were successful, they would be implemented in additional counties during the year. Nevada officials said that their legislature has approved a lump-sum payment program, and state officials are considering the software and eligibility system changes needed before implementing the program. Page 21 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 lump-sum equivalent of 3 months of cash assistance—$1,353 for a family of three—and 3 months of Medicaid coverage. Utah allocates its diversion payments over a 3-month period so that the payment does not make recipients automatically ineligible for Medicaid.33 At the end of the 3-month period, eligibility workers determine whether the families are eligible for any additional months of Medicaid coverage. In Maryland, as is the case in California, Colorado, and Ohio, counties have broad authority to implement state programs on the basis of their own priorities. Thus, welfare avoidance grants are optional benefits that caseworkers may offer welfare-eligible families. Maryland welfare officials told us that about 1,600 welfare avoidance grants—for items such as car repair and dental services—have been awarded statewide over the first 2 years of welfare reform. Baltimore city welfare officials, who manage over 50 percent of the state’s welfare caseload, have defined welfare avoidance grants as one-time payments, the equivalent of up to 3 months of cash assistance (ranging from $1,050 to $1,197 for a three-member household) that caseworkers may offer welfare-eligible families so that the head of a household can continue working or accept a bona fide job offer. However, Baltimore city officials said that because very few of their cash assistance applicants have jobs or genuine job offers, very few have met the local criteria for caseworkers to offer these welfare avoidance grants. In fact, few welfare avoidance grants have been given to families in Maryland’s large urban areas. One point of difference between Baltimore city and Maryland’s counties has to do with car repairs. Baltimore city officials do not consider the need for car repairs as a valid reason for offering avoidance grants because the area has a public transportation system, while other areas in the state do include car repair as an acceptable need. Officials in the remaining seven states attributed low participation in their voluntary diversion programs both to the small amounts of money that families are offered and to other benefits they might forgo in accepting the lump-sum payment. For example, only 12 families in Florida have accepted lump-sum payments since the state implemented the program in November 1997. Florida officials hypothesized that more families have not found the program an acceptable alternative to cash assistance because the payment is small in comparison with the benefits they could receive as cash assistance recipients. A family of three, for instance, is limited to a single payment of $606—the equivalent of 2 months of cash assistance. To receive that payment, the family must prove eligibility by completing the standard welfare application and, possibly, forgo food stamps and 33 Under the AFDC program, qualifying families had generally been limited to $1,000 in liquid assets. Page 22 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Medicaid coverage if the payment raises family resources above the state’s $2,000 cash limit; the family must also agree not to apply for more cash assistance for 3 months. States’ Eligibility States must annually redetermine whether individuals remain eligible for Redetermination Processes Medicaid. As part of the redetermination process, eligibility workers verify Have Added Complexities that family incomes are still within state standards and that families continue to meet any other criteria, such as family composition and for Workers resource limits, applicable to their particular Medicaid eligibility category. Welfare reform and state welfare-to-work strategies have introduced added complexities to the Medicaid redetermination process that appear to have affected workers and, as a result, have the potential to affect beneficiaries. State and local welfare officials reported three ways in which welfare reform has made the redetermination process more burdensome for eligibility workers. First, eligibility workloads per worker have increased in three of the four states we visited—California, Florida, and Maryland—since welfare reform.34 Second, in those states in which TANF and Medicaid redeterminations are still linked, workers often reported added complexities, such as having to monitor beneficiary compliance with state job search, work, and vocational training requirements. Third, because families can now apply for Medicaid separately from cash assistance, workers need to become more familiar with Medicaid eligibility rules, since many states’ eligibility systems are not fully automated. Officials in California, Colorado, Florida, Maryland, and Ohio told us that they are seeing an increase in the number of Medicaid-only cases relative to the number of cash assistance cases. Beneficiary advocates we spoke with in Florida, Maryland, and California indicated that the added pressures on eligibility workers can strain worker-beneficiary relations and, in some cases, make communication between the parties so difficult that eligible families do not get the information they need to apply for or retain Medicaid coverage. The growing number of Medicaid-only cases concerns eligibility workers, who previously handled very few of these cases and consider Medicaid too 34 Although Medicaid enrollment had declined in these states, it does not appear that the declines occurred in the major urban areas that we visited. For example, workers in Los Angeles County attributed their increased per-worker caseload to the rising numbers of mixed-status households, stating that a single family may have as many as four separate cases representing different categories of Medicaid eligibility, such as those based on citizenship status (citizens and noncitizens), income, and medical need. Florida workers attributed their increased caseloads to staff reductions and turnover. Page 23 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 complex given its many eligibility categories for differing income, resource, and family composition criteria. The addition of the section 1931 eligibility category has added to this problem. Officials in the 21 states we contacted reported numbers of welfare-related Medicaid eligibility categories that ranged from almost 30 to over 100. According to state officials and workers, the proliferation of eligibility categories is challenging for workers in most states and is particularly troublesome for workers in states with computer systems that have not kept up with welfare policy changes. For example, we were told that workers in Florida must either manually determine Medicaid eligibility or understand the policies well enough to verify the accuracy of the state’s computerized eligibility determinations. California officials told us that workers must manually determine Medicaid eligibility for the section 1931 eligibility category. They also said that most of California’s programming expertise has been devoted to ensuring that more vital state systems are year 2000 compliant. Wide Variation in Although the Medicaid statute entitles families moving from welfare to Beneficiary Access to work to as much as 12 months of transitional Medicaid coverage, the Transitional Medicaid extent to which families receive the benefit and the length of coverage vary considerably by state. Among the states with data that we contacted, Exists Across States transitional Medicaid participation rates ranged from about 4 percent of the families losing cash assistance in Idaho to 94 percent of such cases in Connecticut. Within our 21-state sample, 6 states—California, Connecticut, Delaware, South Carolina, Vermont, and Utah—have pre-welfare-reform waivers to provide 24 months or more of coverage.35 Officials from some states identified several barriers to full beneficiary use of transitional Medicaid benefits, such as periodic income reporting requirements for beneficiaries and a lack of program knowledge among eligibility workers. Several states have found ways to overcome these barriers and make enhanced use of the benefits afforded by transitional Medicaid. Moreover, HCFA has recommended via a legislative proposal that beneficiary income reporting requirements be eliminated from transitional Medicaid. Transitional Medicaid Use Receipt of transitional Medicaid varied considerably among the states we Varies by State contacted, and we were unable to obtain consistent data on program 35 California has a waiver to provide 12 months of transitional Medicaid to people losing eligibility because of marriage or reunification of spouses. Additionally, California provides 12 months of state-funded coverage to adults 19 years and older who have exhausted the 12 months of federal/state-funded coverage. Page 24 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 participation for all 21 states in our sample.36 Furthermore, among the 16 states that could provide us information on transitional Medicaid, there was little consistency in tracking and interpreting data on program participation. For example, an Idaho survey of 14,772 cash assistance cases closed during state fiscal year 1998 showed that 636 families (4 percent) received transitional Medicaid. Idaho officials said they were not alarmed by this low participation rate because, in their estimation, most of the families losing cash assistance were still enrolled in Medicaid either under the new section 1931 eligibility category for low-income families or as children in the state’s Medicaid expansion program.37 A Connecticut survey of the 2,190 families leaving cash assistance and scheduled for exit interviews in January 1998 showed that 2,050 families (94 percent) received transitional Medicaid. Maryland officials reported that in federal fiscal year 1998, 7,206 individuals received transitional Medicaid and in 1997, 7,227 did so; these figures represent about 21 percent and 18 percent of those losing cash assistance in federal fiscal years 1998 and 1997, respectively.38 Although officials in Delaware and Vermont did not provide specific information on transitional Medicaid participation rates, they surmised that most families that lose cash assistance in their states receive the additional months of coverage. Both states received waivers to provide more than 12 months of transitional Medicaid even before federal welfare reform. Delaware provides 24 months of transitional Medicaid coverage, and the state’s eligibility determination system showed that 80 percent of the families enrolled in transitional Medicaid kept coverage for the full 24-month period. Vermont officials similarly believe that participation in their state’s 36-month transitional Medicaid program is high. In addition to an almost 3-percent increase in adult and child Medicaid enrollment between 1995 and 1997, Vermont officials estimated that about 18 percent of the state’s population received some form of publicly subsidized health insurance coverage. Income Reporting HCFA and state officials noted that quarterly beneficiary income reporting Requirements Can Limit requirements can pose barriers to family receipt of transitional Medicaid Transitional Medicaid benefits. Transitional Medicaid entitles certain families who are losing Participation Medicaid as a result of employment or increased income to an additional 36 Lack of data has been a consistent problem in understanding the availability and use of transitional Medicaid. See Welfare to Work: Implementation and Evaluation of Transitional Benefits Need HHS Action (GAO/HRD-92-118, Sept. 29, 1992). 37 As shown in table 1, between 1995 and 1997, Idaho’s Medicaid enrollment decline was 10.8 percent, compared with 49.0 percent for welfare. 38 According to data from ACF, 61,096 individuals lost cash assistance in Maryland between January 1997 and September 1998. Page 25 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 year of Medicaid coverage. Reporting requirements can pose barriers for families leaving cash assistance at two points: (1) entering transitional Medicaid and (2) maintaining this entitlement for the full period of eligibility. In the first case, the failure to notify eligibility workers of employment can prevent families from being enrolled in transitional Medicaid. Theoretically, a head of household would report increased earnings, be removed from cash assistance, and be placed on transitional Medicaid. However, many heads of households do not notify their eligibility workers that they have obtained employment; thus, once disqualified from cash assistance, these heads of households are not automatically rolled over into transitional Medicaid eligibility. Thus, the eligible family never receives transitional Medicaid. In the second case, families participating in transitional Medicaid can have their benefits terminated if they fail to meet statutory reporting requirements established under section 1925 of the Social Security Act. Although the Medicaid statute entitles families to 12 months of transitional Medicaid assistance—in two 6-month segments—each 6-month period of coverage has its own eligibility criteria and income reporting requirements. In all but 3 of the 21 states we contacted, beneficiaries must comply with the following statutory requirements to obtain and maintain a full year of transitional Medicaid coverage.39 • To receive the first 6 months of transitional Medicaid, families must notify the state—typically through the family’s eligibility worker—of their employment and income status. Although there is no income eligibility limit during this period, families must also submit an income report to the state by the 21st day of the 4th month of transitional Medicaid coverage. • To receive the second 6 months of transitional Medicaid coverage, family income minus child care expenses may not exceed 185 percent of the federal poverty level. Families also must submit quarterly income reports by the 21st day of the 1st and 4th months of the second 6-month period. During this period, families may also be required to pay premiums for Medicaid coverage, and the state can reduce the level of Medicaid benefits and services to which they are entitled. 39 Before welfare reform, Connecticut, Delaware, Maryland, and Oregon obtained welfare and Medicaid waivers to eliminate the requirement for quarterly income reporting, thus ensuring an uninterrupted period of transitional Medicaid coverage. The BBA also permits states, as part of their HCFA-approved state Medicaid or SCHIP plan, to guarantee 12 months of continuous Medicaid/SCHIP eligibility for low-income children, without additional family income reporting requirements. A Connecticut official informed us that the state’s waiver of transitional Medicaid reporting requirements will expire in October 2001, and, unless the state is granted an extension of the waiver, the state will have to comply with what it terms a burdensome requirement. Page 26 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 In both instances, failure to report beneficiary income status can result in the termination of transitional Medicaid benefits, unless the family can show good cause for its failure to report on a timely basis. Our review of states showed that beneficiary income reporting requirements can affect whether families receive transitional Medicaid coverage for the full period for which they may be entitled. Some state officials said that although eligibility workers explain the availability of and conditions for receiving transitional Medicaid and provide new beneficiaries with program information, few families comply with the reporting requirements, and many do not respond to termination notices alerting them to loss of coverage. For example, Colorado, Florida, and Oklahoma officials told us that families typically receive only 6 months of transitional Medicaid, generally because of families’ failure to submit the required quarterly income reports. A study of Maryland cash assistance cases closed between October 1996 and September 1997 showed that 19 percent of the cases were closed because of increased income, making the families eligible for transitional Medicaid. Over 50 percent of the closed cases were coded by eligibility workers as administrative closures—for example, closures resulting from failure to submit required reports or to complete the redetermination process. Wisconsin beneficiary advocates became similarly concerned that the state’s automated eligibility determination system was not appropriately shifting closed cash assistance cases into transitional Medicaid following increased complaints from beneficiaries that their cases had been improperly terminated. In previous work, we noted that 14 states did not have policies for informing families about transitional Medicaid at the time of either application for or redetermination of cash assistance.40 Advocates also noted that state cash assistance termination notices can be difficult to understand, and beneficiaries may fail to see how such notices affect their Medicaid eligibility. In commenting on a draft copy of this report, Maryland and Wisconsin officials informed us that they had begun taking steps to reduce the number of administrative closures of cash assistance cases, including creating outreach posters and flyers and carrying out mass mailings to alert beneficiaries of the importance of reporting earnings information. In addition, Wisconsin reported having simplified the text of its cash assistance termination notices and has begun a longer-term effort to overhaul all of its system-generated notices. 40 See GAO/HRD-92-118, Sept. 29, 1992. Page 27 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Our contacts with the states indicated that while many states expect eligibility workers to provide beneficiaries with information on transitional Medicaid, only nine states—California, Georgia, Florida, Indiana, Maryland, Nevada, South Carolina, Vermont, and Wisconsin—reported having developed specific materials in easy-to-understand language for workers and beneficiaries. Four of these nine states—Georgia, Florida, Maryland, and South Carolina—use consistent materials developed by the Southern Institute on Children and Families, while the others developed worker training or educational materials in response to perceived local needs.41 Several States Have Strategies Several states have initiatives in place to facilitate beneficiaries’ eligibility and Incentives to Increase Use for transitional Medicaid. As we have previously reported, difficult of Transitional Medicaid trade-offs exist between the need for program integrity and ease of enrollment for beneficiaries.42 In this regard, states—and HCFA in its oversight capacity—must balance efforts to simplify and streamline eligibility processes with efforts to ensure that benefits go only to qualified individuals. Transitional Medicaid, which is available for a limited time after an individual moves from welfare to work, has been the focus of some states’ strategies to increase family receipt of Medicaid. Connecticut, which has an approved waiver from HCFA to provide 24 months of transitional Medicaid, also has waiver authority to eliminate quarterly income reporting.43 At the end of the time limit for cash benefits, the recipient is asked to participate in an interview, at which time eligibility for ongoing Medicaid is explored. State officials indicated that if the person does not attend the exit interview, they rely on their own records to determine if the family currently has earned income and grants transitional Medicaid coverage if this is the case. If the family does not have earned income, however, or does not provide other information needed to determine ongoing eligibility, the case is closed. If the family subsequently reports earned income or other information indicating ongoing eligibility within 6 months of the case closure, Connecticut initiates transitional Medicaid coverage. 41 The Southern Institute on Children and Families is an independent, nonprofit public policy organization founded in 1990 that tries to improve opportunities for disadvantaged children and families in the South. 42 See Medicaid: Demographics of Nonenrolled Children Suggest State Outreach Strategies, (GAO/HEHS-98-93, Mar. 20, 1998). 43 Oregon provides the required 12 months of transitional Medicaid coverage and also obtained a waiver before welfare reform to eliminate quarterly income reporting for transitional Medicaid beneficiaries. Page 28 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Kansas revised its computer systems so that eligible families leaving cash assistance or Medicaid are automatically transferred to an alternative health program, such as transitional Medicaid, one of the expansion categories for children, or the state’s SCHIP program. In addition, Kansas workers randomly contact families who are leaving cash assistance to determine their health insurance status and to ensure that they obtain the additional months of Medicaid coverage for which they are eligible. As a result, Kansas officials estimated that about 70 percent of the families leaving cash assistance or Medicaid receive transitional coverage. Indiana and Michigan officials informed us that they, too, have taken steps to improve participation in transitional Medicaid. Indiana instituted a statewide campaign to train eligibility workers about the importance of entering earnings information in the state’s eligibility system. As of November 1998, Indiana officials reported that 13,126 families were receiving transitional Medicaid—an increase of 117 percent since May 1998. Michigan officials also reported a significant improvement. Between October 1992, when the state began its present welfare reform initiative, and November 1998, participation in transitional Medicaid increased more than fourfold—from 28,301 to 125,493 individuals. In Michigan, eligibility workers trigger transitional coverage for families whose earnings are likely to make them ineligible for Medicaid in the upcoming quarter. Officials in South Carolina, Utah, and North Dakota encourage increased participation in transitional Medicaid by contacting families with closed cash assistance cases to determine whether these families have obtained the additional months of Medicaid coverage they may be entitled to receive. Both South Carolina and Utah have pre-welfare-reform waivers to provide 24 months of transitional Medicaid. South Carolina officials told us that they have used county-level goal setting and surveys of closed cash assistance cases to increase enrollment in their state’s transitional Medicaid program. The results of a February 1999 survey showed that the percentage of families receiving transitional Medicaid had increased from about 75 percent between October and December 1996 to about 77 percent between October and December 1997. As a quality control measure, Utah officials use system-generated monthly lists of closed Medicaid cases to contact families to determine whether they have received their 24 months of coverage. North Dakota eligibility workers report preferring that families leaving cash assistance receive transitional Medicaid because, if those families leave Medicaid and reapply, they are likely to be placed in Page 29 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 an eligibility category that requires more burdensome monthly income reporting and monitoring. HCFA Initiative Seeks to In view of concerns that beneficiary reporting requirements are limiting Eliminate Beneficiary Income the use of the transitional Medicaid benefit, HCFA has proposed legislation Reporting Requirements for aimed at simplifying transitional Medicaid. In particular, this proposal Transitional Medicaid would eliminate beneficiary reporting requirements for transitional Medicaid benefits for the full period of required eligibility (up to 1 year). Essentially, the failure to report income on a quarterly basis would no longer result in a beneficiary’s removal from Medicaid enrollment. To date, no action has been taken on this proposal, which has been submitted as a part of the President’s fiscal year 2000 budget. Some States Have Initiated Six states we contacted—California, Florida, Georgia, South Carolina, Medicaid Outreach and Utah, and Wisconsin—have initiated or adapted their Medicaid outreach Beneficiary Education and education programs to specifically address any confusion among beneficiaries following welfare reform. Aside from eligibility-related issues Campaigns to Lessen involving noncitizens, confusion about whether receiving Medicaid counts Confusion Over Welfare against the 5-year limit for welfare benefits, and uncertainty about the Reform impact of TANF sanctions on Medicaid, beneficiary advocates were concerned that welfare reform would deter eligible low-income families from seeking Medicaid coverage. To address these issues, the welfare reform law set aside $500 million in Medicaid funds that states could use for a variety of Medicaid-related administrative costs following welfare reform. The law also offered an enhanced Medicaid matching rate for outreach and beneficiary education activities. However, as of December 31, 1998—the date of the most recent available data—HCFA documents showed that the states had claimed only $25.4 million from the fund; the 21 states we contacted accounted for $7.4 million of the expenditures.44 Individually, some states have initiated efforts to counter any confusion that may have resulted from welfare reform. For instance, South Carolina contracted with the Southern Institute for Children and Families to produce education and outreach materials for distribution to beneficiaries and employers on post-welfare-reform benefits, such as Medicaid, that low-income working families might be eligible to receive. This effort was the result of an outreach project involving several of the state’s larger counties. In addition, South Carolina randomly surveys 500 families 44 According to HCFA officials, the agency does not track the states’ specific uses of the set-aside funds because a variety of administrative and outreach purposes are permissible. Page 30 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 quarterly after their welfare cases have closed to determine, among other things, if the families have health insurance or are Medicaid-eligible. State officials use the survey results to provide feedback to the counties. South Carolina officials believe that performance goals related to job placements have had the effect of an added incentive to counties to follow up with families to ensure that eligible families do not lose Medicaid coverage. Education and outreach efforts pose additional challenges for states with large immigrant populations, such as California and Florida. State and local officials in California told us that citizenship and residency concerns within the state’s immigrant communities have had a significant chilling effect on new applications. For example, in February 1998, Los Angeles County initiated a project to enroll 100,000 of the area’s estimated 300,000 uninsured low-income children in Medicaid by September 1999. By December 1998, only 35,000 to 40,000 additional children had been enrolled, despite expanded community-based outreach. Both beneficiary advocates and county officials attributed the low enrollment to the immigrant communities’ concerns that receiving Medicaid, even for children who are citizens, might jeopardize relatives’ pending applications for citizenship or changes in residence status. Florida officials noted a similar effect in their state, where immigrant families decline to apply for Medicaid because of concerns about jeopardizing their immigration status. Florida officials have been working with numerous community-based organizations and housing projects to counteract the misunderstanding or mistrust that remains within immigrant communities. In light of the fears and confusion among immigrants regarding this issue, the administration has recently published a proposed rule to clarify the circumstances in which individuals can accept certain public benefits without fear of negative immigration consequences.45 The proposed rule specifies a list of benefits, prepared by HHS, the Immigration and Naturalization Service (INS), and the State Department, that immigrants can receive without affecting their admission to the United States or their resident status. Under current law, before admitting someone as a legal permanent resident, the INS or State Department must conclude that the individual is not likely to become a “public charge”—that is, a person whose main source of support is from government programs. Medicaid and SCHIP are among the benefits specified in the proposed rule that would 45 64 Fed. Reg. 28,675 (May 26, 1999). Page 31 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 be exempt from the “public charge” test for immigrant admission, adjustment, or deportation.46 Some states have put a significant amount of effort into developing enrollment outreach programs, as well as increasing the number of locations at which eligibility to facilitate enrollment and workers are available to inform beneficiaries and providers that Medicaid eligibility is no longer tied to cash assistance. Several examples follow. • Georgia is using nearly 150 “Right From the Start” Medicaid outreach eligibility workers to enroll Medicaid-eligible children and to act as intermediaries for families who are seeking only Medicaid coverage and do not wish to go to a local welfare office to apply. • Utah has been able to reach beyond the traditional “outstation” locations by placing additional workers in schools, Indian reservations, and large medical clinics.47 The state has also allowed families to apply for Medicaid by telephone and through the mail. Additionally, Utah’s Department of Workforce Services, which oversees the TANF program, also accepts Medicaid applications by telephone. • In Wisconsin, where the welfare and Medicaid programs are separately administered, officials were particularly concerned about the confusion this separation could cause beneficiaries and providers. In an effort to avoid such confusion, Wisconsin has increased the number of outstationed locations in Milwaukee—one of the state’s larger urban areas—and contracted with advocates to assist beneficiaries in navigating the new system. SCHIP Outreach and While officials in most of the 21 states we contacted reported that Related Simplifications outreach for SCHIP has had a positive spillover effect on Medicaid May Increase Future enrollment among children, officials in 6 states specifically suggested that such efforts may have directly contributed to enrollment increases or Medicaid Enrollment stabilization in their Medicaid programs. Although firm data are not yet available, officials from the states in our sample estimated that perhaps as 46 While Medicaid would not be considered in public charge determinations, the proposed rule specifies an exception: Medicaid or similar state programs would be considered in limited circumstances if they were needed to pay for long-term care, in the form of nursing home or institutionalized care for the individual. The proposed rule provides, however, that the need for long-term care alone would not automatically result in a public charge determination. INS and State Department officials would need to consider other factors required by law (such as age, health, family status, and assets), and determinations would be made on a case-by-case basis. 47 Section 4602 of the Omnibus Budget Reconciliation Act of 1990 (P.L. 101-508) added the requirement that states “outstation” eligibility workers at locations other than local welfare offices, allowing mothers and children to apply for Medicaid at the sites where they receive health care. Page 32 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 many as an additional 135,000 children have recently been enrolled in Medicaid as a result of SCHIP outreach and the requirement that states screen all SCHIP applicants for Medicaid eligibility and enroll those who qualify. For instance, Michigan officials reported that they enrolled two children in Medicaid for each SCHIP enrollee. However, since SCHIP was enacted in 1997 and program implementation was just getting under way in 1998 and 1999 in most states, these spillover enrollment effects are not reflected in our 1997 Medicaid data. While state officials did not report a similar spillover effect on adult enrollment, SCHIP has resulted in simplified Medicaid applications and redetermination processes for children that may also facilitate enrollment among adults. For instance, states such as California, Kansas, Utah, and Vermont have begun to simplify their Medicaid enrollment and application processes in the following ways: • California has shortened its 28-page joint Medicaid/SCHIP booklet on child eligibility to a 4-page application. • In Kansas, since January 1999, low-income families have been able to submit Medicaid redetermination information by mail and are no longer required to meet personally with an eligibility worker. • Utah had instituted several innovative procedures even before SCHIP, including allowing application by mail or telephone and redetermination by mail, telephone, or facsimile. • Vermont has created a centralized Health Access and Eligibility Unit to receive and process mailed applications for those applying only for state medical assistance. SCHIP has also sparked certain states to consider and implement a variety of ways to make their enrollment and application processes less burdensome, including providing applications at alternative locations such as schools, Head Start centers, and community action agencies. Other states have adopted mail-in applications and community-based worker assistance, and one state is considering the feasibility of accepting applications over the Internet. Despite federal protections to ensure that low-income families retain Conclusions health insurance regardless of whether they are receiving cash assistance, it has become more complicated for eligible low-income families to establish and keep Medicaid coverage with the advent of welfare reform. States are challenged with identifying and enrolling families that no longer Page 33 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 qualify for cash assistance yet continue to retain Medicaid eligibility. Some states have taken advantage of the flexibility under welfare reform by using protections provided by the law to ensure that Medicaid coverage is sustained for low-income families that are transitioning to work. Other states have found it increasingly difficult to communicate to both beneficiaries and workers that Medicaid coverage can be maintained even though changes in welfare policies may limit or deny cash assistance. National declines in Medicaid enrollment raise questions about whether states have been able to “de-link” welfare and Medicaid policies in a manner that consistently ensures Medicaid coverage for eligible individuals. Transitional Medicaid is a protection offered to families at a critical juncture in their efforts to move from welfare to work. Employment in low-wage positions frequently does not provide adequate access to affordable health insurance, making Medicaid coverage an important benefit. However, there are indications that procedural difficulties with income reporting—coupled with a lack of national data and the apparently disparate use of this benefit by the states—are limiting the extent to which beneficiaries are receiving transitional Medicaid and maintaining their eligibility for it. Before welfare reform, states were able to obtain authority from HCFA to waive certain beneficiary reporting requirements. Presently, however, states without waivers must comply with section 1925 of the Social Security Act, which requires beneficiary income reporting even though income level does not affect eligibility for the first 6 months of transitional Medicaid. As a result, families that do not comply with this requirement can be terminated from transitional Medicaid, despite their income eligibility for this entitlement. There is precedent for a less burdensome approach, by which states could be allowed to lessen or eliminate beneficiary income reporting requirements. For example, the BBA allowed states to guarantee a longer period of Medicaid coverage for children, regardless of changes in a family’s financial status or size. Similarly, HCFA has proposed eliminating beneficiary income reporting requirements. Our work suggests that removing reporting requirements would be beneficial to increasing the use of transitional Medicaid, provided that sufficient safeguards remained in place to ensure that only those who are qualified receive the benefits. Information on the extent to which transitional Medicaid is implemented across the states is scarce. HCFA is responsible for overseeing the states’ implementation of this entitlement and is in the position to serve as a Page 34 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 conduit of technical assistance and dissemination of states’ best practices in implementing transitional Medicaid. Doing so could heighten understanding of systemic barriers and provide states with strategies to foster and maintain transitional Medicaid coverage for eligible families. To further facilitate families’ making the transition from welfare to work Recommendation to and to prevent income-eligible families from being terminated from the Congress Medicaid for procedural reasons, we recommend that the Congress consider revising section 1925 of the Social Security Act. Specifically, the Congress may wish to allow states to lessen or eliminate periodic income reporting requirements for families receiving transitional Medicaid coverage, provided that states offer adequate assurances that the benefits are reserved for those who are eligible. Actions in this regard could facilitate uninterrupted health insurance coverage for families that are moving from cash assistance to the workforce. In order to ensure that eligible individuals leaving cash assistance do not Recommendations to lose Medicaid coverage, we recommend that the Administrator of HCFA the Administrator of HCFA • determine the extent to which transitional Medicaid is reaching the eligible population and • provide states with guidance or other appropriate technical assistance regarding best approaches for implementing transitional Medicaid in a manner that facilitates the full and appropriate use of this entitlement for eligible beneficiaries. We provided ACF, HCFA, and officials from the 21 states in our sample an Agency and Other opportunity to review a draft of this report. While ACF reviewed the report, Comments it did not suggest any changes to its content. HCFA concurred with our conclusions and recommendations and highlighted steps it has taken to ensure that states understand Medicaid eligibility and the enrollment options families have following welfare reform. HCFA also noted a number of studies it is sponsoring, along with HHS and the Office of the Assistant Secretary for Planning and Evaluation, to better understand the factors contributing to declining enrollment. In particular, HCFA plans to use the results of a 6-state study performed by an independent contractor as the basis for a more extensive longitudinal analysis of individual Medicaid eligibility in 8 to 10 states. Additional Page 35 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 studies, which are planned or under way, include comparisons of national trends in pre- and post-welfare-reform Medicaid enrollment and expenditures. HCFA also commented that it plans to issue additional guidance on transitional Medicaid, conduct outreach to beneficiaries, and propose legislative changes to make access to transitional Medicaid less burdensome. Finally, HCFA cited plans to provide on-site technical assistance to the states to further assist in the coordination between TANF and Medicaid. Specifically, HCFA officials intend to visit every state to ensure that states are taking full advantage of their opportunities and that states are meeting the challenges posed by changes in the BBA and welfare reform. We agree that the unique character of each state’s welfare and Medicaid programs warrants an individualized review of state-level activities. We believe that efforts in this regard should be based on a comprehensive analysis of state-level activities, including an evaluation of the experiences and barriers particular to individual states. Further, we caution HCFA about relying on 2082 data as the primary indicator of Medicaid enrollment, given the data limitations we found.48 (See app. I.) HCFA’s written comments are provided appendix III. In addition to the problematic transitional Medicaid reporting requirements, some responding states also identified other barriers or challenges to transitional coverage and Medicaid, in general.49 Several states cited as barriers the requirement that beneficiaries must have received Medicaid coverage on the basis of AFDC-related criteria in 3 of the previous 6 months to be eligible for transitional Medicaid. According to officials in Florida, because the median length of time that families stay on TANF is 3 months, about half of the state’s recipients leave TANF before meeting the 3-month criteria. Another barrier or challenge to Medicaid enrollment, as noted by one state’s official, is the “disconnect” between cash assistance and Medicaid—two programs that once worked in tandem now, at times, appear to have competing goals. While the TANF program emphasizes self-sufficiency and employment, Medicaid encourages coverage for all eligible individuals. According to this official, reconciling the two programs’ goals poses a continuing challenge for states implementing welfare reform. 48 “2082” is an annual state-submitted report designed to collect statistical data on Medicaid. 49 We sent a draft of this report to officials in all 21 states in our sample: 17 responded; Delaware, Georgia, Idaho, and North Dakota did not. Page 36 GAO/HEHS-99-163 Medicaid After Welfare Reform B-281152 Several states also expressed the concern that our 1995 to 1997 enrollment data do not reflect the effects of policy changes and program expansions implemented since 1997. We agree that the time frame for this analysis represents a “snapshot” of state experience and may not reflect the evolving nature of Medicaid enrollment in individual states. As enrollment begins to stabilize or to reverse previous declines in some states and more current data become available, further analysis to determine the status of Medicaid enrollment as it relates to welfare reform would be warranted. Several states provided technical comments, which we incorporated as appropriate. As arranged with your offices, unless you announce its contents earlier, we plan no further distribution of this report until 14 days after its issuance date. At that time, we will send copies to the Honorable Donna Shalala, Secretary of HHS; the Honorable Nancy-Ann Min DeParle, Administrator of HCFA; the Honorable Olivia Golden, Administrator of ACF; directors of the programs in the 21 states we contacted; and interested congressional committees. Copies of this report will also be made available to others upon request. If you have any questions about this report, please contact me at (202) 512-7114. Other GAO contacts and staff acknowledgments are in appendix IV. Kathryn G. Allen Associate Director, Health Financing and Public Health Issues Page 37 GAO/HEHS-99-163 Medicaid After Welfare Reform Contents Letter 1 Appendix I 40 Limitations of HCFA Enrollment Data 42 Scope and Methodology Appendix II 44 Analysis of States’ Medicaid Enrollment and Welfare Participation Data Appendix III 51 Comments From the Health Care Financing Administration Appendix IV 58 GAO Contact and Staff Acknowledgments Related GAO Products 60 Tables Table 1: Percentage Changes in Medicaid Enrollment and Welfare 8 Participation in Our State Sample, 1995-97 Table 2: Percentage Changes in Child and Adult Medicaid 14 Enrollment in Our State Sample, 1995-97 Table 3: Medicaid Coverage of Pregnant Women, Infants, and 16 Children in Our State Sample as of September 30, 1997 Table 4: Comparison of 21 States’ Welfare Diversion Policies as of 20 April 1999 Table II.1: Changes in Adult/Child Medicaid Enrollment Between 44 1995 and 1997 Table II.2: Changes in Welfare Participation Between 1995 and 46 1997 Page 38 GAO/HEHS-99-163 Medicaid After Welfare Reform Contents Table II.3: Comparison of Changes in Medicaid and Welfare 49 Enrollment Between 1995 and 1997 Figure Figure 1: Changes in States’ Medicaid Enrollment, 1995-97 10 Abbreviations ACF Administration for Children and Families AFDC Aid to Families With Dependent Children BBA Balanced Budget Act of 1997 HCFA Health Care Financing Administration HHS Department of Health and Human Services INS Immigration and Naturalization Service SCHIP State Children’s Health Insurance Program TANF Temporary Assistance for Needy Families Page 39 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix I Scope and Methodology To analyze Medicaid enrollment for families and children following welfare reform, we examined state-level data from two sources: (1) state-provided monthly Medicaid enrollment data and (2) the Health Care Financing Administration’s (HCFA) federal fiscal year data on the states’ annual enrollment.50 We chose 1995 and 1997 for our analysis because 1995 provided a baseline for enrollment before the 1996 enactment of welfare reform, and 1997 was the most current year for which HCFA enrollment data were available when we initiated our work. We limited our analysis to nonelderly and nondisabled adult and child enrollment because this segment of the Medicaid population was the most likely to have been enrolled in the states’ cash assistance programs that were affected by welfare reform. To report the change in Medicaid enrollment between 1995 and 1997, we used state-provided data because average monthly data provided a better indicator of changes in states’ Medicaid enrollment than the cumulative, annual count of enrollees that HCFA reports. Moreover, we found significant inconsistencies in the HCFA data for federal fiscal years 1995 and 1997, which are described in greater detail at the end of this appendix. We analyzed average monthly enrollment data for calendar years 1995 and 1997 that we collected by contacting the 50 states and the District of Columbia.51 When states did not provide 12 months of data for each year, we extrapolated the data provided to derive annual averages. Finally, we obtained state-specific welfare participation data from the Administration for Children and Families (ACF). To review the effects of minimum wage employment on Medicaid eligibility, we used the Department of Labor’s 1997 minimum wage data and the minimum income disregards required by the former AFDC program that applied to the first 4 months of employment. This approach enabled us to determine the extent to which heads of three-member households might work and continue to qualify for Medicaid coverage. We estimated that family income from working 52 weeks at 40 hours per week, at the 1997 minimum wage of $4.75 per hour, would be approximately $823 per month. After deducting the standard disregards required by the former Aid to Families With Dependent Children (AFDC) program and still applicable in determining Medicaid eligibility ($30 plus one-third of $793), we calculated that family income would be approximately $529. If states also 50 HCFA collects and publishes annual Medicaid enrollment on a state-by-state basis as part of the agency’s 2082 reporting format. 51 We were unable to obtain comparable monthly data for the District of Columbia, Rhode Island, and West Virginia. Page 40 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix I Scope and Methodology exercised their option to disregard a portion of work- and child care-related expenses, family income could be even less. According to the most recent data available on the optional disregards, states disregarded an average of $102 per month for work-related expenses and $137 for child care expenses in 1995. Applying these averages to $529 results in a net income of $290 per month. With countable family income of $290 per month, in all but 4 of the 21 states included in our sample, heads of three-member households can work full-time at the minimum wage and continue to qualify for Medicaid coverage. In three of the four states—South Carolina, Indiana, and Kentucky—families can be considered Medicaid-eligible even if their income is above the standard used to determine eligibility for cash assistance. In Colorado, gross income for a three-member family must be below $778 per month for the state to apply the income disregards, and net income must be below $421 per month. As a result, three-member households in Colorado can work only 36 hours per week at the 1997 minimum wage if they are to remain Medicaid-eligible. Under the former AFDC program, states determined the amount of income families of varying sizes needed for a minimal standard of living—the “need standard”—and set payment standards that represented the maximum AFDC cash assistance payment families were entitled to receive. Nationally, most states’ AFDC payments were below the states’ need standards.52 To identify the procedures and protections states are using to enroll Medicaid-eligibles, we judgmentally selected 21 states to contact for additional review, over-sampling for states with declines in Medicaid enrollment to focus our analysis on those policies and practices that may have contributed to declines in enrollment. We initially selected the 15 states with the largest declines in Medicaid enrollment, based on a preliminary analysis of HCFA data, and subsequently expanded the sample by adding states with relatively stable or increased enrollment. The 21 states represented about 46 percent and 45 percent of Medicaid enrollment in 1995 and 1997, respectively, as well as approximately 39 percent of total program expenditures for fiscal year 1997. In addition to geographic diversity, the states had varying degrees of experiences with and approaches to welfare reform. We visited four states and several locales within the states—California (Sacramento and Los Angeles), Florida (Tallahassee and Miami), Maryland (Baltimore and Prince George’s County), and Wisconsin (Madison and Milwaukee). We interviewed by 52 Under welfare reform, a South Carolina family of three, for example, may have monthly “countable” income of up to $667 and remain Medicaid-eligible because the state’s need standard is $668. Page 41 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix I Scope and Methodology telephone, collecting and analyzing documentation on Medicaid eligibility and application processes from officials in the 17 remaining states—Colorado, Connecticut, Delaware, Georgia, Idaho, Indiana, Kansas, Kentucky, Michigan, Nevada, North Dakota, Ohio, Oklahoma, Oregon, South Carolina, Utah, and Vermont. Using structured interview protocols in each of the four site-visit states, we interviewed knowledgeable state and local Medicaid and welfare officials, beneficiary advocates, and eligibility workers. From state-level officials, we obtained and analyzed information and documentation on state preapplication policies, such as diversion assistance and up-front job search requirements, Medicaid application procedures and locations, eligibility determination policies, transitional Medicaid and former welfare recipients’ health insurance status, TANF sanctions that can affect Medicaid coverage, and state outreach strategies. Our local welfare office interview protocol covered office organization and eligibility worker training, outreach, initial applicant contact, preapplication activities (diversion assistance and up-front job search requirements), and Medicaid application procedures and eligibility determination policies. For the other 17 states, we obtained information on state Medicaid eligibility criteria, eligibility determination processes and computer systems, transitional Medicaid, outreach strategies, preapplication activities (diversion and up-front job search requirements), application procedures, and health insurance status of former welfare recipients. Also, we obtained and reviewed various reports and studies and interviewed officials representing organizations including the American Public Human Services Association (formerly known as the American Public Welfare Association), the Center on Budget and Policy Priorities, the Children’s Defense Fund, the George Washington University’s Center for Health Policy Research, the National Eligibility Workers Association, the National Governors’ Association, the National Health Law Program, the Southern Institute on Children and Families, and The Urban Institute. HCFA’s enrollment data represent an attempt to provide an unduplicated Limitations of HCFA annual count of Medicaid enrollees, whereas state monthly enrollment Enrollment Data data show the number of individuals enrolled in the program each month. For our analysis of changes in Medicaid enrollment, we relied primarily on the average monthly Medicaid enrollment data that we obtained directly from the states because of significant inconsistencies that we found in Page 42 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix I Scope and Methodology HCFA’s enrollment data for federal fiscal years 1995 and 1997. For example, we found duplicate counts in some of HCFA’s state data as well as inconsistencies in HCFA’s use of reporting categories. West Virginia reported that HCFA may have double-counted adult and child enrollees for fiscal year 1995, thus substantially overstating the extent of Medicaid declines between 1995 and 1997. Similarly, HCFA’s data for Oregon showed about an 18-percent increase in adult and child enrollment due to HCFA’s overcounting the number of infants and children in 1997, while Oregon’s monthly data reflected a 13-percent decline. Oregon’s Medicaid director confirmed that enrollment had indeed declined between 1995 and 1997. Louisiana officials told us that HCFA’s 1997 data inappropriately categorized most of the state’s adult and child enrollees as aged, resulting in HCFA’s reporting a nearly 50-percent decline in adult and child enrollment, rather than the 7-percent decline reflected by the state’s average monthly enrollment data for the same period. HCFA officials acknowledged that comparing the 2 years’ data could have been problematic because in fiscal year 1997 the agency changed its reporting format and categories. HCFA officials noted that steps were being taken to improve the overall reliability of future years’ enrollment data. HCFA officials believe that the Balanced Budget Act requirement that states use the Medicaid Statistical Information System reporting format to electronically submit all Medicaid claims and enrollment information as of January 1999 will improve categorical consistency among the states. They believe outside contractor assistance in screening state data will be helpful as well. Page 43 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Using 1995 and 1997 data to compute states’ average annual monthly enrollment, the aggregate national decline for the adult and child portion of Medicaid enrollment was 7.4 percent. The median decline was about 7 percent. Medicaid enrollment among the nonelderly and nondisabled adults and children ranged from a 19-percent decline in Wisconsin to an approximately 26-percent increase in Delaware. Enrollment declined by 10 percent or more in 12 states, declined between 3 and 10 percent in 20 states, declined or increased 3 percent or less in 12 other states, and increased 5 percent or more in 4 states. See table II.1. Table II.1: Changes in Adult/Child Medicaid Enrollment Between 1995 Percentage enrollment Average monthly and 1997 change, enrollment State 1995-97 1995 1997 Alabama –4.2 289,333 277,041 Alaska +0.2 52,197 52,306 Arizonaa +2.8 322,904 331,908 Arkansasa –0.5 139,175 138,472 California -7.6 4,189,509 3,869,454 Colorado -9.6 169,957 153,592 Connecticut -1.0 224,128 221,935 Delaware +26.2 49,085 61,953 District of Columbiab Floridac -13.4 988,805 855,888 Georgiaa -4.6 674,374 643,301 c Hawaii –13.9 153,103 131,834 Idahoa –10.8 55,746 49,745 Illinois –6.8 1,083,802 1,010,397 Indianaa –9.0 337,791 307,429 Iowa –5.8 230,139 216,742 Kansas –10.5 133,295 119,265 a Kentucky –10.1 828,403 744,418 Louisianad –7.1 499,265 463,999 d Maine –4.2 107,791 103,297 Maryland –7.2 319,799 296,843 Massachusetts +1.2 409,713 414,639 Michigane –8.9 688,646 627,561 Minnesota –0.8 340,330 337,529 Mississippi –13.4 254,801 220,536 a Missouri –5.2 436,945 414,276 (continued) Page 44 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Percentage enrollment Average monthly change, enrollment State 1995-97 1995 1997 a Montana –8.6 76,518 69,947 Nebraska +5.1 99,977 105,047 Nevada –11.5 65,480 57,943 New Hampshiree +1.2 60,296 61,034 New Jersey –4.9 489,753 465,959 New Mexico +13.7 182,543 207,507 New York –9.4 2,248,274 2,037,802 North Carolina –1.5 526,780 518,650 North Dakota –8.0 31,110 28,628 Ohio –15.9 900,347 756,916 Oklahoma –9.3 199,383 180,838 Oregon –13.0 329,786 286,779 Pennsylvania –17.4 1,253,478 1,035,142 Rhode Islandb South Carolina +2.4 231,882 237,338 South Dakota –0.5 39,264 39,081 a Tennessee –0.3 910,494 907,485 Texas –8.6 1,592,553 1,455,059 Utaha –5.0 97,733 92,881 Vermont +2.6 61,950 63,548 Virginia –8.1 362,383 333,092 Washington +13.2 489,427 554,030 b West Virginia Wisconsin –19.0 316,419 256,449 Wyoming –14.5 29,694 25,400 Total for all states –7.4 23,574,557 21,840,914 Total for our sample states –8.9 10,893,628 9,912,703 (Table notes on next page) Page 45 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Note: States in boldface were part of our sample. a The average monthly enrollment for these states was calculated using less than a full year of monthly data. b We were unable to obtain comparable monthly enrollment data for the District of Columbia, Rhode Island, and West Virginia. c Data for Florida and Hawaii may reflect only a portion—approximately 70 percent and 80 percent, respectively—of the state’s nonelderly, nondisabled adult/child enrollment. d Louisiana and Maine officials were unable to provide separate monthly enrollment data for their welfare-related populations. As a result, we arrived at these figures by using each state’s 2082 data submission to HCFA. e Michigan and New Hampshire provided a yearly enrollment figure that reflected their states’ fiscal years—October 1st to September 30th and July 1st to June 30th, respectively. As a result, the monthly enrollment data calculations for Michigan are for October 1994 through September 1995 and October 1996 through September 1997; for New Hampshire, the calculations are for July 1994 through June 1995 and July 1996 through June 1997. Source: GAO analysis of state monthly enrollment data. We calculated changes in welfare participation by using the average of January 1995 and January 1996 recipient data to arrive at the figure for 1995 and the average of January 1997 and January 1998 recipient data for the 1997 figure. Over this period, welfare participation declined on average by about 23 percent, ranging from as much as nearly 56 percent in Wisconsin to less than 7 percent in Alaska. Hawaii was the only state to experience an increase in welfare participation between 1995 and 1997. See table II.2. Table II.2: Changes in Welfare Participation Between 1995 and 1997 Percentage decline in Average annual welfare welfare participation State participation 1995 1997 Alabama –33.3 115,053 76,766 Alaska –6.6 36,348 33,939 Arizona –27.8 183,350 132,368 Arkansas –26.5 62,274 45,792 California –13.5 2,670,487 2,310,530 Colorado –32.2 105,241 71,393 Connecticut –11.5 166,228 147,184 Delaware –15.8 24,734 20,823 District of Columbia –12.9 71,206 62,000 Florida –35.2 616,433 399,608 Georgia –30.4 378,285 263,348 (continued) Page 46 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Percentage decline in Average annual welfare welfare participation State participation 1995 1997 Hawaii +7.0 65,949 70,565 Idaho –49.0 23,799 12,129 Illinois –17.8 686,622 564,353 Indiana –36.8 172,154 108,820 Iowa –24.2 97,418 73,890 Kansas –37.0 76,131 47,995 Kentucky –20.3 185,162 147,559 Louisiana –34.7 248,714 162,493 Maine –21.2 58,646 46,222 Maryland –31.2 217,844 149,960 Massachusetts –25.2 264,374 197,872 Michigan –26.9 573,964 419,638 Minnesota –14.5 176,203 150,616 Mississippi –37.3 139,674 87,564 Missouri –25.4 248,824 185,541 Montana –27.8 33,435 24,138 Nebraska –7.5 40,346 37,313 Nevada –29.3 41,169 29,118 New Hampshire –31.2 26,595 18,287 New Jersey –23.0 307,492 236,692 New Mexico –25.6 103,881 77,287 New York –18.3 1,233,599 1,007,952 North Carolina –25.7 299,961 222,729 North Dakota –27.0 14,286 10,424 Ohio –23.5 591,012 452,417 Oklahoma –34.0 118,917 78,471 Oregon –42.2 99,896 57,740 Pennsylvania –24.5 582,182 439,714 Rhode Island –11.1 61,531 54,673 South Carolina –32.9 127,635 85,628 South Dakota –28.6 17,237 12,303 Tennessee –38.8 273,651 167,457 Texas –27.9 739,992 533,221 Utah –26.2 44,309 32,681 Vermont –16.8 26,791 22,292 Virginia –31.6 177,753 121,623 (continued) Page 47 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Percentage decline in Average annual welfare welfare participation State participation 1995 1997 Washington –13.1 283,479 246,258 West Virginia –27.2 103,054 75,019 Wisconsin –55.6 199,307 88,507 Wyoming –54.3 14,483 6,613 Total for all states –23.4 13,227,097 10,127,510 Total for our sample states –23.4 6,473,778 4,956,260 Note: States in boldface were part of our sample. Source: ACF’s AFDC/TANF recipient data. Analyzing state-provided monthly Medicaid enrollment data for the nonelderly and nondisabled adults and children between January 1995 and December 1997 and comparable years’ welfare data from ACF, we found that welfare participation declined nationally 1.2 times more than Medicaid enrollment. State-by-state analysis showed some variance in the states’ experiences, as shown in table II.3. State ratios ranged from 1.9 in Wyoming—where welfare participation declined by over 54 percent and Medicaid declined by 14.5 percent—to a ratio of .8 in Hawaii—where welfare participation increased 7 percent while Medicaid declined by nearly 14 percent. However, 38 states were within (+/-) .2 of the national ratio. In addition, a correlation analysis of the data showed a statistically significant relationship between changes in welfare participation and changes in Medicaid enrollment between the 2 years (r = .39, p < .01). In general, states that experienced a decline in welfare participation also had a decline in Medicaid enrollment. While the correlation is statistically significant, only 15 percent of the change in Medicaid enrollment may be explained by its relationship to the change in welfare participation (r2 = .15). Consequently, there are factors in addition to welfare reform that influenced Medicaid enrollment between 1995 and 1997. Page 48 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data Table II.3: Comparison of Changes in Medicaid and Welfare Enrollment 1997 enrollment as a Between 1995 and 1997 proportion of 1995 Ratio of enrollment welfare-to- Medicaid Welfare Medicaid State enrollment enrollment changea Alabama 0.96 0.67 1.4 Alaska 1.00 0.93 1.1 Arizona 1.03 0.72 1.4 Arkansas 0.99 0.74 1.4 California 0.92 0.74 1.4 Colorado 0.90 0.68 1.3 Connecticut 0.99 0.89 1.1 Delaware 1.26 0.84 1.5 District of Columbiab Florida 0.87 0.65 1.3 Georgia 0.95 0.70 1.4 Hawaii 0.86 1.07 0.8 Idaho 0.89 0.51 1.8 Illinois 0.93 0.82 1.1 Indiana 0.91 0.63 1.4 Iowa 0.94 0.76 1.2 Kansas 0.89 0.63 1.4 Kentucky 0.90 0.80 1.1 Louisiana 0.93 0.65 1.4 Maine 0.96 0.79 1.2 Maryland 0.93 0.69 1.3 Massachusetts 1.01 0.75 1.4 Michigan 0.91 0.73 1.2 Minnesota 0.99 0.85 1.2 Mississippi 0.87 0.63 1.4 Missouri 0.95 0.75 1.3 Montana 0.91 0.72 1.3 Nebraska 1.05 0.92 1.1 Nevada 0.88 0.71 1.3 New Hampshire 1.01 0.69 1.5 New Jersey 0.95 0.77 1.2 New Mexico 1.14 0.74 1.5 New York 0.91 0.82 1.1 North Carolina 0.98 0.74 1.3 North Dakota 0.92 0.73 1.3 (continued) Page 49 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix II Analysis of States’ Medicaid Enrollment and Welfare Participation Data 1997 enrollment as a proportion of 1995 Ratio of enrollment welfare-to- Medicaid Welfare Medicaid State enrollment enrollment changea Ohio 0.84 0.77 1.1 Oklahoma 0.91 0.66 1.4 Oregon 0.87 0.58 1.5 Pennsylvania 0.83 0.76 1.1 Rhode Islandb South Carolina 1.02 0.67 1.5 South Dakota 1.00 0.71 1.4 Tennessee 1.00 0.61 1.6 Texas 0.91 0.72 1.3 Utah 0.95 0.74 1.3 Vermont 1.03 0.83 1.2 Virginia 0.92 0.68 1.3 Washington 1.13 0.87 1.3 b West Virginia Wisconsin 0.81 0.44 1.8 Wyoming 0.86 0.46 1.9 National ratiosb 0.93 0.77 1.2 Our sample 0.91 0.77 1.2 Note: States in boldface were part of our sample. a We derived these ratios by dividing 1997 average monthly Medicaid and welfare participation by 1995 average monthly Medicaid and welfare participation. b We were unable to obtain comparable enrollment data for the District of Columbia, Rhode Island, and West Virginia. Page 50 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 51 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 52 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 53 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 54 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 55 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 56 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix III Comments From the Health Care Financing Administration Page 57 GAO/HEHS-99-163 Medicaid After Welfare Reform Appendix IV GAO Contact and Staff Acknowledgments Carolyn Yocom, (202) 512-4931 GAO Contact In addition, Carol Carter, Enchelle Bolden, Christine DeMars, JoAnn Staff Martinez, and Craig Winslow made key contributions to this report. Acknowledgments Page 58 GAO/HEHS-99-163 Medicaid After Welfare Reform Page 59 GAO/HEHS-99-163 Medicaid After Welfare Reform Related GAO Products Food Stamp Program: Various Factors Have Led to Declining Participation (GAO/RCED-99-185, July 2, 1999). Welfare Reform: Public Assistance Benefits Provided to Recently Naturalized Citizens (GAO/HEHS-99-102, June 23, 1999). Children’s Health Insurance Program: State Implementation Approaches Are Evolving (GAO/HEHS-99-65, May 14, 1999). Welfare Reform: Information on Former Recipients’ Status (GAO/HEHS-99-48, Apr. 8, 1999). Welfare Reform: States’ Experiences in Providing Employment Assistance to TANF Clients (GAO/HEHS-99-22, Feb. 26, 1999). Welfare Reform: Status of Awards and Selected States’ Use of Welfare-to-Work Grants (GAO/HEHS-99-40, Feb. 5, 1999). Welfare Reform: Few States Are Likely to Use the Simplified Food Stamp Program (GAO/RCED-99-43, Jan. 29, 1999). Year 2000 Computing Crisis: Readiness of State Automated Systems to Support Federal Welfare Programs (GAO/AIMD-99-28, Nov. 6, 1998). Welfare Reform: Early Fiscal Effects of the TANF Block Grant (GAO/AIMD-98-137, Aug. 18, 1998). Welfare Reform: Child Support an Uncertain Income Supplement for Families Leaving Welfare (GAO/HEHS-98-168, Aug. 3, 1998). Welfare Reform: Many States Continue Some Federal or State Benefits for Immigrants (GAO/HEHS-98-132, July 31, 1998). Welfare Reform: States Are Restructuring Programs to Reduce Welfare Dependence (GAO/HEHS-98-109, June 17, 1998). Medicaid: Early Implications of Welfare Reform for Beneficiaries and States (GAO/HEHS-98-62, Feb. 24, 1998). (101762) Page 60 GAO/HEHS-99-163 Medicaid After Welfare Reform Ordering Information The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. Orders by mail: U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists. For information on how to access GAO reports on the INTERNET, send an e-mail message with "info" in the body to: email@example.com or visit GAO’s World Wide Web Home Page at: http://www.gao.gov PRINTED ON RECYCLED PAPER United States Bulk Rate General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. G100 Official Business Penalty for Private Use $300 Address Correction Requested
Medicaid Enrollment: Amid Declines, State Efforts to Ensure Coverage After Welfare Reform Vary
Published by the Government Accountability Office on 1999-09-10.
Below is a raw (and likely hideous) rendition of the original report. (PDF)