The Basic Features of CHIP

In an article published at the beginning of this year, Sara Rosenbaum and her colleagues at the Center for Health Policy Research identified some of the key factors that influenced how the new program is configured:²

CHIP is a reflection of numerous political and policy themes: health policymakers’ concerns about the continuing problem of uninsured children . . ; belief on the part of the Clinton Administration … and Congress that there should be at least some federal response to the problem of health coverage affordability; states’ strong desire for flexibility in the coverage of children; child advocates’ demands for minimum legislative protections; and observers’ concerns about the “crowd-out” effects of government insurance on private coverage. … The final legislation is an attempt to blend all of these issues and concerns into one program; the result is an unusual statute that is far more complex than it first appears to be.

States are not required to participate in the CHIP program. (Indeed, there is no mandate that they participate in Medicaid itself and, for many years, Arizona did not have a Medicaid program.) But, for the first round of CHIP funding, only four states—Alaska, Hawaii, Washington and Wyoming—have not yet submitted CHIP plans. (A fifth, Vermont, withdrew the plan it initially submitted.) The decision of 45 states and the District of Columbia to participate in CHIP represents a tacit acceptance on their part of the terms of state participation under CHIP.

CHIP offers states the option of covering uninsured children with family incomes up to 200 percent of the federal poverty level (FPL), using capped federal funds with a significantly higher federal matching rate. For states whose Medicaid eligibility levels already exceed 150 percent of the FPL, eligibility limits under CHIP may exceed the current Medicaid levels by as much as 50 percentage points. As a result, some states have chosen CHIP eligibility standards as high as 300 percent of the FPL. (See appendix on page 12 for the dollar amounts that various percentages of the federal poverty level represent.) In addition, the CHIP statute offers the states some latitude in how they define income for eligibility purposes; in other words, by disregarding certain specified income, states may make CHIP benefits available to an even wider population of children.

The state matching rate under CHIP is 30 percent less than the historic state Medicaid rate. Thus, if a state’s current rate is 50 percent under Medicaid, it drops to 35 percent under CHIP. However, the federal matching rate for CHIP is capped at 85 percent of total program costs; thus, state matching requirements range between 15 and 35 percent (as contrasted with 23 to 50 percent for Medicaid).

CHIP also offers each state a choice of three basic approaches to structuring its program. It may (a) create a new or enlarge an existing state children’s health insurance program, (b) expand its current Medicaid program by raising eligibility limits, or (c) choose a combination of the two approaches. If it chooses to create or enlarge a state program, it can cap or limit enrollment—so that everyone who meets eligibility criteria need not be served—but it is required to cover certain minimum benefits as defined in specified “benchmark” plans or solicit the approval of the secretary of health and human services for whatever alternative it offers. (The secretary must assure that any alternative offers comprehensive coverage generally comparable to benchmark plans.) If a state opts to operate CHIP as a stand-alone program independent of Medicaid, it has much greater flexibility to impose certain eligibility criteria (such as those related to age, length of residency in the state, and disability status) than is permitted under the Medicaid program. If it opts instead to expand Medicaid, it must offer new enrollees the same benefits as current Medicaid beneficiaries and is obligated to cover all children who satisfy its new eligibility levels; in essence, the state would be augmenting the individual entitlement under Medicaid.

CHIP funds cannot be used to cover children who are eligible for but not enrolled in Medicaid under the eligibility provisions a state had in place as of March 31, 1997. Accordingly, children must be screened for Medicaid eligibility under those pre-CHIP criteria; if they are found to be eligible under the old Medicaid standards, they must be enrolled in Medicaid—but at the lower, non-CHIP federal match rate applicable under Medicaid. States are also bound by maintenance of effort requirements, so that, if they opt to expand Medicaid under CHIP, they may not impose income and resource eligibility limits more restrictive than those used as of June 1, 1997. The BBA stipulates that CHIP is to be the payer of last resort with respect to all third-party payers, except for the Indian Health Service.

Under CHIP, states may impose cost-sharing requirements in the form of premiums and copayments. However, if they opt for Medicaid expansion under CHIP, they are bound by existing Medicaid limits on cost-sharing, which tend to be fairly restrictive. Yet there are also limits on the cost-sharing that can be implemented under a non-Medicaid CHIP program. Furthermore, any revenues accruing from cost-sharing must be used to offset or reduce matchable federal spending under CHIP.

The BBA indicates that CHIP is not meant to be a new entitlement program for individual children, but one for the states. (However, if the states opt to expand Medicaid as their approach to using CHIP funds, they in effect extend an existing individual entitlement to more of their residents.) This basic message about the non-entitlement is reinforced in the funding levels authorized for the program.

For fiscal year (FY) 1998 through FY 2001, Congress authorized a total of $4.275 billion annually for CHIP, which declines to $3.150 billion for FY 2002 through FY 2004. Authorized funding rises again to $4.050 billion in FY 2005 and FY 2006, and $5.000 billion in FY 2007.

The BBA allows individual states to roll over their allotments for up to two additional fiscal years, to some extent cushioning the decline in federal funding. But as Cindy Mann of the Center on Budget and Policy Priorities pointed out in an analysis published last fall:

Even without accounting for the effects of health care inflation and other factors that would be likely to cause state health care costs to rise over time, the total amount of funds allocated to states under the child health block grants will be sliced almost 26 percent in 2002 and remain at that reduced level until 2005.³

It is also important to recognize that, since the program is not an entitlement, CHIP funding is not open-ended. The enhanced federal matching rate for state expenditures under CHIP is available only up to the limit of the state CHIP allocation. Basically, the only way that a state may secure additional federal funding beyond its CHIP allocation is to opt at the outset for an expansion of its Medicaid program to cover additional children; under this option, once a state’s CHIP allotment is exhausted, it may draw down additional federal funds through the traditional Medicaid program, but at the lower federal matching rate available under Medicaid.

State CHIP allotments for FY 1998 through FY 2000 are based on the state’s share of the nation’s uninsured children in families with incomes below 200 percent of poverty, adjusted using a state cost factor related to average annual wages per employee in the health services industry. In FY 2001, the allotment formula shifts to one that weights the number of uninsured children at 75 percent and introduces a simple count of low-income children, weighted at 25 percent. Beginning in FY 2002, the formula gives equal weight (50–50) to the number of low-income uninsured children and the number of low-income children.

As pointed out in a recent memorandum by Patrick Purcell of the Congressional Research Service, this shift in the allocation formula has significant regional implications:⁴

The distribution of low-income uninsured children and all low-income children in the United States are very different. … Together, the states in the South and West accounted for 72% of all uninsured children in the United States with family incomes of 200% of poverty or less over the period from 1994 to 1996, while the Northeast and Midwest had 28% of the nation’s low-income uninsured children. In contrast, among all children with family incomes at or below 200% of poverty, 59% resided in the South and West, while 41% lived in the states of the Northeast and Midwest. Consequently, when the formula for allocating federal matching funds for S-CHIP begins to change in 2001 … more funds will be allocated to states in the Northeast and Midwest and fewer federal matching dollars will be allocated to states in the South and West.

Possible Structural Deficiencies in the BBA’s CHIP Provisions

A recent analysis by Frank Ullman, Brian Bruen, and John Holahan of the Urban Institute suggested that if states were able to use the maximum allotments available to them, about six million children might be covered under CHIP, yielding a dramatic reduction in the number of uninsured children.⁵ The authors identified four major reasons why states will not be able to realize this full potential:

• First, they noted that a significant number of uninsured children in low-income households—between 1.6 million and 4.7 million, depending on the source of the data—are already eligible for but not enrolled in Medicaid. The BBA explicitly prohibits the use of CHIP funds to cover them.
• Second, they estimated that there are only 2.9 million uninsured children with family incomes below 200 percent of the FPL who are actually eligible for CHIP. An additional 300,000 children might be covered in states that could offer CHIP coverage to families with incomes over 200 percent of the FPL. The authors concluded: “These results suggest that states could have difficulty spending their CHIP funds under current program rules.”
• Third, they found that about 3.0 million uninsured children live in families with incomes too high to qualify for CHIP. They suggested that some of them might be made eligible if states chose to exclude part of their families’ income and assets in determining eligibility or if Congress raised CHIP eligibility levels, should a significant amount of CHIP funds go unused.
• Fourth, they questioned the equity of the distribution of federal funds among the states:

  States that currently have broad levels of coverage for children will receive fewer federal dollars per child than if federal funds were allocated simply on the basis of the state’s share of low-income children. Meanwhile, states that have not already enacted comprehensive coverage for children get more federal money per poor child under the CHIP allocation formula.

They also highlighted a flaw in the CHIP matching rates:

The program makes a greater change in incentives to add coverage in higher-income states than in lower-income states. The percentage point reductions in state matching requirements [in comparison to regular Medicaid matching rates] are greater in a high-income state than in a low-income state, e.g., greater in Connecticut than in Mississippi. Low-income states already had high federal matching rates and the enhanced matching rates are only slightly better. … Whether they will respond to a small change in incentives is unclear.

This Urban Institute report evoked an immediate, strongly worded response from the Children’s Defense Fund (CDF). In addition to disputing elements of the methodology used by the Urban Institute researchers, the CDF response questioned whether the report was not too narrowly focused on the early years of CHIP, especially FY 1998. CDF pointed out that CBO had essentially already reached a similar conclusion about early underutilization of CHIP funds by the states. However, CBO noted:⁶

States will need to manage their programs carefully and take advantage of the option to roll over part of any year’s allocation for up to two succeeding years. But because of the start-up time necessary for states to develop their programs, submit plans to HCFA, and have those plans approved, most states will probably not be able to spend their full allotments for the first two years of the program anyway. The slow start in effect provides an automatic cushion for the leaner years of the program.

CDF went on to declare: “CHIP’s funding formula is tailored to CHIP’s purpose: covering uninsured children. CHIP funding levels are highest in the states with the largest proportion of uninsured children.” It seems clear that Congress intended to target CHIP funds to uninsured children, at least in the first years of the program. Congress also explicitly ruled out substituting CHIP funds for Medicaid, despite the effect this might have on the “states which couldn’t wait,” but went on to expand Medicaid prior to CHIP.

Special Dilemmas for States with Broad-Based Medicaid Coverage

Nonetheless, developments in three states with broad-based Medicaid coverage—Minnesota, Vermont, and Washington—demonstrate the somewhat untoward effects of the BBA’s CHIP provisions on such states.

Minnesota already has a Section 1115 demonstration waiver that was implemented in July 1995 and built on the existing MinnesotaCare program, which covers low-income, uninsured people. Under the waiver, the state had already extended Medicaid eligibility to 275 percent of the FPL and had extended coverage to an additional 52,000 children and pregnant women. While CHIP would have allowed the state to extend Medicaid eligibility to 325 percent of the FPL, Minnesota officials chose instead to extend the eligibility limit from 275 percent of poverty to 280 percent; they apparently opted for this modest increase so that they could obligate their FY 1998 CHIP allotment. At the same time, they sought to amend the state’s 1115 waiver to allow the state to spend more than CHIP allows on administrative and/or planning expenditures (limited to 10 percent of the total state CHIP allotment) as well as to secure relief from CHIP’s maintenance-of-effort requirements. The amendment would have allowed the state to utilize more fully its $28.5 million allotment under CHIP.

HCFA denied the waiver amendment request, citing its prior policy guidance to the states that, since CHIP is a new program, it is not at this stage eligible for Section 1115 waivers, which are intended to foster new and innovative demonstration programs. As a result of HCFA’s interpretation and the state’s seeming unwillingness to choose other alternatives, it is estimated that Minnesota will be able to extend coverage to fewer than 50 additional children under CHIP.⁷

Vermont initially submitted to HCFA a CHIP plan that would have raised eligibility to 300 percent of the FPL through an expansion of its existing Medicaid program, which has operated under a Section 1115 waiver implemented on January 1, 1996. (Under the terms of the waiver, children were already covered in families with incomes up to 225 percent of poverty.) However, on August 5, 1998, Gov. Howard Dean officially withdrew the state’s CHIP application through a letter to President Clinton, which stated:

SCHIP is designed for states that have done the least to provide [Medicaid] coverage, and offers no flexibility to states such as Vermont. … Vermont will only experience an annual net gain of $100,000 over the regular matching rate. To establish a program with separate administrative and eligibility requirements for the small number of uninsured children (approximately 1,000) that we can cover under SCHIP is not cost effective. The added administrative expenses and systems requirements will more than use up the incremental increase in federal matching funds available under SCHIP, as compared to using traditional Medicaid.

In Washington State, the legislature rejected Democratic Gov. Gary Locke’s proposal to utilize CHIP funding to expand the Basic Health Plan, the state’s subsidized insurance plan, from 200 to 250 percent of the FPL to cover an estimated 10,000 uninsured children. Republican legislative leaders insisted that incomes up to that level—especially in rural and lower-cost areas of the state—were too high to merit public subsidies. Some observers point out that, in the past, the state never would have left available federal funds of this magnitude ($47 million for the first year) untapped. A reporter for State Health Watch noted: “Under the heading of ‘no good deed goes unpunished,’ one lobbyist said Washington has been penalized because it took aggressive steps to insure more of its population before passage of CHIP.”⁸

Whether to Participate

As noted above, only three states—Vermont, Washington and Wyoming—have decided not to participate in CHIP in the first year. As previously explained, the decisions of Washington not to participate and of Vermont to withdraw its initial CHIP plan are related to the relative generosity of their Medicaid programs.

Wyoming, on the other hand, offers relatively low-level Medicaid coverage but has not chosen to avail itself of CHIP to expand coverage. Part of the reason may lie in the fact that the state has the seventh highest level of state fiscal effort, according to a study conducted by Urban Institute researchers Toby Douglas and Kimura Flores, who define fiscal effort as “revenue from state and local taxes divided by state personal income.”⁹ Quite simply, given that Wyoming has only the 35th highest median per capita income in the nation, while it exerts such a high level of fiscal effort, it may have simply been beyond the political will or fiscal capacity of the state to raise additional revenues to draw down federal matching funds under CHIP.

Action by Congress in May 1998 may lead Washington and Wyoming to rethink their decisions not to participate in CHIP. In enacting P.L. 105–74, Congress allowed states to obligate CHIP allocations for FY 1998 by submitting plans before September 30, 1999—the end of FY 1999. HCFA has received some very preliminary feelers from the two states about possible reconsideration of their CHIP decisions. Alternatively, some Washington counties may attempt to draw down some or all of the state’s CHIP allotment if the state itself fails to do so.

Medicaid Expansion or Separate State Program?

A number of analysts have looked at the pros and cons of the three approaches available to a state under CHIP—(a) create or enlarge a separate state program to cover uninsured children, (b) expand their Medicaid program, or (c) adopt a combination of both approaches. In a CHIP implementation guide that it published last November, the House Committee on Commerce noted:¹⁰

Fundamental to this analysis is a critical question: which approach—a State-only program or a Medicaid expansion—will best enable a State to expand coverage and services to the largest number of low-income uninsured children? After all, it was for the purpose of providing coverage and services to such children that the S-CHIP was created. … From an eligibility perspective, a State-only S-CHIP program would enable States to expand coverage to more low-income uninsured children than would an S-CHIP-financed Medicaid expansion.

Coming at the issues from a very different perspective, Mann concluded:

A unified system of covering low- and moderate-income children that builds on the current Medicaid program will be most likely to reach and enroll eligible children, take advantage of Medicaid’s bargaining power and cost-efficiencies, and maximize available federal funding, thereby assuring the new child health initiative results in the greatest number of children receiving comprehensive and affordable coverage.¹¹

Alan Weil of the Urban Institute has capsulized the tradeoffs as follows:

The principal reason for states to use new S-CHIP funds to expand Medicaid is that they can build on an existing infrastructure. … However, states that expand Medicaid are expanding a program for which they have long sought more flexibility.¹²

Regardless of the merits of one approach over another, certain characteristics of states and their pre-CHIP Medicaid programs seem to have predisposed them initially to opt for one approach or another under CHIP. For example, it might be conjectured that, if a state already has a separate children’s or family health insurance program, it is more likely to use CHIP to build on that program than to expand Medicaid.

According to a May 1997 Alpha Center study written by Anne Gauthier and Stephen Schrodel, eight states—California, Colorado, Florida, Massachusetts, New Jersey, New York, Pennsylvania, and Washington—already had such programs.¹³ Three of these states—Colorado, New York, and Pennsylvania—have chosen to enlarge their current state programs under CHIP. Four others—California, Florida, Massachusetts, and New Jersey—chose to pursue a combination approach under CHIP. The eighth state with an existing separate state program, Washington, opted not to participate in CHIP. None has opted to expand Medicaid alone.

Viewed from another perspective, of the 23 states that chose either a separate state program or a combination approach under CHIP, 7 already had separate state programs for children or families. Thus, only 17 states chose to develop entirely new state programs under CHIP and, of these, 9 chose a combination approach, relying in part on an expansion of their Medicaid programs.

Another seemingly predisposing factor under CHIP is whether a state already has a Section 1115 demonstration waiver for its Medicaid program. It might be posited that if a state already has an 1115 waiver, it has been afforded a great deal more flexibility than it would have if it operates a more conventional Medicaid program and, thus, might be more inclined to expand its Medicaid program under CHIP than to opt for a separate state program. According to data from HCFA, 17 states currently have operational Medicaid 1115 demonstrations.¹⁴

Of these 1115 waiver states, seven—Arkansas, Maryland, Minnesota, Ohio, Oklahoma, Rhode Island, and Tennessee—have chosen to expand their Medicaid programs under CHIP. (An eighth, Vermont, will apparently expand Medicaid without participating in CHIP.) Only four—Arizona, Delaware, New York, and Oregon—opted to pursue a separate state program approach; New York already has a separate state program, and Arizona’s AHCCCS program is considered by some to be a non-Medicaid program, possibly making its choice of a separate state program largely a matter of semantics. An additional four states with 1115 waivers—Alabama, Kentucky, Massachusetts, and New Jersey—chose a combination approach. Two of these—Massachusetts and New Jersey—already had separate state programs.

Finally, there appear to be some regional patterns in the choice of basic approach under CHIP (Table 1). Of the 12 states in the Midwest, all but 2 have chosen Medicaid expansions. Of the nine states in the Northeast, five have opted for a combination approach. Of the 13 states in the West, 2 have opted not to participate; of the remaining 11, 6 have chosen the separate state program route. Of the 16 states in the South, 9 have opted to expand Medicaid using CHIP funds.

Table 1

Type of CHIP Program, by Region.

Choice of Eligibility Level

As mentioned previously, states have chosen CHIP eligibility levels that range between 100 percent and 300 percent of the FPL. Thirteen states and the District of Columbia have set their CHIP eligibility levels at exactly the 200 percent of the FPL that is generally allowed. Eight states that had Medicaid eligibility limits above 150 percent of the FPL before CHIP’s enactment have set CHIP eligibility levels above 200 percent, three of them—Connecticut, Missouri, and New Hampshire—at 300 percent. Twenty-three states have opted for eligibility levels below 200 percent, five of them—Alabama, Arkansas, Mississippi, North Dakota, and Texas—at 100 percent.

It might be hypothesized that the state’s median per capita income would have some influence on its choice of eligibility level under CHIP, since to some degree, per capita income reflects the fiscal capacity of the states. If the state’s choices of eligibility level are ranked by quintile according to their median per capita income, the distribution shown in Table 2 occurs.

Table 2

States’ Choices of Eligibility Levels under CHIP, Ranked by Per Capita Income Quintile.

While there is not a precise, straight-line relationship between state median income and choice of CHIP eligibility level, there appears to be a pattern in which those states with higher median incomes generally have opted for higher eligibility levels and those with lower per capita incomes generally have opted for lower eligibility levels.

There are also regional patterns in the choice of CHIP eligibility levels. Of the eight states with eligibility limits greater than 200 percent of poverty, five are in the Northeast. Of the 23 states that have chosen eligibility limits lower than 200 percent, 9 are in the South and 7 are in the Midwest.¹⁵

Containing Crowd-Out

One of the major concerns of Congress in enacting CHIP was to assure that it does not supplant—or “crowd out”—existing insurance coverage for the eligible population. Accordingly, the BBA limits CHIP coverage to children without other forms of “creditable coverage,” as defined in the Health Insurance Portability and Accountability Act (HIPAA) of 1996. States have an affirmative obligation to assure that children with existing coverage are not enrolled in CHIP.

In their CHIP plans, states have taken a variety of approaches in addressing crowd-out. According to HCFA data on the 23 state plans approved through mid-July, the most common approach is some kind of waiting period prior to CHIP coverage, an option chosen by 11 states; for most of these states, the period is either three months or six months. (New Jersey imposes a 12-month waiting period. After initially imposing a six-month wait, North Carolina will eventually phase in a 60-day waiting period.) Eight states have made no explicit provision for crowd-out in their plans, suggesting that they will monitor developments and take appropriate action if a significant problem materializes. A number of states will subsidize or supplement existing employer coverage as a means of limiting erosion of private coverage and thereby reducing crowd-out.

The crowd-out issue is one that has caused a great deal of concern to both fiscal conservatives and advocates for the poor. Conservatives point out that, given the limited funds available through CHIP, if subsidies are used by people who are already insured, the capacity of the program to reduce the number of uninsured children will be compromised. Some say that allowing people who are already insured to enroll in CHIP might also skew participation in favor of this population and away from the generally lower-income uninsured.

On the other hand, advocates for the poor contend that some degree of crowd-out is inevitable in a program like CHIP and express concern that the more rigorously measures to prevent crowd-out are enforced, the more likely they are to impede CHIP participation and undermine the ability of the program to reach the maximal number of uninsured children. Larry Levitt and Judith Feder, in a paper the Kaiser Family Foundation issued earlier this year, raised some related concerns:¹⁶

It is important to recognize that currently insured low income families who choose to take advantage of new coverage opportunities do so because it gives them financial relief or better coverage. This relief seems at least as legitimate as the relief recent legislation has provided self-employed families through tax preferences (with no evidence of expanded coverage). Further, denying one group of low income families a benefit awarded to others of similar income seems unfair, especially if the insurance coverage they have entails substantial financial sacrifice.

Thus, while both sides of the crowd-out debate seem to share a basic concern about covering as many children as possible, they seem to differ fundamentally about the desirability of preventing crowd-out from occurring.

Speakers

Debbie I. Chang, co-chair of the DHHS Steering Committee on CHIP Implementation and director for benefits, coverage, and payments in HCFA’s Center for State Operations, will lead off with an overview of CHIP as well as some general observations about the CHIP plans that have been submitted to HCFA and the options the states are choosing.

She will be followed by four state officials who have played a major role in the implementation and management of CHIP in their states: Barbara Ladon, director of the Office of Program Development of the Colorado Department of Health Care Policy and Financing; Sandra Shewry, executive director of the California Managed Risk Medical Insurance Board; Greg Vadner, director of the Division of Medical Services in the Missouri Department of Social Services; and Gwendolyn Williams, commissioner of the Alabama Medicaid Agency. They will talk about how CHIP is configured in each of their states, the planning and decision-making process that produced their CHIP plans, the politics of CHIP in their states, and what they perceive to be the strengths and weaknesses of CHIP statute and regulations.

Following their presentations, Ms. Chang will respond to their remarks before a general discussion between the audience and the panel.