Dennis G. Smith
Center for Medicaid & State Operations
Centers for Medicare & Medicaid Services
Covering Uninsured Kids: Missed Opportunities for Moving Forward
Committee on Energy & Commerce
Subcommittee on Health
U.S. House of Representatives
Tuesday January 29, 2008
Chairman Pallone, Congressman Deal, thank you for inviting me to testify on today’s topic as you renew the important work of reauthorizing the State Children’s Health Insurance Program (SCHIP). The Administration strongly supports this important program and its full reauthorization. Last year, additional funding for the program was provided to ensure stability in the program through March 2009. We look forward to working with all members during this time to achieve the goal of reauthorization through 2013.
The full picture of our commitment to insuring low-income children includes Medicaid as well as SCHIP. Medicaid is approximately four times larger than SCHIP in terms of enrollment of children and just over six times larger in terms of expenditures for children. Total Federal and State Medicaid spending on children will exceed $400 billion over the next five years and $1 trillion over the next ten years. There are important budgetary and programmatic interactions between SCHIP and Medicaid that are appropriate to consider in the context of reauthorization.
When Congress was considering the legislation that became Title XXI more than ten years ago, there was a widely held view that 10 million children in the United States lacked health insurance. It was recognized that many of these children were already eligible for Medicaid but were not enrolled, and that many of these children were uninsured but lived in families with sufficient income to be able to afford coverage. Congress ultimately adopted an approach that was targeted to children with family incomes above existing Medicaid levels who lived in families for which the cost of insurance was beyond their reach. It set a general upper limit of income eligibility at the higher of 200 percent of the federal poverty level (FPL) or 50 percentage points above a state’s Medicaid level. Under the FPL guidelines released last week for 2008, 200 percent of FPL is $42,400 for a family of four and 250 percent of FPL is $53,000 for a family of four. Just by way of comparison: the median income in the United States for a family of four is approximately $59,000.
SCHIP is a unique compound of incentives and checks and balances. Congress rejected the idea of simply re-creating Medicaid and its complexities. States with an approved SCHIP plan are eligible for Federal matching payments drawn from a state-specific capped allotment. While the program provides states with a great deal of program flexibility, including using Medicaid as their vehicle for administering Title XXI, it also creates the expectation that states will adopt policies to stay within their capped allotments. Capped appropriations and capped allotments were critical features of that bipartisan compromise. The legislation appropriated $40 billion over ten years, an amount that would support the number of children thought to be in the target population group. That level of funding clearly was not designed or intended to serve children at all income levels, nor was it intended to create a new entitlement for coverage.
Congress also realized that millions of children were eligible for Medicaid but were not enrolled. To ensure the success of SCHIP and avoid the possibility of creating a new program that would not be taken up by the states, the idea of an enhanced match rate was ultimately adopted as the means of providing states with sufficient incentive to aggressively find and enroll uninsured low-income children. Thus, SCHIP provides a 70 percent federal match rate on an average national basis compared to the 57 percent average match rate for Medicaid. But central to the bipartisan discussion at that time was the question, “for whom is the enhanced match intended?” That question remains central to reauthorization today.
Enrollment Exceeds Expectations
If the goal ten years ago was to enroll 10 million children, then expectations have been exceeded. In 1998, the number of children “ever-enrolled” in Medicaid (enrolled at least for some period of time) was 19.6 million. States enrolled approximately 670,000 children in SCHIP in that first year for a combined total of more than 20 million children. Since then, combined Medicaid and SCHIP enrollment has increased every year. In FY 2006, more than 36 million children were enrolled (at least for some period of time) in Medicaid and SCHIP combined, an increase of 16 million children above the 1998 Medicaid level.
Since 1998, enrollment of children in SCHIP and Medicaid has increased nearly 80 percent, while growth in the total number of children in the U.S. population as well as the number of children in families below 200 percent FPL over the same period has been nominal. Enrollment in Medicaid and SCHIP now exceeds the number of children below 200 percent FPL. Therefore, it is clear that Medicaid and SCHIP are covering children in higher-income families.
“95 Percent Enrollment Goal”
It is because of this tremendous growth in Medicaid and SCHIP enrollment relative to the overall population and to the low-income population specifically that we believe our adopted goal of 95 percent enrollment of low-income children before expanding eligibility to higher income populations is both reasonable, in light of the statutory purpose of SCHIP to serve low-income children, and is achievable.
We anticipate working with states to determine their specific rates of coverage. It is unfortunate that some groups have prejudged compliance as they have relied on flawed national data to make comparisons regarding state performance. For example, it is widely recognized that the Current Population Survey (CPS) undercounts Medicaid participation. In the most recent CPS data released last year, the Census Bureau reported 20.7 million children ever enrolled in FY 2006, when enrollment reported by states for Medicaid and SCHIP combined in that same period was over 36 million.
We believe the 95 percent goal is further supported by last year’s work conducted by the Urban Institute which shows much lower uninsurance rates among Medicaid and SCHIP eligible children than expected. This study was not unanimously received as good news at the time, but we believe it demonstrates that states are far more successful than given credit. Therefore the 95 percent goal is not only achievable, but should be expected and demanded. Indeed, our view is that a number of states are already meeting the 95 percent goal.
We strongly believe, as the future of SCHIP as a program is considered, that states be required to put poor children first before they expand to higher income levels. The federal government has tied financial incentives to performance standards in other public benefits programs with good results.
I want to reaffirm our previously stated position that children currently enrolled in SCHIP should not be affected as we work with states to implement the August 17, 2007 State Health Official (SHO) letter. The guidance sets out procedures and assurances that should be in place when states enroll new applicants with family incomes in excess of 250 percent of the federal poverty level (FPL) – that is, in excess of the median family income in the United States. But the guidance is not intended to affect enrollment, procedures, or other terms for such individuals currently enrolled in State programs.
The goal of SCHIP is to increase the rate of insurance among our nation’s children in low-income families. “Crowd-out” or the substitution of existing coverage does not increase insurance rates, it merely shifts the source of funding. It is a public policy concern because it increases public expenditures without necessarily improving access to care or health status. It is also a concern because, as healthy lives are shifted out of the private sector insurance pools, there is a detrimental impact on those who remain. Insurance fundamentally means the sharing of risk. When the private pool of healthy insured lives shrinks and the risk cannot be spread as widely as before, the cost will rise for those who remain, triggering another cost increase which is likely to displace yet another group of people, whether employers or employees or both.
Crowd-out is not a new topic. There were numerous papers written on Medicaid and crowd-out prior to the enactment of SCHIP and it remains a popular subject today. The pre-SCHIP papers on crowd-out dealt primarily with populations below 200 percent of FPL, many of whom were assumed to not have access to employer-sponsored health insurance or the means to contribute the employee share of costs. There are a variety of opinions on how to define crowd-out, how to measure it, and how to prevent it. In its paper on SCHIP last May, the Congressional Budget Office (CBO) neatly summarized the research on this topic and concluded that, “… in general, expanding the program to children in higher-income families is likely to generate more of an offsetting reduction in private coverage (and therefore less of a net reduction in uninsurance) than expanding the program to more children in low-income families.” The CBO estimates on the SCHIP legislation that the President vetoed reinforce the findings of its May study.
As early as February 1998, the federal government released instructions to the states on how it would review strategies to protect against substitution of private coverage. In a February 13, 1998 State Health Official letter, co-signed by the Director of the Center for Medicaid and State Operations at the Health Care Financing Administration and the Acting Administrator of the Health Resources and Services Administration, the federal government provided that, “States that provide insurance coverage through a children’s only and/or a State plan (as opposed to subsidizing employer-sponsored coverage) or expand through Medicaid will be required to describe procedures in their State CHIP plans that reduce the potential for substitution. … After a reasonable period of time, the Department will review States’ procedures to limit substitution. If this review shows they have not adequately addressed substitution, the Department may require States to alter their plans.”
Another federal agency within the Department of Health and Human Services, the Agency for Healthcare Research and Quality, listed several strategies to prevent crowd-out at that time which included:
- Institute waiting periods (3, 6, or 12 months)
- Limit eligibility to uninsured or under-insured
- Subsidize employer-based coverage
- Impose premium contributions for families above 150 percent of the Federal poverty level
- Set premiums and coverage and levels comparable to employer-sponsored coverage
- Monitor crowd-out and implement prevention strategies if crowd-out becomes a problem
States faced competing pressures as they designed their SCHIP programs. Effective crowd-out strategies were measured against pressures to quickly build enrollment. Decision makers at the state level faced strong public criticism for “turning back” federal funds that would go to other states or be returned to the Federal Treasury.
As the 16 million children were being added to Medicaid and SCHIP, the percent of children between 100 and 200 percent of poverty with private insurance declined. In 1997 according to data from the 2006 National Health Interview Survey, 55 percent of children in families with income at this level had private insurance. But by 2006, the percentage had declined to 36 percent. 
Currently there are 20 jurisdictions (19 states and the District of Columbia) that cover children in families with income greater than 200 percent of FPL, of which 17 jurisdictions cover children in families with income equal to or greater than 250 percent FPL. In addition, there are three states that cover children in families with income thresholds above 200 percent of FPL that apply income disregards in an amount we believe is likely to exceed the 250 percent FPL threshold. Expansions of SCHIP to higher income levels occurred early in the program or just in the past two years. Of the 19 states and the District of Columbia that provide coverage above 200 percent of the poverty level, 13 of them received approval to cover those higher incomes by July 2001 or earlier. Of those 13 states, eight were “qualifying states,” that had increased Medicaid eligibility prior to the creation of SCHIP.
The other seven states that have expanded eligibility above 200 percent FPL occurred in January 2006 or later. With the exception of Hawaii, the eligibility limits were approved as state plan amendments, not as waivers as has been widely reported. After a five-year period in which no state raised their eligibility level, there clearly are growing interests or pressures among additional states to expand eligibility beyond the statutory definition. It is important to understand those interests or pressures in order to design an appropriate response.
Federal responses may be different than the choices made ten years ago and should include approaches outside of SCHIP as well as within the program. One area that seems particularly ripe for a new approach within SCHIP is premium assistance. Perhaps some of the crowd-out effect could have been prevented if SCHIP were used to a greater extent to support private coverage rather than replace it.
SCHIP has been highly successful in the mission it was given to increase coverage among uninsured low-income children. But that success does not mean SCHIP can or will be as successful when populations at higher incomes are involved.
We hope that the lessons of the past will guide how we use the fresh opportunity before us and the Administration looks forward to working with all members to forge reauthorization in the same bipartisan spirit in which SCHIP was created.
 “Eligible But Not Enrolled: How SCHIP Reauthorization Can Help,” September 24, 2007 [available at http://www.urban.org/publications/411549.html].
See http://www.cdc.gov/nchs/data/nhis/earlyrelease/insur200712.pdf. The data are derived from the Family Core component of the 1997–2007 NHIS, which collects information on all family members in each household. Data analyses for the January – June 2007 NHIS were based on 41,823 persons in the Family Core.
Last revised: June 18, 2013