1
Background
In the mid-1990s there were more than 11 million children in the United States without any form of health insurance (Institute of Medicine, 1998). This situation led Congress to enact legislation to establish the State Children’s Health Insurance Program (SCHIP) with an appropriation of $40 million over a 10-year period to be allotted to the states (U.S. Congress, 1997). Subsequent difficulties in estimating the size of the SCHIP eligible population, in bringing data to bear on the enrollment process, and in assessing disenrollment from the program led the Office of the Assistant Secretary for Planning and Evaluation, of the Department of Health and Human Services, to contract with the Committee on National Statistics of the National Research Council to organize and hold a workshop to explore some of the ways in which data analysis could be used to promote achievement of the SCHIP goals of expanding health insurance coverage for uninsured children from low-income families. The charge to the workshop was to bring together state SCHIP officials and researchers to share findings and methods that would inform the design, implementation, and evaluation of SCHIP at the state and national levels.
SCHIP was established to provide health insurance to uninsured children whose family income was too high for Medicaid coverage but too low to allow the family to obtain private health insurance coverage. Like Medicaid, SCHIP is a joint federal-state program, with funding from both sources but administered by the states. Thus, in the states and the District of Columbia there are 51 different programs.
The SCHIP authorizing legislation set out criteria that all programs must meet yet left the states a great deal of latitude in determining the structure of the program, eligibility criteria for children, and program implementation. States had the option of organizing their SCHIP program in the form of a new state children’s health insurance program, expanding Medicaid coverage, or establishing a combination of these two approaches. As of December 2000, 17 states had established separate SCHIP coverage, 17 had expanded Medicaid, and 17 had a combination of these two approaches (Smith and Rousseau, 2001). In addition, within federal guidelines, states set eligibility criteria for enrollment. The legislation requires that the states maintain the Medicaid eligibility they had in place on June 1, 1997; thus SCHIP could only be used to expand eligibility for health insurance to those who would not have been eligible under the preexisting Medicaid program. On one hand, if a state chose to implement a Medicaid expansion, it was required to offer those who would be newly covered the same Medicaid benefits package that was already in place. On the other hand, if a state chose to establish a separate SCHIP, it was required to meet the standards laid out by the act that established the program.
The program went into effect on October 1, 1997, but before states could enroll children in SCHIP, they were required to obtain approval of the plan for their program from the secretary of the U.S. Department of Health and Human Services. The plan was required to detail what the state was already doing in insuring children of low-income families, to estimate the number of uninsured children in the state, to describe proposed SCHIP eligibility requirements and benefits, and to describe how the plan would be coordinated with Medicaid and how outreach would be conducted.
By 6 months after the program went into effect, 6 states had obtained approval for their plans; this number increased to 33 by the end of the first year and to 49 by the end of the second year. Two states did not get started until fiscal year (FY) 2000 (Rosenbach et al., 2001). One reason that states were delayed in obtaining approval for their SCHIP plans was the difficulty they had in estimating the numbers of uninsured children who would meet income eligibility criteria for the program in their state and the distribution of eligible children within the state. The Current Population Survey (CPS) is used to allocate federal funds for SCHIP to the states, but the size of the sample is too small to estimate numbers of eligible children for all but the largest staes, let alone for geographic areas within states and to estimate the total number of eligible children in many states reliably. Estimates of the
numbers of uninsured children will also be needed to evaluate the success of the program in reducing the uninsurance rate among children. States encountered a variety of impediments: available data on insurance often defined “uninsured” as lacking insurance at a point in time, while some states set a 6-month period without health insurance as a requirement for inclusion in the program; furthermore, family income data in available surveys was often categorized in a way that made it difficult to determine whether a child would meet the income eligibility criteria for SCHIP.
Once they had an approved program, states also encountered difficulties in enrolling children in SCHIP as rapidly as they had predicted. A paucity of good information on the characteristics of uninsured children whose family income fell in the range specified by the state made it difficult to target the program to eligible families. During the first three years of the program, most states were unable to enroll as many children as their funding allotment would allow. However, the allotment was based on the unreasonable assumption that each state would have complete enrollment on the day the SCHIP legislation went into effect. Even though states were allowed to spend funds allotted in a given fiscal year over a 3-year period, only 45 percent of the $4 billion allotted in FY 1998 had been spent by the end of FY 2000. This left $1.9 billion to be distributed among the states that had spent all of their FY 1998 allotment (Kenney et al., 2000). The SCHIP program covered between 2 and 3 million previously uninsured children each year during its first three years; even so, there were an estimated 11 to 12 million uninsured low-income children at the end of that time, an unknown number of whom were eligible for SCHIP.
As states increased the numbers of children enrolled in SCHIP, many also documented low retention rates in the program. However, there was uncertainty about whether children were leaving SCHIP because their families had obtained other insurance for them, either Medicaid or private; because of failure to complete the paperwork requirements for reenrollment; or because their families were dissatisfied with the program.
The workshop was designed to provide a forum for substantive interaction among state policy makers and analysts, federal policy makers, and researchers and academics. The presentations provided opportunities for states to share their experiences in making innovative use of data to target enrollment and improve and monitor the success of their programs. Researchers examining SCHIP, often using nationally representative databases, presented their analyses and suggested ways that states might adapt the research methods to better understand state-level SCHIP issues. The work
shop agenda is included in Appendix A, and the presentations are summarized briefly in Appendix B. Appendix C contains a brief description of each of the national surveys mentioned in this report.
Following the meeting, the panel met to summarize the workshop discussions and to draw conclusions about ways to enhance data collection and analysis in order to strengthen and target enrollment and retention efforts and to support evaluation of the SCHIP program. This report contains the panel’s summaries of the workshop discussions around three issues: determining the size of the eligible population (Chapter 2), enrollment (Chapter 3), and disenrollment (Chapter 4). Chapter 5 discusses several cross-cutting themes that affect multiple program goals. At the end of each of Chapters 2-5, the panel lists conclusions focused on the capacity to evaluate and monitor the SCHIP program. Reviewing the SCHIP activities of the states not included in the workshop was beyond the scope of the panel’s charge. The conclusions are those of the panel and do not necessarily reflect the opinions of either the workshop participants or the sponsors of the workshop. Although the panel sometimes based its conclusions directly on suggestions made by workshop participants, the conclusions evolved from subsequent discussion and synthesis of the issues, and some were not explicitly discussed in the workshop itself. The final chapter is a summary of the highlights of the workshop.