CHIPing Away at the Problem of Uninsured Children

Originally published by the Center for Studying Health System Change (HSC). HSC was a nonpartisan policy research organization funded principally by the Robert Wood Johnson Foundation.

James D. Reschovsky, Peter J. Cunningham

The State Children's Health Insurance Program (CHIP), signed into law in August 1997, represented the most ambitious public health coverage expansion in more than 30 years. Its success would hinge on whether state officials could actually locate and enroll the children who qualified. Research from the Center for Studying Health System Change (HSC) revealed that uninsured children were far from a monolithic group, and that effective CHIP implementation would require programs tailored to specific populations and local market realities. This Issue Brief examines why children lacked health insurance and what those patterns meant for CHIP's rollout.

The Scope of CHIP

Established by the Balanced Budget Act of 1997, CHIP gave states block grants to expand coverage for children through one of three approaches: broadening Medicaid eligibility, creating a new program to subsidize private insurance for children, or combining both strategies. Congress authorized $40 billion over ten years to fund these efforts. Every state and the District of Columbia -- except Washington and Wyoming at the time -- had developed or was actively developing an implementation plan.

States had wide latitude in structuring their programs. Twenty chose to expand Medicaid alone, 14 built new programs subsidizing private coverage, and 15 used a combination -- though most of the combination states leaned heavily on new programs, using Medicaid expansion only for specific subgroups. HSC analysis found that 33 percent of all uninsured children would become newly eligible through CHIP.

Whether CHIP would actually reduce the number of uninsured children depended less on whether states chose a public or private insurance model and more on how well they carried out enrollment. HSC's findings made clear that understanding the specific reasons children were uninsured was a prerequisite for effective program design. Though uninsured children came from diverse circumstances, they were predominantly children of working parents -- four out of five had a parent in the labor force. Yet nearly 80 percent lived in families earning below 200 percent of the federal poverty level, according to the HSC Household Survey.

Why Children Were Uninsured

Most American children obtained health coverage either through a parent's employer or through a public program. Children fell through the cracks for four main reasons: their parents had no access to employer-sponsored coverage; parents chose not to enroll their children in an available plan; gaps existed in public insurance eligibility; or parents did not enroll children who actually qualified for public programs.

Limited Access to Employer Plans

The single biggest factor was that parents could not get employer-sponsored insurance, affecting roughly three out of four uninsured children. This happened for multiple reasons: parents were not employed, their employers did not provide health benefits, or they did not meet eligibility requirements for the plan being offered. Among these, the employer not offering coverage at all was the most common, affecting two in five uninsured children. One in five had no working parent in the household. Another 15 percent had parents who were employed but did not qualify for the plan because they had not been on the job long enough or did not work sufficient hours.

Parents Who Turned Down Available Coverage

About 9 percent of uninsured children had parents who were offered employer coverage but declined it. The dominant reason, cited by 84 percent of these respondents, was the cost of premiums, deductibles, copayments, and uncovered services. Only 12 percent said they did not feel they needed insurance.

In other cases, parents had enrolled themselves but did not extend coverage to their children -- accounting for 15 percent of uninsured children. The most likely explanation was cost: while 98 percent of employees were at firms that offered dependent coverage, employers frequently subsidized family plans less generously than employee-only plans, pushing the additional expense onto parents.

Gaps in Public Insurance Eligibility

When Congress enacted CHIP, roughly 5.7 million children -- 65 percent of all uninsured children -- were not eligible for Medicaid. CHIP was designed to close some of that gap, but an estimated 2.7 million children, about 31 percent of all uninsured children, would remain ineligible for any federally supported coverage even after CHIP was fully implemented.

Eligible Children Not Enrolled

Before CHIP, more than one-third of uninsured children -- some 3.1 million -- already qualified for Medicaid but were not enrolled. These children made up about 16 percent of all Medicaid-eligible children. Parents may not have known their children qualified, may not have seen the need, or may have avoided enrollment due to the stigma attached to Medicaid. Administrative hurdles in the eligibility determination process, language barriers, and other obstacles also played a role.

Changes in welfare policy were likely to make this enrollment gap worse. Historically, children became enrolled in Medicaid when their parents signed up for welfare benefits. But as public insurance eligibility expanded beyond the welfare population, and welfare reform reduced the welfare rolls, states needed to find new ways to reach eligible children outside the welfare office.

Wide Variation Across Communities

While often discussed as a national problem, children's uninsurance varied enormously from one community to the next. Across HSC's 12 study sites, children's uninsurance rates ranged from 4.4 percent in Lansing, Michigan, to 22.4 percent in Miami, Florida -- a fivefold difference.

Four factors consistently explained the largest share of variation across communities, listed in order of importance: the proportion of parents with an employer offer of insurance; the share of children from Spanish-speaking families; the percentage of Medicaid-eligible children; and the distribution of family income. These four factors mapped directly onto the four reasons children lacked coverage.

Employer offer rates drove much of the local variation. In Miami, only 54 percent of children had a parent who was offered workplace insurance, compared with 73 percent nationally. If Miami's offer rates had matched the national average, the community's child uninsurance rate would have dropped by four percentage points.

Hispanic children were the demographic group most likely to be uninsured, even after accounting for income, offer rates, and other factors -- with more than one in four lacking coverage. But the data showed that language, not Hispanic ethnicity per se, was the predominant factor. Larger populations of Spanish-speaking families helped explain the elevated uninsurance rates in Miami and Orange County.

State Medicaid eligibility rules also mattered: variations in these rules accounted for up to two percentage points of difference in child uninsurance rates across the 12 sites. And income distribution among lower-income families, particularly the concentration of families in poverty, explained additional variation. Communities like Little Rock and Miami had relatively high child uninsurance rates in part because of their concentrations of poor families who were less likely to enroll in private insurance even when it was available.

Policy Responses for Reaching Uninsured Children

CHIP had the potential to substantially reduce the number of uninsured children in the country, but only if states pursued creative, aggressive strategies to identify and enroll eligible children. The research pointed to several program design features that could help states reach children who fell through coverage gaps due to parental job transitions or other disruptions.

States could direct outreach to unemployment offices, avoid imposing waiting periods for children who became uninsured after a parent lost or changed jobs, and provide 12-month continuous eligibility even when family circumstances changed. Insurance market reforms -- such as limiting waiting periods for new employees and restricting the ability of firms to deny coverage to part-time workers or those with preexisting conditions -- could also help. Portability provisions to ensure continuity of coverage were another tool.

Because employer-based benefits remained unaffordable for many lower-income families, states needed to be careful about imposing premiums, deductibles, and copayments on CHIP participants. Since roughly one-quarter of CHIP-eligible children had parents with access to employer insurance, states could also consider coordinating with private employers or subsidizing employer-sponsored premiums. Under Medicaid, states already had the authority to pay premiums and cost-sharing for beneficiaries with private coverage when doing so was cost-effective, though few had used this option due to administrative complexity.

Many states had enacted or were considering measures to reduce the stigma of public coverage by making Medicaid and CHIP resemble private insurance as closely as possible. Streamlining eligibility determinations was another common strategy.

Outreach to Hispanic Families

Forty percent of uninsured Medicaid-eligible children were Hispanic, and CHIP-eligible children were twice as likely to be Hispanic as the general child population. Outreach directed at neighborhoods with large Hispanic populations and at Hispanic community, religious, and cultural institutions was critical. Sensitivity to language barriers was particularly important: Hispanic Medicaid-eligible children whose parents were not fluent in English were far more likely to be uninsured than those with English-fluent parents. This suggested language was a major enrollment barrier, tied to both awareness and acceptance of public programs -- and potentially to immigration status.

Outreach Through Health Care Settings

Only one uninsured child in five had no usual source of medical care, meaning the vast majority were already connected to the health care system in some form. Emergency rooms, hospital outpatient clinics, and public clinics presumably informed uninsured patients about public coverage options as a matter of course, but it was less clear whether private physicians did the same. About a third of uninsured Medicaid-eligible children used a doctor's office as their regular source of care, and among CHIP-eligible children, that figure rose to 41 percent. These numbers suggested that encouraging physicians in private practice to participate in outreach could meaningfully expand enrollment.

Survey Methodology

Data for this analysis came from the 60-site Household Survey, conducted as part of HSC's Community Tracking Study to monitor changes in the health care system and their effects on consumers and providers. The survey was carried out through telephone interviews and in-person interviews of families in households without a telephone between July 1996 and July 1997. The data file included observations on over 60,000 individuals in nearly 33,000 families, with about 11,600 families having children. Questions covered insurance coverage, health status, access to and use of health care, satisfaction with the health care system, and demographic characteristics. The data were weighted to represent the U.S. population.

Sources and Further Reading

The target population for CHIP included children under age 19 in families with incomes less than 200 percent of the federal poverty level or 50 percentage points above the state's income eligibility levels as of August 22, 1997. Some states expanded eligibility beyond these income levels.

This Issue Brief is adapted from "CHIPing Away at the Problem of Uninsured Children: Why Children Lack Health Insurance and Implications for the New State Children's Health Insurance Programs," Research Report No. 02, published by HSC.