SCHIP, Medicaid Expansions Lead to Shifts in Children's Coverage

SCHIP, Medicaid Expansions Lead to Shifts in Children's Coverage

Issue Brief No. 59
December 2002
Peter J. Cunningham, James D. Reschovsky, Jack Hadley

Recent expansions of the State Children's Health Insurance Program (SCHIP) and Medicaid produced notable shifts in how children are covered. New data from the Center for Studying Health System Change (HSC) reveal that the share of low-income children lacking insurance fell from 20.1 percent in 1997 to 16.1 percent in 2001, driven by substantial growth in public program enrollment. However, the net impact of these coverage gains was tempered by a concurrent decline in private insurance (from 47% in 1997 to 42.3% in 2001). Part of the drop in private coverage stemmed from the substitution of public insurance for private plans.

An Ambitious Effort to Cover Kids

Congress created SCHIP in 1997 with the goal of lowering the number of low-income children who lacked health insurance, particularly those in families without access to employer-sponsored coverage or Medicaid. Under the program, states had the option to broaden eligibility within their existing Medicaid programs or to set up separate child health insurance programs. To limit the possibility that privately insured children might migrate to SCHIP and Medicaid, Congress mandated that states implement strategies to prevent those already covered by private insurance from enrolling. A number of states imposed waiting periods (generally three to six months without insurance) before permitting enrollment; others gathered data on applicants' current and prior insurance status; and some introduced cost-sharing requirements higher than what is typical for public insurance serving low-income populations.

States also employed a range of strategies to boost enrollment among children who qualified, such as lowering administrative hurdles, streamlining application procedures, and launching or expanding outreach campaigns to publicize the program and motivate parents to sign up their eligible children.

By late 2001, enrollment across all SCHIP-related programs had reached 3.5 million children. Additionally, Medicaid enrollment among those eligible under pre-SCHIP guidelines grew as well, partly due to enhanced outreach and simplified enrollment procedures.

Fewer Uninsured, But Private Coverage Declines, Too

Findings from the Community Tracking Study (CTS) Household Survey spanning 1997 to 2001 show that among children in families earning below 200 percent of the federal poverty level (roughly $36,000 for a family of four in 2001), the proportion enrolled in SCHIP or Medicaid rose by nearly eight percentage points -- from 28.4 percent in 1997 (before SCHIP was enacted) to 36 percent in 2001. This expansion in public coverage drew from both the uninsured population and those who previously held private insurance. The share of low-income children without any insurance fell four percentage points, from 20.1 percent in 1997 to 16.1 percent in 2001, while the share with private coverage dropped 4.7 percentage points, from 47 percent to 42.3 percent over the same period. The most pronounced coverage changes occurred among children in families with incomes between 100 percent and 200 percent of poverty -- the core SCHIP target population.

Table 1: Health Insurance Coverage, Children Age 19 and Under

All Children Age 19 and Under: Private Insurance: 70.8% (1997), 69.3% (1999), 69.9% (2001), change -0.9%. SCHIP/Medicaid: 14.2 (1997), 15.4 (1999), 16.8 (2001), change 2.6#. Other: 3.6 (1997), 3.8 (1999), 3.9 (2001), change 0.3. Uninsured: 11.5 (1997), 11.5 (1999), 9.4* (2001), change -2.1#.

200% of Poverty or Higher: Private Insurance: 89.3 (1997), 89.8 (1999), 86.0* (2001), change -3.3#. SCHIP/Medicaid: 3.1 (1997), 2.7 (1999), 5.6* (2001), change 2.5#. Other: 2.9 (1997), 2.6 (1999), 2.8 (2001), change -0.1. Uninsured: 4.7 (1997), 4.9 (1999), 5.5 (2001), change 0.8.

Less Than 200% of Poverty: Private Insurance: 47.0 (1997), 41.5* (1999), 42.3 (2001), change -4.7#. SCHIP/Medicaid: 28.4 (1997), 32.6* (1999), 36.0 (2001), change 7.6#. Other: 4.6 (1997), 5.5 (1999), 5.7 (2001), change 1.1. Uninsured: 20.1 (1997), 20.5 (1999), 16.1* (2001), change -4.0#.

Between 100-200% of Poverty: Private Insurance: 63.8 (1997), 56.2* (1999), 57.1 (2001), change -6.7#. SCHIP/Medicaid: 13.4 (1997), 20.8* (1999), 24.2 (2001), change 10.8#. Other: 4.0 (1997), 5.1 (1999), 5.8 (2001), change 1.8#. Uninsured: 18.9 (1997), 17.9 (1999), 13.0* (2001), change -5.9#.

Less Than 100% of Poverty: Private Insurance: 25.5 (1997), 23.0 (1999), 23.6 (2001), change -1.9. SCHIP/Medicaid: 47.6 (1997), 47.3 (1999), 50.9 (2001), change 3.3. Other: 5.4 (1997), 5.9 (1999), 5.5 (2001), change 0.1. Uninsured: 21.5 (1997), 23.7 (1999), 20.0 (2001), change -1.5.

Notes: "Other" encompasses military insurance, Indian Health Services, Medicare, and other public programs. * Change from previous survey is statistically significant at p<.05 level. # Change from 1997 to 2001 is statistically significant at p<.05 level. Source: HSC Community Tracking Study Household Survey.

Coverage Changes Vary by State

The coverage expansions resulting from SCHIP differed significantly from one state to another. For instance, some states were already insuring large numbers of low-income children through Medicaid and other state-operated programs, so they did not need to widen eligibility as much as others. Thirteen states plus the District of Columbia (accounting for roughly 40% of low-income children) implemented substantial eligibility expansions (defined as an increase of 50 percentage points or more in the share of eligible low-income children), while six states (representing about 16% of low-income children) made no eligibility changes.

States that broadened eligibility by 50 percentage points or more experienced the most dramatic shifts in public and private coverage rates. In those states, the proportion of low-income children with SCHIP or Medicaid coverage surged nearly 14 percentage points, from 24.5 percent in 1997 to 38.3 percent in 2001. By contrast, states with more modest or no eligibility expansions saw public coverage increase by only about three percentage points -- a change that was not statistically significant. Nearly all of the decrease in private insurance among low-income children between 1997 and 2001 was concentrated in states with the largest eligibility expansions.

By comparison, the proportion of low-income children who were uninsured fell significantly in both states with small eligibility expansions and those with larger ones. Although the magnitude of the drop was somewhat greater in states with the biggest expansions (5.3 percentage points versus 3.4 percentage points for states with smaller or no changes), that gap is much narrower than the differential observed for shifts in public and private coverage.

Table 2: Changes in SCHIP/Medicaid Coverage Among Low-Income Children Age 19 and Under

Percent with SCHIP/Medicaid: 28.4% (1997), 36.0% (2001), change 7.6%*. In states with large increase in eligibility: 24.5 (1997), 38.3 (2001), change 13.8*. In states with smaller or no increase: 30.9 (1997), 34.1 (2001), change 3.2.

Percent with Private Insurance: 47.0 (1997), 42.3 (2001), change -4.7*. In states with large increase in eligibility: 46.1 (1997), 37.0 (2001), change -9.1*. In states with smaller or no increase: 47.5 (1997), 46.4 (2001), change -1.1.

Percent with Other Coverage: 4.6 (1997), 5.7 (2001), change 1.1. In states with large increase in eligibility: 5.3 (1997), 5.9 (2001), change 0.6. In states with smaller or no increase: 4.2 (1997), 5.5 (2001), change 1.3.

Percent Uninsured: 20.1 (1997), 16.1 (2001), change -4.0*. In states with large increase in eligibility: 24.1 (1997), 18.8 (2001), change -5.3*. In states with smaller or no increase: 17.4 (1997), 14.0 (2001), change -3.4*.

Notes: Large increase in eligibility is defined as an increase of 50 percentage points or higher in the percent of low-income children eligible for SCHIP/Medicaid coverage, when state rules are applied to a standardized population of low-income children. * Change is statistically significant at p<.05 level. Source: HSC Community Tracking Study Household Survey.

Reaching Out to Eligible Families

The fact that the rate of uninsured children declined even in regions with minimal or no expansion in eligibility indicates that additional factors -- particularly outreach initiatives to increase participation and reduce bureaucratic obstacles -- also played a role. Many parents of uninsured children who qualify for SCHIP or Medicaid are unaware the programs exist, do not think their children are eligible, lack interest, or are put off by burdensome enrollment procedures.

States undertook significant efforts to connect with families whose children might be eligible for SCHIP and to lower administrative barriers to enrollment. Federal and state governments devoted substantial resources to advertising campaigns, websites, and toll-free phone lines to promote sign-ups. States collaborated with schools, healthcare providers, private employers, and social service organizations to identify eligible children and encourage their parents to complete enrollment.

Because enrollment processes had frequently been cited as obstacles to Medicaid participation, most states also worked to simplify their procedures for both SCHIP and Medicaid. Common reforms included shortening application forms, eliminating the requirement for in-person interviews or asset tests, and permitting presumptive eligibility -- granting temporary coverage before a formal determination so that children could receive healthcare services immediately.

Anecdotal evidence and case study results suggest these activities are successfully boosting participation among eligible children. Greater participation may also help explain why SCHIP enrollment nearly doubled between 2000 and 2001, despite the fact that most major eligibility expansions had taken place before 2000. CTS data show that participation rates among low-income children eligible for SCHIP or Medicaid increased from 60 percent in 1999 to 66 percent in 2001, with particularly large gains in communities that had the highest rates of uninsured children.

Coverage Expansions Result in Some Substitution

The expansion of SCHIP also led to some displacement of private coverage by public insurance, a phenomenon sometimes called crowd out. Substitution occurs when children who enroll in SCHIP and Medicaid would have been covered by private insurance in the absence of the public program expansions. This can happen when parents take advantage of free or lower-cost public coverage by directly switching their children from private to public plans (a practice states were required to try to prevent). But substitution can also unfold indirectly over time, as the availability of public coverage creates additional pathways for children whose family economic circumstances change.

A multivariate analysis of CTS data found that roughly one-quarter of the growth in public coverage among children in families earning less than 200 percent of poverty between 1997 and 2001 involved replacement of private insurance with public coverage. Among children in families with incomes between 100 percent and 200 percent of poverty -- the primary SCHIP target group -- approximately 39 percent of the increase in SCHIP or Medicaid enrollment reflected substitution.

These figures are broadly consistent with earlier estimates of substitution that occurred when Medicaid eligibility was broadened in the late 1980s and early 1990s (though estimates from that era vary widely due to different data sources and analytical methods). Given the considerably higher rates of private coverage in the SCHIP target population, one might have expected substitution to be greater than during the earlier Medicaid expansions. However, substitution under SCHIP may also have been held in check because states were required to adopt explicit measures to prevent direct switching from private to public coverage. While some states merely collected data on the extent of substitution with an implicit commitment to act if it proved significant, others imposed concrete safeguards -- most commonly requiring children to be uninsured for a specified period (typically three to six months) before enrolling, or collecting information about prior insurance coverage.

While the findings confirm that some substitution of public for private coverage did take place, this does not necessarily mean that the safeguards designed to prevent direct switching were ineffective. The substitution observed over the four-year window captured by the CTS surveys may involve far more complex dynamics than simple plan-switching.

Research has shown that movement between different types of insurance is far more fluid than periodic surveys can capture. For example, some children experience temporary gaps in coverage or brief Medicaid enrollment when a parent loses a job. Without SCHIP, many of these children might eventually have returned to private insurance once their parents found new or better employment or purchased individual coverage -- but they stay enrolled in SCHIP instead. Most existing crowd-out protections do not address this indirect form of substitution.

State Budget Cuts Could Imperil Coverage Gains

States have begun to realize the promise of SCHIP in reducing the ranks of uninsured children. Although the expansion of public coverage displaced some private insurance, more recent trends indicate the program is also bringing down the number of uninsured children. Nevertheless, the sluggish national economy, escalating private insurance costs, and widening state budget shortfalls threaten to stall further progress or even reverse the gains already achieved.

Confronting growing deficits and expanding Medicaid budgets, most states prioritized cost containment measures, including controls on prescription drug spending, reductions or freezes in provider reimbursement, benefit cuts, higher beneficiary copayments, and restrictions on Medicaid eligibility. SCHIP largely escaped direct cuts to eligibility or benefits, though some states scaled back their outreach activities.

Increasing unemployment and premium hikes will reduce the availability and affordability of private insurance for many parents of low-income children. SCHIP and Medicaid serve as a crucial safety net for children who lose private coverage when their parents become unemployed or when families can no longer keep up with the rising cost of private insurance. Consequently, any eligibility reductions driven by state fiscal pressures will place more children at risk of going without coverage entirely.

Data Source and Methods

This Issue Brief draws on findings from the HSC Community Tracking Study Household Survey, a nationally representative telephone survey of the civilian, noninstitutionalized population conducted in 1996-97, 1998-99, and 2000-01. For clarity, single calendar years (1997, 1999, and 2001) are used in the discussion. Data were supplemented with in-person interviews of households without telephones to ensure proper representation. Each survey round collected information on roughly 60,000 individuals, including more than 10,000 children. Response rates across the three rounds ranged from 59 percent to 65 percent.

Additional details on survey methodology can be found at www.hschange.org.

Notes

1. Smith, Vernon K., and David M. Rousseau, SCHIP Program Enrollment: December 2001 Update, Kaiser Commission on Medicaid and the Uninsured (July 2002).

2. Bruen, Brian K., and John Holahan, Acceleration of Medicaid Spending Reflects Mounting Pressures, Kaiser Commission on Medicaid and the Uninsured (May 2002); Rosenbach, Margo, et al., Implementation of the State Children's Health Insurance Program: Momentum Is Increasing After a Modest Start, Mathematica Policy Research, Inc. (January 2001).

3. The size of the eligibility expansions in each state in the CTS is determined by computing the percent of low-income children eligible for SCHIP/Medicaid coverage for 1997 and 2001. This is done by applying the state- and year-specific eligibility criteria to a standardized population of low-income children based on the 1996-97 CTS Household Survey. Using a standardized population to compute eligibility holds constant differences in population characteristics across states and over time, which could result in some of the variation in eligibility being due to factors other than the differences and/or changes in eligibility standards.

4. A multivariate analysis that also controls for individual characteristics, changes in health insurance costs and other market factors confirms that increases in eligibility for SCHIP/Medicaid coverage directly led to a decrease in private insurance coverage relative to SCHIP/Medicaid coverage between 1997 and 2001. A multivariate analysis of the effects of eligibility expansions on changes in coverage between 1997 and 1999 can be found in Cunningham, Peter J., Jack Hadley and James Reschovsky, "The Effects of SCHIP on Children's Health Insurance Coverage: Early Evidence from the Community Tracking Study," Medical Care Research and Review, Vol. 59, No. 4 (December 2002). These results, updated to include data for 2000-01, can be found at www.hschange.org.

5. Kenney, Genevieve, and Jennifer Haley, "Why Aren't More Uninsured Children Enrolled in Medicaid or SCHIP?" New Federalism: National Survey of America's Families, No. B-35, Urban Institute (May 2001).

6. Felland, Laurie E., and Andrea M. Benoit, Communities Play Key Role in Extending Public Health Insurance to Children, Issue Brief No. 44, Center for Studying Health System Change, Washington, D.C. (October 2001).

7. Smith and Rousseau, op. cit.

8. Cunningham, Peter J., "SCHIP Making Progress: Participation Increases as Children's Uninsurance Declines," Working Paper, Center for Studying Health System Change, Washington, D.C. (July 2002).

9. Estimates of substitution were obtained from multivariate regression analyses of the effects of eligibility increases on changes in coverage between 1997 and 2001, while also controlling for changes in individual characteristics, health insurance costs and other factors that may be associated with coverage changes. The regression results were used to simulate coverage rates for 2000-01, assuming that eligibility remained at 1996-97 levels. The simulations were then used to compute the proportion of the total increase in SCHIP/Medicaid coverage between 1997 and 2001 that is due to the SCHIP/Medicaid-related decrease in private insurance coverage. More detail on the methodology used to derive estimates of substitution is provided in Cunningham, Hadley and Reschovsky, op. cit.

10. Dubay, Lisa, Expansions in Public Health Insurance and Crowd-Out: What the Evidence Says, The Henry J. Kaiser Family Foundation (1999).

11. Bennefield, Robert L., "Who Loses Coverage and for How Long?" Current Population Reports, U.S. Census Bureau (July 1998).

12. Bruen and Holahan, op. cit.

13. Smith, Vernon, and Victoria Wachino, Medicaid Spending Growth: Results from a 2002 Survey, Kaiser Commission on Medicaid and the Uninsured (September 2002).

14. Howell, Embry, Ian Hill and Heidi Kapustka, SCHIP Dodges the First Budget Ax, Health Policy Online, No. 3, Urban Institute (September 2002).

Issue Briefs are published by the Center for Studying Health System Change. President: Paul B. Ginsburg. Director of Public Affairs: Richard Sorian. Editor: The Stein Group.

Sources and Further Reading

Centers for Medicare & Medicaid Services — CHIP Overview — Official CMS resource on the Children's Health Insurance Program, including eligibility, enrollment data, and program structure.

Kaiser Family Foundation — Children's Health Coverage Through Medicaid and CHIP — KFF analysis of children's coverage trends under Medicaid and CHIP, including data on enrollment shifts and crowd-out effects.

U.S. Census Bureau — Health Insurance Coverage Statistics — Federal data on health insurance coverage trends, including children's uninsured rates referenced in this research.

Commonwealth Fund — SCHIP Reauthorization and Coverage Trends — Research examining the policy debate around SCHIP reauthorization, including crowd-out concerns and coverage substitution dynamics.

Robert Wood Johnson Foundation — HSC Research on Children's Coverage Shifts — RWJF-funded HSC research on how SCHIP and Medicaid expansions affected the balance between public and private children's health coverage.