A parent earns too much for Medicaid and too little to buy family coverage in a private market, holds a job that insures the worker alone or insures nobody at all, and faces a child’s fever, fracture, or asthma attack with no coverage card to present at a clinic desk. That position, stranded between an entitlement built for the poorest households and an employer based market priced for groups, is the operative puzzle of the Children’s Health Insurance Program. Federal law held an answer for a child below a state Medicaid income line and no answer for a child just above that line, and the distance between the two positions was measured in deferred visits, unpaid bills, and preventable illness carried into classrooms. A statute written for that gap had to decide how far federal responsibility reached into working households, and the decision could not be separated from the form of money chosen to carry the responsibility, because the form of money determined whether coverage rested on an enforceable claim held by a child or on a funded place inside a program with a ceiling set in advance. The puzzle was therefore fiscal and moral at once, and a solution that answered only one half of the puzzle had no path through the Congress that convened in January 1997.

The calendar made the choice sharper. In 1996, one year before the children’s statute, Congress ended the federal entitlement to cash assistance that had operated since 1935 and replaced the entitlement with a capped block grant, paid to states under state plans at a fixed sum, with broad state discretion over benefits and rules. In 1997, Congress reached for the same instrument in a different field and for a different population. The Balanced Budget Act of 1997 created Title XXI of the Social Security Act, offered states an enhanced federal matching rate for children’s coverage, divided a capped national total into state allotments, and left each state to design a program inside the total that Congress had named. The namable claim at the center of this profile can be stated in a single line: A block grant for children’s health. A claim of that shape invites a verdict borrowed from welfare policy, where a fixed federal grant, held against inflation and recession, eroded in real value while need rose and fell with the economy. The children’s record refuses that verdict in a simple form and refines the verdict instead, because the allotments described in the sections ahead were funded generously relative to measured need in ordinary years, rebased on actual spending after a decade of operation, backed by redistribution and a contingency fund, and increased repeatedly at reauthorization, while the failure the structure reliably produced arrived on a calendar date, at the expiry of funding authority, rather than through slow erosion of a frozen sum.

The route through the profile follows the money first and the politics second, because the politics becomes legible only after the fiscal machinery is visible. An opening account of the statute and the allotment formula gives way to the problem of uninsured children in the middle 1990s and the coalition that answered the problem, led in the Senate by a Republican committee chairman from Utah and a Democratic health policy veteran from Massachusetts, financed in part by a tax on cigarettes, and preceded by state experiments in Pennsylvania, Florida, and Minnesota. Attention then turns to the enhanced matching rate and the arithmetic that made governors want a place inside a capped program, to the three state options and the entitlement consequences that followed from each option, to the three rebuildings of the funding in 2009, 2010, and 2015, to the lapse of funding authority in the autumn of 2017 and the restoration that followed in early 2018, and finally to the coverage record and the promises the statute never made. Readers following a statute of this complexity across many sittings can keep your statute notes, citations, and case chronologies together free on VaultBook while tracing an allotment formula in one section, a Senate bargain in another, and a funding lapse two decades later in a third.

Two cautions govern the reading. First, coverage figures belong to the surveys that produced the figures, and a rate drawn from the National Health Interview Survey and a rate drawn from the Current Population Survey describe the same population through different instruments, so each figure in the record section carries the name of the survey behind the figure. Second, attribution for the steep decline in child uninsurance after 1997 belongs jointly to the children’s program, to Medicaid expansions for children, and later to the health law of 2010, and no honest account assigns the whole decline to Title XXI alone. Held against those cautions, the comparison that frames the profile remains instructive. The welfare block grant of 1996 and the children’s allotment of 1997 shared a capped form, a state plan, and state discretion, and diverged in funding generosity, in rebasing, in matching incentives, and in the constituency that defended the program at each deadline. That divergence, traced through structure, coalition, match, options, reauthorizations, lapse, and record, is the evidence on offer for a narrower and more useful lesson about capped federal grants than any general verdict can supply.

What the Children’s Health Insurance Program Is and How Its Money Works

The statute answers a precise and uncomfortable question that the American health coverage system had left open for decades. A parent works, earns wages, and supports a household on an income that a state Medicaid program counts as too high for enrollment, and the job that produces those wages carries no health coverage for the worker’s children, or carries coverage for the worker alone. The child in that household belongs to no category that the older public programs recognize, and belongs to no private arrangement that the parent’s employment supplies. Illness, injury, and the routine care that childhood requires do not wait for a category to appear. The question the statute answers is therefore not whether government should enter the field of child health coverage in the abstract, but how far a federal commitment should extend toward children whose families work for a living, earn beyond the limits of the program built for the poorest households, and receive nothing from an employer in place of that public help. Every structural choice in the statute flows from that question, because the population it names sits between two systems, and a program built for a population between systems must define its boundaries with unusual care.

The formal answer arrived in a single legislative vehicle. The Children’s Health Insurance Program, enacted as Title XXI of the Social Security Act, was created by the Balanced Budget Act of 1997, Public Law 105-33, signed on August 5, 1997, and codified at 42 U.S.C. 1397aa and following. That placement inside the Social Security Act mattered for more than organizational convenience. The Act already housed the federal government’s principal health coverage commitments, and a new title placed beside them announced that child health coverage had become a permanent subject of federal social legislation rather than a temporary demonstration or a discretionary pilot. At the same time, a separate title, with separate financing rules and separate conditions, announced just as clearly that the new program was not an extension of the older titles by another name. The distinction between placement and separation runs through the entire design. The program borrows the administrative habits, the federal and state partnership, and much of the vocabulary of the coverage titles that preceded it, and then departs from them at the point that counts most, the point where federal money stops.

In one clear pass, the program does a limited and well defined thing. The federal government offers matching funds to states for the purpose of providing health coverage to children, states draw those funds under state plans that the federal government has approved, and the Centers for Medicare and Medicaid Services within the Department of Health and Human Services administers the program at the federal level. A state that wishes to participate submits a plan describing how coverage will be furnished, which children will be reached, and how the program will be organized, and federal approval of that plan opens the door to federal payment. Nothing in that basic transaction guarantees a place to any individual child in the way an entitlement would, and nothing in it compels a state to participate on terms the state has not accepted through its own plan. The transaction is instead a structured offer, generous in its matching rate, bounded in its total, and conditional on a plan. States accept the offer because the matching terms reward acceptance handsomely, and the federal government gains a child coverage program in states whose own legislatures might never have financed one alone. The simplicity of that pass can mislead a reader about the machinery underneath it, because the conditions attached to the money determine what kind of coverage actually reaches a child.

The child at the center of the statute carries a defined statutory description. A targeted low-income child is, in general terms, an uninsured child under the age of 19 living in a family whose income rises above the thresholds that the child’s state applies under Medicaid. Each element of that description performs work. The age element confines the program to childhood and draws a boundary at the point where adult coverage categories were expected to take over. The insurance element confines federal payment to children who lack coverage, directing the money toward a gap rather than toward replacing arrangements already in place. The income element operates in two directions at once, because a family must earn too much for Medicaid and yet count as low-income under the standards the state establishes for the new program. States set the upper income limits for that band, and many states set them well above 200 percent of the federal poverty level, a choice that reflects how far up the wage scale the absence of dependent coverage extends. A child outside any one of those elements falls outside the federal payment structure, however sympathetic the child’s circumstances might appear to an administrator reading an application.

That band between Medicaid eligibility and secure private coverage deserves close attention, because the statute’s drafters treated it as a distinct social fact rather than as a residual category. Families inside the band work in the sectors of the economy where dependent health benefits had always been least reliable, in small firms, in service employment, in jobs structured around hourly wages without a benefits package attached. A parent in such employment might earn enough across a year to clear a Medicaid income threshold calculated for a household in poverty, and yet face the full price of a child’s medical care alone, with no group plan through which to spread that risk. The statute does not treat that exposure as a personal failure of planning or bargaining, and it does not treat Medicaid’s income line as a fair measure of a family’s ability to buy coverage in a private market priced for groups. It treats the band itself as the problem, and it sizes the program to the band, reaching upward as far as states choose to set their limits while refusing to convert the program into a general subsidy for families at every income.

The defining structural fact of the program, the fact from which its politics, its administration, and its vulnerabilities all descend, is that federal money arrives as a capped allotment rather than as an open promise. Congress fixes a sum for the program, that sum is divided among participating states as individual allotments, and federal payments to a given state for its child health program cannot exceed the allotment assigned to that state. The ceiling is not a guideline, a target, or an estimate that administrators may exceed when need runs high. The ceiling is the legal limit of federal liability. A state that spends beyond it spends its own money alone, restructures its program to spend less, restricts enrollment within the flexibility the statute allows, or waits for further congressional action. That single feature separates the program from the largest federal health coverage commitment of the twentieth century, and the separation was deliberate, debated, and understood by the legislators who voted for it. A capped allotment converts an open social commitment into a budgeted one, and budgeting the commitment changes the behavior of every institution that touches the program, from a congressional appropriations committee to a state agency deciding how widely to advertise enrollment.

Medicaid, against which the new program was measured at every stage of drafting, rests on the opposite architecture. Enacted as Title XIX of the Social Security Act in 1965, Medicaid obliges the federal government to match state spending on covered services for eligible individuals at a rate fixed by formula, with no ceiling on the total federal payment a state may draw. Eligibility under Medicaid carries the character of an individual entitlement, so a person who meets the governing criteria holds a claim to coverage that a state administering the program must honor, and federal payment follows that claim wherever it leads. When enrollment swells in a recession, when medical prices climb, when a costly treatment becomes standard care, federal Medicaid spending rises automatically with state spending, because the federal obligation attaches to the expenditure itself rather than to a sum set in advance. Congress influences that spending through eligibility law, benefit law, and the matching formula, but between statutory changes the money flows as the claims arise. The contrast gave the drafters of 1997 a clean choice between two fiscal temperaments. One temperament trusts the entitlement to define the spending. The other fixes the spending and lets the program define itself within the total. Title XXI chose the second temperament while borrowing the first program’s administrative skeleton.

What is the difference between a capped allotment and an open-ended match?

A capped allotment sets a fixed ceiling on federal payments to a state for a period, so federal liability ends at that ceiling. An open-ended match sets no ceiling, so federal payments rise with every dollar of eligible state spending. The first fixes federal exposure in advance; the second follows state spending wherever eligibility leads.

The depth of that difference becomes visible in the sequence of decisions each structure forces. Under an open-ended match, the consequential federal decision happens at the moment eligibility is defined, because every later payment is a consequence of that definition working itself out through clinics, hospitals, and state claims systems. Under a capped allotment, the consequential federal decision repeats on a calendar, because each period’s allotment must be set, divided, and defended as a sum of money, and a sum of money invites a debate that an entitlement formula suppresses. Legislators who would hesitate to vote against coverage for a named category of children may vote without hesitation for a smaller total figure in a budget table, since the total carries no child’s name and no family’s story. Administrators learn that arithmetic quickly. A program financed by allotment lives inside budget politics in a way a program financed by entitlement largely escapes, and the history of federal child coverage after 1997, with its recurring rounds of extension legislation and its recurring arguments about the size of the fund, illustrates the exposure that the structure creates. The cap does not merely limit money. The cap schedules the political argument about money, again and again, at intervals the statute itself produces.

An allotment, however, is not an arbitrary number assigned by negotiation alone, and the statute’s method of building one reveals what Congress thought a fair division of a fixed sum required. The formula rests principally on each state’s share of the nation’s low-income children and each state’s share of the nation’s low-income uninsured children, with an adjustment for health care costs. The first measure, the share of low-income children, anchors the division in the size of the population the program exists to serve, so a state where low-income childhood is common draws a larger portion of the national total than a state where such childhood is rare. The second measure, the share of low-income uninsured children, sharpens that anchor toward need as the statute defines it, because a low-income child who already holds coverage, public or private, does not present the gap the program was built to close, while a low-income child without coverage presents exactly that gap. Weighting the division toward uninsured children directs the capped sum toward the places where the defining problem concentrates, rather than spreading the sum evenly across populations with very different rates of uninsurance.

The health cost adjustment completes the formula’s logic by recognizing that a dollar of coverage does not buy the same quantity of care in every state. Medical prices, provider markets, wage levels in health care employment, and the ordinary cost of delivering a given service differ across the country, and a formula that divided money strictly by head count would systematically underfund coverage in high cost states and overfund it in low cost states relative to the care each dollar must purchase. An adjustment tied to health care costs corrects the head count toward purchasing power, so that allotments approximate equal capacity to buy coverage for equal numbers of uninsured low-income children. The correction also carries a judgment about fairness between states. Population alone would have been simpler to calculate and simpler to defend, and cost differences could have been left for states to absorb from their own revenues. The statute instead absorbs part of that difference at the federal level, treating the geography of medical prices as a national condition that a national formula should recognize. Taken together, the three elements, low-income children, low-income uninsured children, and health costs, produce an allotment that claims to be a measurement rather than a bargain, even though every measurement of that kind rests on choices about which facts count and how heavily each one weighs.

Inside that discipline, one further ceiling governs how a state may use the money it draws, and its small percentage carries an outsized significance for what the program is. A state may spend no more than ten percent of the funds it draws on administration, outreach, and related activities that do not themselves purchase coverage for children. Nine tenths or more of every federal dollar drawn under the program must therefore land on coverage, on the premiums, services, and care that constitute the program’s reason for existing. The limit answers a suspicion that attends every grant program, the suspicion that money appropriated for a sympathetic population will settle into salaries, contracts, systems, and publicity before any of it reaches the population named in the statute. By fixing the administrative share in the statute itself, Congress removed that allocation from annual administrative discretion and made the coverage-first character of the spending a matter of law rather than of good intention. A state that builds an elaborate apparatus to run a small program finds the apparatus itself constrained by the same rule that constrains a state running a large one, because the percentage applies to funds drawn rather than to the ambition of the bureaucracy drawing them.

The ten percent allowance deserves a careful reading in the other direction as well, because the activities it permits are not waste by another name. Administration includes the eligibility determinations, enrollment systems, claims processing, and oversight without which coverage exists only on paper, and outreach includes the work of finding children whose families do not know that a program exists for them. A program aimed at a population between systems faces a particular outreach problem, since the families it serves have no established relationship with a welfare office and no employer benefits department to hand them a form. Children who are eligible and unenrolled receive nothing from an appropriation, however generous the appropriation may be, so the statute’s decision to let a bounded share of funds support outreach amounts to a recognition that coverage must be delivered as well as financed. The cap holds both truths at once. It treats administration and outreach as necessary, funds them, and refuses to let them grow at the expense of the coverage they exist to serve. A percentage that began as an accounting rule therefore expresses a theory of the program, in which the overwhelming purpose of the money is care for children and every other purpose justifies itself as a means to that care.

The final piece of the allotment machinery softens the rigidity of a fixed division without surrendering it, and it does so by attending to a plain fact about forecasts. No formula, however carefully built, predicts each state’s actual need in each period with precision. One state draws its full allotment while enrollment climbs beyond expectation, another state draws only part of its allotment while enrollment lags behind the population figures that produced the division, and a system that locked each sum inside its original state would waste money in one place while children went uncovered in another. The statute therefore provides, in principle, for the redistribution of unspent allotments. Funds that a state leaves unspent after a limited period of years move through a federal redistribution toward states facing shortfalls, so that the national total Congress appropriated works as fully as the pattern of state need allows. Redistribution preserves the cap at the level where the cap was designed to bind, the level of total federal spending, while relaxing it at the level of the individual state, where a forecast error could otherwise punish children for the inaccuracy of a formula. The mechanism also changes state incentives at the margin, because an allotment hoarded without use does not remain a permanent reserve, and an allotment spent responsibly does not leave a state defenseless against a surge that arrives after careful planning.

Read as a whole, the money mechanics express a single coherent settlement between ambition and restraint that the politics of 1997 made possible and that the statute then locked into law. The ambition appears in the matching offer itself, in rates generous enough to draw states into covering a population that no federal law had previously reached as a class, and in a formula that sends the largest shares of the fund toward the states where low-income uninsured children concentrate. The restraint appears in the ceiling on each state’s draw, in the percentage reserved against administrative absorption, and in the periodic return of the entire program to Congress for further funding decisions. Neither half of the settlement cancels the other. A state receives strong inducement and real money, and receives both inside boundaries that the state cannot enlarge by spending. A child gains a route to coverage that had not existed in federal law before August 1997, and gains it through a structure that promises the coverage as a funded program rather than as an individual legal claim in the separate program designs the statute permits. Understanding the program begins with holding those two halves together, because later arguments about its size, its rules, and its survival were, at bottom, arguments about whether the settlement of 1997 struck the balance in the right place.

The Problem of Uninsured Children and the Coalition That Answered It

By the middle of the 1990s, children occupied a large part of the gap in American health coverage. The National Health Interview Survey, conducted by the National Center for Health Statistics, recorded 13.9 percent of children under age 18 as uninsured at the time of interview in 1997. The Current Population Survey, conducted by the Census Bureau, put the 1997 figure for children at about 15.0 percent, a share that the Department of Health and Human Services analysis of the survey translated into 10.7 million children. The National Health Interview Survey classified a child as uninsured based on coverage held around the time of a face to face household interview, producing an estimate for an average month of the survey year. The Current Population Survey asked respondents in the spring about coverage during the previous calendar year and, by its formal definition, classified a person as uninsured only after a full calendar year without coverage, even though Department of Health and Human Services analysts noted that many respondents appeared to answer with their status at the interview in mind. Named side by side, the surveys told a consistent story: roughly one child in seven lacked health insurance in 1997.

The children behind the percentages shared a recognizable position in the coverage system. Medicaid, enacted in 1965 as Title XIX of the Social Security Act, covered children in families with very low incomes, and expansions enacted in the late 1980s and the early 1990s had raised the income lines for infants, young children, and pregnant women. Above those lines stood families whose earnings disqualified their children from Medicaid and whose jobs offered no affordable family coverage or no coverage at all. A 1997 Congressional Record statement reported that employer provided dependent coverage had fallen from 61 percent of children in 1988 to 54 percent in 1993. Parents in low wage service work, in small firms, in part time jobs, and in jobs with waiting periods or employee premium shares that consumed a large fraction of pay faced a private market that offered few child only policies and priced family policies beyond reach. Children were inexpensive to insure relative to adults, childhood illness carried consequences for schooling and development that lasted into adult life, and no named participant in the 1997 debate defended a position that children in working families ought to go without care. The dispute concerned the instrument, and the instrument had to survive a decade that had already destroyed one comprehensive answer.

That destroyed answer was the Clinton administration plan of 1993 and 1994. President Bill Clinton addressed a joint session of Congress on September 22, 1993, and presented a proposal for universal coverage built around required employer contributions, purchasing cooperatives, federal standards for benefits, and a National Health Board with authority over spending. A White House task force led by First Lady Hillary Clinton had prepared the plan, and Representative Richard Gephardt introduced the Health Security Act in the House in November 1993 while Senator George Mitchell introduced the measure in the Senate. The scale that President Clinton presented as the virtue of the plan became the ground of attack for the Health Insurance Association of America, which financed the Harry and Louise advertisements against the proposal, and for small business organizations that opposed an employer mandate. Senate Majority Leader George Mitchell offered a compromise in August 1994 that delayed employer requirements and exempted small businesses, and in September 1994 Senator Mitchell announced that the compromise had failed and that health reform would wait for a later Congress. No version of the bill reached a final vote in either chamber. The November 1994 elections then gave the Republican Party control of the House of Representatives and the Senate for the first time since the Congress of 1953 and 1954. Republican leaders treated any new federal health commitment as suspect when the commitment carried an open ended entitlement, a mandate on employers, or a new federal administrative structure. A proposal that hoped to become law after 1994 had to be narrower, cheaper, and easier to explain.

A child only, state run design fit those constraints in a way that no adult inclusive design could. Sponsors stated the reasons openly: childhood coverage cost less per person than adult coverage, preventive care in childhood averted expensive treatment afterward, and working parents formed a constituency that members of both parties described as deserving of help. A state run structure answered the objection that had followed the 1993 plan from its introduction to its collapse, the objection that a federal program would displace state judgment and private arrangements under a single national template. A federal grant that set broad goals and left eligibility bands, benefit choices, contracting, and outreach to state governments matched machinery that existed in every state capital, where governors of both parties had administered Medicaid for three decades. A capped federal commitment answered the budget objection. Senate Majority Leader Trent Lott and the Republican budget leadership insisted during the spring of 1997 that new spending fit inside the budget agreement negotiated with the White House, and a fixed allotment for children let supporters name the full federal cost in advance in a way that an entitlement, obligated to pay for every eligible child who enrolled, never allowed. A fixed allotment could also run short, and a state could respond with enrollment limits that an entitlement forbade. That trade sat at the center of the 1997 design from the beginning, accepted by Senator Edward Kennedy as the price of Republican votes and defended by Senator Orrin Hatch as the discipline that made federal help compatible with a balanced budget.

State governments supplied the evidence that such a design could operate, because several states had built child coverage programs with state money before Congress acted. Pennsylvania enacted a Children’s Health Insurance Program in 1992, and Pennsylvania officials financed the state program in part with a two cent tax on a pack of cigarettes, according to accounts given by Pennsylvania officials during the 2017 funding lapse. The Pennsylvania program offered free or low cost coverage to children in families with incomes above the Medicaid lines, used private insurers to deliver benefits, and treated the child as the unit of coverage rather than the family case at a welfare office. The federal statute placed the preexisting programs of Pennsylvania, Florida, and New York inside the new framework as programs that met federal standards, and Pennsylvania officials described the state program as a model for the federal effort. The Pennsylvania experience answered a practical question with an operating program: a state could identify uninsured children above the Medicaid line, contract for a defined benefit, charge modest premiums to families with higher incomes, and enroll children whose parents had no route to coverage through employment.

Florida supplied a second demonstration with a different administrative idea. Florida Healthy Kids grew out of a 1988 University of Florida study and a pilot program financed jointly by the State of Florida and the Robert Wood Johnson Foundation. Florida used school districts to group children into purchasing pools, moving the insurance contract from the employer to the school and making coverage portable for a child whose parent changed jobs, a feature that held particular value for a child with a preexisting condition. Families that qualified for no public aid and could afford no private policy bought a product designed for children, with premiums that reflected the lower cost of child coverage. The evaluation described in the Congressional Record statement recorded a 70 percent decrease in emergency room visits per enrollee and gains in primary care visits and immunizations, along with reports from teachers of improved school attendance. Florida demonstrated that enrollment could follow the school roster rather than the welfare roll, that a child only risk pool could support a marketable product, and that preventive visits rose when the price barrier fell.

Minnesota supplied a third demonstration with a broader ambition. MinnesotaCare grew out of a children’s insurance effort that Minnesota had begun in 1988, and under a federal waiver approved in 1995, MinnesotaCare covered children and pregnant women at incomes that reached 275 percent of the federal poverty level, with premiums set on a sliding scale tied to family income. The Minnesota Department of Human Services administered MinnesotaCare, and Minnesota officials defended the program in Washington as proof that a state could run subsidized coverage outside the Medicaid structure while coordinating with Medicaid for families whose incomes fell. Taken together, the state experiments demonstrated working parts that federal drafters could assemble: an income band above Medicaid, premiums graduated by income, contracts with private plans, enrollment through schools and community institutions, and a benefit built around primary and preventive care. The experiments also demonstrated the limits of state action alone. Each program depended on a state tax or a state appropriation, each covered a fraction of the uninsured children in a single state, and no state program could spread the cost across the national tax base or set a floor under poorer states. Senator Kennedy pointed to the state record in Senate floor statements as evidence for feasibility and to the gaps in the state record as evidence for federal money.

One year before Congress wrote the children’s program, a different statute had settled the form that federal money for family support would take. President Clinton signed the Personal Responsibility and Work Opportunity Reconciliation Act on August 22, 1996, and the statute ended the Aid to Families with Dependent Children program, a federal entitlement to cash assistance that had operated since 1935, and replaced the program with Temporary Assistance for Needy Families, a block grant that paid states a fixed sum under a state plan and left states broad discretion over benefits and rules. The Congressional Research Service history of the Temporary Assistance for Needy Families block grant described the change as a replacement of an entitlement with a broad purpose block grant. The Children’s Health Insurance Program, enacted one year after the 1996 welfare law replaced an entitlement with a capped block grant, used the same capped logic for health coverage. Federal payments to a state under the children’s program stopped at the state allotment, state plans described the program that a state intended to run, and flexibility over design was the stated compensation for the federal ceiling. In cash assistance and in child health coverage alike, Congress of the middle 1990s chose a fixed federal sum, a state plan, and state discretion over an individual federal guarantee.

Who were the Senate architects of the children’s health proposal?

Senator Orrin Hatch of Utah, a Republican and chairman of the Senate Judiciary Committee, and Senator Edward Kennedy of Massachusetts, a Democrat and the ranking Democrat on the Senate Labor and Human Resources Committee, jointly authored the children’s health proposal introduced in March 1997.

The pairing itself was the story that every account of the 1997 effort placed first, because the two names carried opposite expectations. Senator Hatch represented Utah, held the chairmanship of the Judiciary Committee in a Republican Senate, and had built a Senate career that began in 1976 with a conservative program of limited federal spending. Senator Kennedy represented Massachusetts, served as the leading Democrat on the Committee on Labor and Human Resources, and had pursued national health coverage as a central goal since the 1970s. Neither sponsor alone commanded the same reach across the divided Senate, and together the two senators removed the easiest objections available to each side. Senator Hatch gave the proposal standing with Republican leaders who controlled the Senate calendar and with a Judiciary Committee chairman’s network across the Republican conference. Senator Kennedy gave the proposal committee drafting experience, a record on health legislation that reached back decades, and credibility with children’s advocacy organizations that had pressed the issue through the failed reform of 1993 and 1994. Chris Jennings, the White House health policy adviser, treated the joint bill in a memorandum to President Clinton dated March 13, 1997 as the event that required a presidential response, and summarized the proposal as a departure from the bill that Senator Kennedy had introduced with Senator John Kerry during the previous September.

The substance of the joint bill showed the negotiation inside the partnership. The Child Health Insurance and Lower Deficit Act, announced on March 13, 1997, proposed a grant program of 20 billion dollars for children’s health coverage, paid to states that would contribute a state share and set eligibility within federal limits. The Jennings memorandum described the design as fixed funding with state flexibility and no entitlement, in contrast to the earlier Kennedy and Kerry proposal for a fully federal entitlement that subsidized children in families with incomes up to 300 percent of the poverty level. Senator Kennedy accepted a capped grant, a state contribution, and state discretion that his earlier bill had rejected. Senator Hatch accepted a new federal program, a federal tax increase to pay for the program, and a federal role in child coverage that reached families well above the poverty line. A Republican chairman who endorsed a tobacco tax and a Democratic leader who endorsed a block grant structure had each spent something that colleagues in the respective parties valued, and the joint sponsorship signaled a possible majority for a bill drawn at the chosen dimensions.

The financing device completed the design. The Child Health Insurance and Lower Deficit Act proposed an increase of 43 cents in the federal tax on a pack of cigarettes, from 24 cents to 67 cents, with comparable increases for other tobacco products, and Senator Kennedy stated the estimate on the Senate floor: the increase would raise about 30 billion dollars over five years, with 20 billion dollars directed to children’s coverage and the balance directed to deficit reduction, the destination that gave the bill the second half of its title. Senator Kennedy defended the source of the money on the Senate floor in March 1997 by tying the tax to the harm that justified the expenditure, arguing that smoking imposed direct health costs and lost productivity on the nation and that a higher price for cigarettes would deter teenagers from starting to smoke. Senator Hatch made the same deterrence argument in Senate debate, citing the share of smokers who began before age 18. The tax also solved a legislative problem. A dedicated excise increase also let the sponsors present the program as paid for at introduction, with a surplus for deficit reduction, inside a budget that Republican leaders and the White House had agreed to balance.

The first test of the coalition came on ground that the sponsors had not chosen. In May 1997, President Clinton and the Republican congressional leadership completed a budget agreement for fiscal 1998, and the Senate took up the budget resolution that carried the agreement. Senator Hatch and Senator Kennedy offered an amendment to the fiscal 1998 budget resolution that would have placed the children’s proposal, with the 43 cent cigarette tax, inside the budget framework. Senate Majority Leader Trent Lott condemned the amendment as a measure outside the budget agreement and, in remarks reported by Government Executive, called the amendment a deal breaker. The amendment divided the Clinton administration against a proposal that the administration favored in principle. First Lady Hillary Clinton served as a prominent administration advocate for children’s coverage, and President Clinton supported the concept of the Hatch and Kennedy proposal, as Senator Lott acknowledged in the debate reported by Government Executive. Budget leaders then appealed to the administration to defend the agreement that the White House had negotiated. Administration officials then lobbied Democratic senators to protect the agreement, President Clinton committed to Senator Lott that the administration would work for votes against the amendment, and a motion to table the amendment carried by 55 to 45. Government Executive reported that both Senator Hatch and Senator Kennedy attributed the result to the administration lobbying.

The defeat on the budget resolution redirected the proposal rather than ending the effort, because the budget agreement and the reconciliation process that implemented the agreement left a narrower door open. Reconciliation bills moved under special Senate rules that barred a filibuster, and the Balanced Budget Act of 1997, the reconciliation measure that implemented the agreement, became the vehicle for a scaled version of the children’s initiative. The negotiation that followed the May tabling produced a compromise that funded the children’s initiative at roughly 24 billion dollars over the first five years, the figure that Senator Kennedy cited on the Senate floor in September 1997 while defending the tobacco revenue that supported the initiative. The tax, by contrast, fell sharply. In place of the 43 cent increase that the two sponsors had proposed in March 1997, the enacted compromise raised the federal cigarette tax by 15 cents a pack in two steps. The Congressional Research Service summary of federal tobacco taxes recorded the schedule that the Balanced Budget Act enacted: the first step raised the tax by 10 cents a pack to 34 cents on January 1, 2000, and the second step raised the tax by an additional 5 cents a pack to 39 cents on January 1, 2002.

President Clinton signed the Balanced Budget Act of 1997, Public Law 105-33, on August 5, 1997, and Title XXI of the Social Security Act took a place in the statute books beside the Medicaid title that Congress had enacted in 1965. Senator Kennedy had surrendered the entitlement structure that his health bills had carried for decades, and Senator Hatch had surrendered the position that new coverage belonged outside the federal budget. The Clinton administration had lobbied against the amendment that carried the administration’s own priority and then signed the compromise that the lobbying had made necessary. The state experiments in Pennsylvania, Florida, and Minnesota had shown that the machinery worked at state scale. The coalition of a Republican Judiciary Committee chairman from Utah and a Democratic labor committee leader from Massachusetts had shown that a majority existed for one particular shape of answer: children only, state administered, federally financed through a tax on cigarettes, and capped at a sum that Congress named in advance. That shape, rather than any larger ambition, was the answer that the politics of 1997 allowed, and the statute of August 5, 1997 gave the shape a title, an allotment formula, and a decade of consequences that later Congresses would inherit.

The Enhanced Match: How the Formula Made States Want In

A matching rate is a price, and the price Congress set in 1997 was designed to change behavior. The Children’s Health Insurance Program, enacted as Title XXI of the Social Security Act by the Balanced Budget Act of 1997, Public Law 105-33, asked states to build coverage for uninsured low-income children who earned too much, under state rules, to qualify for Medicaid. Persuasion alone had no line in a state budget. A matching formula had a line in a state budget. By paying a larger federal share for each dollar a state spent on the new program than the federal government paid for a dollar spent on Medicaid, the statute lowered the state price of covering a child and raised the cost, in forgone federal money, of refusing to participate. Generosity was the invitation. A capped federal allotment was the condition attached to the invitation, and the two parts belonged together from the first day of the program.

Title XXI did not invent a matching system. The new program borrowed the matching system of Medicaid, Title XIX of the Social Security Act, enacted in 1965, and then improved the terms for children covered under Title XXI. Medicaid financing rested on shared payment. A state spent money on medical assistance for eligible people under an approved state plan, and the federal government reimbursed a percentage of that spending. The percentage was called the federal medical assistance percentage, abbreviated as FMAP. The rate applied to most Medicaid service expenditures, it was determined for each state, and it was redetermined on a regular cycle rather than negotiated case by case. A governor and a budget office could plan around a published percentage, and a state that spent more on covered services drew more federal money at the same percentage, because Medicaid matching operated without a fixed federal ceiling for an individual state.

The regular FMAP embodied a judgment about fiscal capacity. The formula compared a state’s per capita income with per capita income for the United States as a whole, and it moved in the opposite direction from income. A state with per capita income below the national average received a higher federal percentage, and a state with per capita income above the national average received a lower federal percentage. The comparison was squared in the statutory formula, a construction that widened the effect of an income difference, and the formula used a factor of 0.45 so that a state with per capita income exactly equal to the national average received a federal share of 55 percent and bore a state share of 45 percent. In plain terms, a dollar of medical spending was harder for a poorer state to raise than for a richer state, and the formula compensated by shifting a larger part of the cost to the federal treasury, which drew revenue from the whole country.

Statutory limits kept that compensation within bounds. The regular FMAP had a floor of 50 percent and a ceiling of 83 percent. The floor meant that no state, however wealthy by the per capita income measure, paid more than half the cost of most Medicaid services from state funds, and the richest states clustered at that floor, receiving 50 cents from the federal government for each dollar of covered spending and supplying 50 cents from state sources. The ceiling meant that no state, however poor by the same measure, received more than 83 cents on the dollar under the regular formula, and a state at the ceiling bore at least 17 cents of each dollar. Between floor and ceiling, the national average rate sat in the high fifties before the 2010 health law, consistent with a formula centered on 55 percent for an average-income state and pulled upward by higher rates for lower-income states. In a state with a 60 percent FMAP, a 100 dollar Medicaid service cost drew 60 dollars of federal money and 40 dollars of state money. In a state at the 50 percent floor, the same service drew 50 dollars of each. The service was identical. The state price differed by ten dollars, and the difference followed from income rather than from need for care or provider prices.

How is the enhanced federal match for the program calculated?

The enhanced rate starts with a state’s regular Medicaid matching rate and reduces the state share by 30 percent. A state responsible for 40 percent under Medicaid becomes responsible for 28 percent, so the federal share rises from 60 percent to 72 percent, within statutory limits.

Section 2105(b) of the Social Security Act, codified at 42 U.S.C. 1397ee(b), stated the enhancement as a reduction in the state share, and that form deserves attention because a reduction in a share and an addition to a rate describe the same transaction from opposite sides of a ledger. The regular FMAP divided each dollar between the federal government and a state. The part that remained after the federal percentage was subtracted from 100 percent was the state share. Title XXI reduced that state share by 30 percent. A state share of 50 percent, cut by 30 percent of itself, fell by 15 points to 35 percent, and the federal share therefore rose by 15 points from 50 percent to 65 percent. A state share of 40 percent, cut by 30 percent of itself, fell by 12 points to 28 percent, and the federal share therefore rose by 12 points from 60 percent to 72 percent. The reduction was proportional rather than flat. Wealthier states at the Medicaid floor gained the largest number of points, because the enhancement operated on the largest state shares, and poorer states gained fewer points, a pattern partly corrected by the floor and ceiling around the enhanced rate.

The same formula reads in a second form, and the equivalence between the two forms is exact rather than approximate. The enhanced FMAP equals the regular FMAP plus 30 percent of the gap between the regular FMAP and 100 percent. The gap between a 50 percent FMAP and 100 percent is 50 points, and 30 percent of that gap is 15 points, producing an enhanced rate of 65 percent. The gap between a 60 percent FMAP and 100 percent is 40 points, and 30 percent of that gap is 12 points, producing an enhanced rate of 72 percent. The gap between a 70 percent FMAP and 100 percent is 30 points, and 30 percent of that gap is 9 points, producing an enhanced rate of 79 percent. Each statement yields the same result, because shrinking the state share by a fraction and adding that fraction of the distance to full federal payment describe one movement along a scale from zero to 100.

Title XXI then bounded the enhanced rate with a floor of 65 percent and a ceiling of 85 percent. The floor guaranteed that no participating state received less than 65 cents of federal money for each dollar of covered spending under the program, whatever the income comparison produced, and a state with a 50 percent regular rate reached 65 percent by formula alone, exactly at the floor. The ceiling held payment at 85 percent for a state whose calculated rate rose above that level, leaving a state share of at least 15 percent. The compressed range, running from 65 percent to 85 percent across all states, was narrower than the Medicaid range running from 50 percent to 83 percent, and compression itself carried a policy meaning. Differences in fiscal capacity continued to matter, because a poorer state generally received a higher enhanced rate, but the differences mattered less than under Medicaid, because the floor lifted wealthy states far and the ceiling restrained the poorest states. For a program seeking participation from every state, compression was a sensible trade, since the states most tempted to stay out were wealthy states facing the lowest Medicaid rates.

Measured across states, the enhancement moved the average federal share from the high fifties under Medicaid to about 71 percent under the Children’s Health Insurance Program in the period before the increase enacted in 2010 took effect. A difference of roughly 15 percentage points, averaged nationally, translated into a substantial change in the state cost of a given coverage decision. Consider coverage that cost 10 million dollars in total spending. At a 55 percent Medicaid rate, the state share was 4.5 million dollars. At a 71 percent enhanced rate, the state share was 2.9 million dollars. The same coverage required 1.6 million dollars less in state appropriations under the enhanced rate, and a fixed state appropriation purchased considerably more total coverage under Title XXI than under Title XIX. The enhancement was calibrated to change that arithmetic without eliminating a state contribution altogether. A state contribution remained mandatory in structure, because federal payment arrived as a match on state spending, and a state that spent nothing drew nothing.

The enhancement, however, operated only inside a capped allotment, and that pairing was the central fiscal design of Title XXI. Federal payments to a state under the program could not exceed the state’s allotment, a sum distributed under a formula resting principally on the state’s share of low-income children and low-income uninsured children, with an adjustment for health care costs. Medicaid matching, by contrast, operated as an open-ended commitment to states, with federal payment following eligible spending without a state allotment ceiling of that kind. Title XXI inverted one half of the Medicaid bargain and preserved the other half. The rate was better than Medicaid, and the total was limited in a way Medicaid totals were not. Generosity and limit were the same mechanism seen from two sides, because the federal government could afford to offer a higher percentage precisely on spending that had a predetermined maximum, and a state could accept a maximum precisely because the percentage made each state dollar inside the maximum go further than a Medicaid dollar went. Remove the cap, and the enhanced rate became an open-ended exposure that the 1997 budget agreement was constructed to avoid. Remove the enhancement, and the cap offered states little reason to expand.

The pairing also disciplined the meaning of the enhanced rate in practice. Under Medicaid, each additional state dollar drew a fixed federal contribution across the whole range of eligible spending. Under Title XXI the promise held until a state reached the end of available allotment funds, subject to the redistribution arrangements under which unspent allotments flowed after a limited period of years toward states facing shortfalls. Beyond available funds, the marginal federal contribution fell away, and a state contemplating enrollment growth had to reckon with financing that growth at state expense, at a less favorable rate under another authority, or not at all. The statute also limited the share of drawn funds a state could devote to administration, outreach, and related non-coverage activities to 10 percent, pushing allotment money toward coverage itself. A budget officer therefore saw two numbers at once, a percentage that flattered expansion and a total that bounded expansion.

For governors and state budget officers, the fiscal calculus combined those two numbers with facts that every administration understood. Children were a comparatively inexpensive population to insure, because children used less medical care on average than older adults or people with disabilities, and the targeted population sat in families with income above Medicaid thresholds. The enhanced match lowered the state price of closing that gap when a state decided whether to file a plan, which income bands to cover, and whether to proceed through a Medicaid expansion, a separate child health program, or a combination of both. A state dollar committed to the program returned more federal money than a state dollar committed to Medicaid, inside a federal total that a budget office could treat as predictable for the allotment period, without the open-ended exposure that made Medicaid forecasting difficult. Predictability had independent value in states with balanced budget requirements, where an entitlement that grew faster than forecast forced cuts elsewhere or tax action. The ceiling sold the program to fiscal conservatives at the same time that the percentage sold the program to coverage advocates.

Hesitation at the state level tended to gather around the same two numbers read less favorably. A capped allotment meant that a successful program, enrolling children faster than projections, could exhaust federal support and confront a state with choices among raising state spending, restricting enrollment in a separate program where federal law permitted enrollment limits, or seeking relief through redistribution and later federal action. A high match meant that a state became accustomed to a federal contribution that later Congresses could reduce. Those risks were structural rather than hypothetical, and the funding lapse after September 30, 2017 demonstrated that federal support under Title XXI depended on periodic congressional action in a way Medicaid support did not. In ordinary operation before that lapse, many states did not spend full allotments, and the discipline of the cap expressed itself mainly through the appropriations calendar and the recurring need for reauthorization. Governors accepted a bounded federal promise on favorable terms, in preference to an unbounded state obligation at ordinary prices or no program at all.

The exchange became markedly more favorable for a defined four-year period under the Patient Protection and Affordable Care Act, Public Law 111-148, signed March 23, 2010. The 2010 health law raised the enhanced FMAP by 23 percentage points for expenditures from October 1, 2015 through September 30, 2019, and it provided that the increased rate could not exceed 100 percent. A flat addition of 23 points differed from the proportional enhancement of 1997, lifting each state by the same number of points until the 100 percent ceiling intervened and, for states with enhanced rates above 77 percent before the increase, eliminating the state share within available allotment funds altogether. A state at the 65 percent floor moved to 88 percent, bearing 12 cents of each dollar. A state at 72 percent moved to 95 percent, bearing 5 cents of each dollar. A state at 79 percent moved to the 100 percent ceiling, bearing no part of the matched cost while allotment funds lasted. Averaged across states, the federal share during the period from October 1, 2015 through September 30, 2019 was about 93 percent, far above the roughly 71 percent average under the standard enhanced formula and far above the high fifties average of regular Medicaid matching.

A federal share averaging about 93 percent changed state budget arithmetic without changing the underlying allotment structure. Because federal payment remained bounded by allotments, raising the percentage did not raise the maximum federal payment in proportion to the increase, and a state exhausted a fixed allotment with less total spending when each dollar drew more federal cents. From a state perspective, the increase operated largely as relief on the state share required to draw a given allotment, freeing state funds that had been budgeted as match, rather than as an equivalent expansion of total program resources. The 2010 law also required states to maintain eligibility standards, methodologies, and procedures in effect on March 23, 2010 for children through September 30, 2019, a maintenance of effort obligation that prevented states from answering the richer match by narrowing the population to which the match applied. The combination made the period an unusually protected one for children’s coverage in state budgets.

Congress addressed the end of the increase when long-term funding was restored in January 2018, after funding authority had expired at the end of September 30, 2017 and a lapse of 114 days had followed. The continuing resolution signed January 22, 2018, Public Law 115-120, carrying the HEALTHY KIDS Act, funded the program for six years, for fiscal 2018 through fiscal 2023, retroactive to October 1, 2017, and the statute wrote a step-down from the 23 point increase into law as an enacted schedule. Under that schedule, the increase was 11.5 percentage points for the year beginning October 1, 2019, exactly half the 23 point increase, applicable through September 30, 2020, and the standard enhanced formula applied from October 1, 2020, with no addition beyond the rate produced by the 30 percent reduction in the state share and the floor of 65 percent and ceiling of 85 percent. Stated strictly as the schedule enacted in January 2018, the sequence ran from 23 additional points through September 30, 2019, to 11.5 additional points for the following year, to the standard formula thereafter. A phased reduction gave budget officers dates and percentages for forecasts, in contrast to an abrupt return from an average federal share of about 93 percent to an average near 71 percent in one transition.

The history of the matching formula, from the Balanced Budget Act of 1997 through the schedule enacted in January 2018, traced a consistent logic under changing numbers. The regular Medicaid FMAP measured fiscal capacity through per capita income and confined the federal share between 50 percent and 83 percent. The enhanced FMAP lifted that share by reducing the state portion by 30 percent, confined the result between 65 percent and 85 percent, and placed the richer rate inside a capped allotment so that federal generosity per dollar coexisted with federal control over total dollars. The 2010 law added 23 points for four years beginning October 1, 2015, subject to a 100 percent ceiling, and raised the average federal share to about 93 percent for that period. The January 2018 extension then halved the addition for the year beginning October 1, 2019 and restored the standard formula from October 1, 2020, as an enacted schedule known to states in advance. The formula made states want in because the formula changed the price of a decision that legislatures had to pay for, and the cap made Congress willing to change that price because the cap limited what the changed price could cost.

Three Ways to Build the Program, and Why the Choice Decides the Entitlement

The Balanced Budget Act of 1997 did not create a single national children’s insurance product. Title XXI of the Social Security Act created a fund, a set of federal conditions, and a choice of institutional forms, and each state expressed the choice in a state plan submitted to the Secretary of Health and Human Services for approval. The plan was the instrument through which a state declared whom the program would cover, at what income levels, with what benefits, and under what cost sharing, and federal payment followed the approved plan rather than a freestanding federal promise to a child. A state could expand Medicaid with Title XXI money, could build a separate child health program outside Medicaid, or could combine the two approaches for different groups of children. The decision shaped the benefit package a child received, the amount a family could be charged, and whether an eligible child held a legally enforceable claim to coverage at all or held only a place in a program that a state could close when funds ran short.

The plan requirement gave the federal government a point of control without operational control of enrollment or delivery. A state described the design in the plan, the Centers for Medicare and Medicaid Services reviewed the plan against Title XXI, and amendments were required when a state changed eligibility, benefits, or cost sharing in material respects. Within federal boundaries, variation was intentional. Congress set a target population of uninsured low income children who were not eligible for Medicaid, set an enhanced federal matching rate to induce participation, and capped the federal contribution through state allotments. Every route drew the same enhanced match within the allotment, but the legal character of coverage diverged at the point of selection.

In a Medicaid expansion program, Title XXI money paid for Medicaid coverage itself. The child was enrolled in Medicaid, the state Medicaid plan governed the coverage, and the expenditure was matched at the enhanced rate in Section 2105(b) of the Social Security Act rather than at the regular Medicaid rate, so long as the expenditure fell within the state’s Title XXI allotment. The subsidy was richer than ordinary Medicaid financing, but the coverage delivered with that subsidy was Medicaid coverage in substance and in law. A state taking this route did not design a new benefit package under Title XXI standards and did not gain the flexibility on premiums and enrollment limits that Title XXI gave to a separate program. The state instead extended an existing entitlement program upward to children defined in the Title XXI plan. Eligibility determinations, hearing rights, benefit rules, and provider arrangements operated through the Medicaid framework administered since the enactment of Medicaid in 1965.

Children covered through a Medicaid expansion received the Medicaid benefit package, and for children that package included early and periodic screening, diagnostic, and treatment services, commonly known by the initials EPSDT. Federal Medicaid law described EPSDT as screening services at periodic intervals, diagnostic services when a screening indicated a need for further examination, and treatment services necessary to correct or ameliorate defects and physical and mental illnesses or conditions discovered through screening. Screening encompassed health and developmental history, physical examination, immunizations appropriate to age and health history, laboratory tests including lead toxicity screening where indicated, and health education. The treatment obligation reached beyond a narrow list of services in a state plan. Where a screening identified a condition, Medicaid law required provision of medically necessary services of a kind coverable under the federal Medicaid statute to correct or ameliorate the condition, even when the state plan for adults did not include the service in the same scope. For a child with a developmental delay or a complex physical need, that structure mattered more than the label on the eligibility card, because the inquiry began with medical necessity for the child rather than with resemblance to a commercial insurance package.

Cost sharing in a Medicaid expansion followed Medicaid limits rather than the more permissive Title XXI schedule for separate programs. Medicaid law in the period after the 1997 enactment confined charges for children to nominal amounts in most circumstances, barred premiums for most low income children, and exempted many children’s services from charges altogether, including preventive services associated with EPSDT screening. A family with a child in a Medicaid expansion faced a legal structure designed to keep financial barriers minimal. A family with a child in a separate program faced a structure in which Congress expressly authorized premiums, enrollment fees, deductibles, copayments, and coinsurance within federal ceilings. The enhanced match did not erase that difference, because the match described how governments shared the cost of coverage, not how much of the cost could be shifted to the family under the applicable title.

Medicaid is an individual entitlement program, and that character defined the expansion route. A child who met the eligibility criteria in the approved Medicaid plan held a claim to enrollment and to covered medically necessary services, and the state obligation did not end where a fixed allotment was exhausted. Title XXI modified the financing of an expansion child’s coverage when the allotment ran out, but Title XXI did not convert the expansion child into a beneficiary of a capped grant program for purposes of the right to coverage. If a state exhausted the Title XXI allotment available for Medicaid expansion children, coverage for those children continued, and federal payment continued under Title XIX at the regular Medicaid matching rate. The state share rose because the enhanced match no longer applied, but the child’s coverage survived the exhaustion of the Title XXI fund. The cap operated as a limit on the enhanced subsidy, not as a limit on the entitlement itself, and a shortfall produced a fiscal problem for the state rather than a waiting list for expansion children.

Does a child in a separate state program hold an entitlement to coverage?

No. A child who meets the eligibility rules for a separate program holds no individual entitlement to enrollment. The state may limit enrollment to the funds available under its Title XXI allotment, close enrollment, maintain a waiting list, and require a waiting period before coverage begins.

The separate program route allowed a state to build child health coverage outside Medicaid while drawing Title XXI funds under an approved Title XXI plan. A separate program had its own eligibility rules within federal parameters, its own benefit package under Section 2103 of the Social Security Act, its own cost sharing schedule under Section 2105, and, in many states, its own name and administrative identity distinct from Medicaid. Policymakers in the debates of 1997 argued that a program styled after private insurance would reach families who resisted a welfare associated medical program, and administrators and legislators valued the fiscal control that came from matching enrollment to the allotment with a directness that Medicaid law denied.

Section 2103 supplied the benefit standard for a separate program, and the standard was built by reference to coverage outside Medicaid. The statute required child health assistance through a separate program to consist of benchmark coverage, benchmark equivalent coverage, coverage that included the Medicaid benefit package, or coverage approved by the Secretary. Benchmark coverage meant coverage at least equal to one of three models named in the statute. The first model was the standard Blue Cross and Blue Shield preferred provider option plan for federal employees under the Federal Employees Health Benefits Program. The second model was a health benefits plan generally available to state employees. The third model was coverage through a health maintenance organization that held the largest insured commercial, non Medicaid enrollment in the state. Each model tethered a public children’s program to a package sold or provided in the employment based market, and the tether expressed a judgment in the 1997 law that adequate children’s coverage could be defined by resemblance to mainstream coverage rather than by adoption of the Medicaid child benefit in full.

Benchmark equivalent coverage added actuarial discipline to that reference system without requiring identity of benefits. A state could offer a package with an aggregate actuarial value at least equivalent to one of the benchmark packages, provided the package included coverage for categories of basic services specified in Section 2103. The basic categories included inpatient and outpatient hospital services, physician surgical and medical services, laboratory and x ray services, and well baby and well child care including age appropriate immunizations. Where the chosen benchmark included categories of additional services, a benchmark equivalent package had to include coverage in those categories with an actuarial value of at least 75 percent of the actuarial value in the benchmark. The additional categories named in the statute were prescription drugs, mental health services, vision services, and hearing services. An actuary certified the equivalence under federal rules, using a standardized population and standardized utilization and price factors. Secretary approved coverage gave a further path where a state proposed a package that met Title XXI objectives through a design that did not fit neatly into a benchmark comparison, subject to federal approval and to statutory requirements for basic services. Dental services appeared among the service categories a separate plan could cover in the original framework, while the stronger dental guarantee enacted in the reauthorization of 2009 belongs to the history of that later law and receives full treatment in the section devoted to reauthorization.

Comparison with EPSDT clarified what the benchmark system did and did not promise. The benchmark test measured actuarial value across a population and required inclusion of service categories, rather than imposing a child specific duty to furnish every medically necessary coverable service needed to correct or ameliorate a condition found on screening. Limits on therapy visits, exclusions for particular treatments, and narrower definitions of medical necessity could therefore operate in a separate program in ways that Medicaid EPSDT law constrained for a Medicaid child, and for a child requiring intensive therapy or specialized equipment the difference could decide whether a service was a covered benefit or an expense outside the program.

Title XXI permitted a state with a separate program to impose premiums, enrollment fees, deductibles, copayments, and coinsurance, subject to protections in Section 2105. States used premiums most visibly, often on a sliding scale that rose with family income, and enrollment fees appeared in some state designs in the early years of the program. Federal law imposed an aggregate ceiling. Total cost sharing for a family, including premiums and other charges under the separate program, could not exceed 5 percent of family income for the period to which the cost sharing applied. States were required to inform families of the ceiling and to apply procedures that prevented charges from passing that level. The rule placed an outer boundary on the shift of program cost to households.

A more specific protection governed children in the poorest families served by a separate program. For a child in a family with income below 150 percent of the federal poverty level, cost sharing could not exceed the amounts permitted under Medicaid, which meant nominal charges rather than commercial levels of copayment or coinsurance. Cost sharing could not be imposed for well baby and well child services, including age appropriate immunizations, and the schedule could not favor children from higher income families over children from lower income families. Above 150 percent of the poverty level, federal law allowed charges to rise, with regulatory limits tied to the cost of a service and with the aggregate 5 percent ceiling in force across the schedule. The result was a graduated system that preserved a Medicaid like floor for children closest to Medicaid eligibility and allowed a more insurance like contribution from families with higher incomes. Enforcement also differed from Medicaid practice in many separate programs, where failure to pay a premium could lead to disenrollment after required notices and grace periods under the approved plan.

Waiting periods formed another tool available in a separate program and illustrated the anti substitution purpose that ran through Title XXI. A state could require a period of uninsurance before a child became eligible, so that a family could not drop employer sponsored dependent coverage and move a child immediately into public coverage when private coverage remained available. States set waiting periods of varying lengths in the years after 1997, commonly measured in months, and paired waiting periods with exceptions for involuntary loss of coverage, loss of employment, expiration of coverage under federal continuation law, and other circumstances specified in the approved plan.

The absence of an individual entitlement gave legal meaning to each of those tools and to the allotment cap behind the separate program. Because Title XXI created a capped grant for a separate program rather than an entitlement for each eligible child, a state could limit enrollment when projected spending approached available funds. Federal law permitted enrollment controls, including closing enrollment to new applicants, freezing enrollment at a stated level, and maintaining a waiting list for children who met eligibility criteria but could not be admitted within available resources. A child on a waiting list held eligibility in an administrative sense without holding coverage. When federal funding authority lapsed at the end of September 2017, contingency planning in states with separate programs, including enrollment freeze announcements in Alabama and Connecticut in December 2017, demonstrated the operational consequence of that legal design under fiscal stress.

In a combination program, a state used a Medicaid expansion for one group of children and a separate program for another group under a single Title XXI framework. The typical division followed income, with Medicaid expansion coverage for children in families with incomes closest to the Medicaid thresholds and separate program coverage for children at higher incomes, where Congress had authorized more flexible benefits and cost sharing. Some divisions also followed age, reflecting Medicaid eligibility rules that had historically covered younger children at higher income levels than older children. Entitlement status followed the component in which a child sat. A child in the Medicaid expansion component held the Medicaid entitlement and the continuation protection at the regular match after allotment exhaustion. A child in the separate component held no such entitlement, even though both children drew support from Title XXI and both families encountered outreach materials that used a single program name in many states. States adopted combinations to reconcile competing objectives, preserving EPSDT depth for the lowest income band while gaining premiums, benchmark benefits, and enrollment controls for a higher band.

Premium assistance gave a further variant that cut across the distinction between direct public coverage and private coverage, and the variant requires careful statement because federal authority developed in stages. Title XXI allowed a state, with federal approval, to use program funds to help purchase coverage through an employer sponsored plan for an eligible child, in circumstances where use of the private plan satisfied Title XXI requirements and federal cost and benefit protections. The reauthorization law of 2009 strengthened and clarified premium assistance in Section 2105(c)(10) of the Social Security Act. Under the definition applied in that framework, qualified employer sponsored coverage was group health coverage through an employer that qualified as creditable coverage, was made available on a nondiscriminatory basis, and received an employer contribution of at least 40 percent of the premium. A state premium assistance subsidy under the 2009 framework paid the employee share of the premium attributable to coverage of the eligible child or, in defined circumstances involving a family cost effectiveness comparison, support for family coverage, and enrollment was voluntary for the family rather than a compulsory substitute for direct Title XXI coverage. Where an employer plan benefit package fell short of Title XXI requirements, wraparound coverage was required to fill the gap for the child under the rules described by the Congressional Research Service in analysis of the 2009 law.

New York enacted Child Health Plus through state legislation in 1990, and children began receiving coverage under the program in August 1991. Child Health Plus served uninsured children who were not eligible for Medicaid, initially concentrated on younger children, and operated as a state subsidized insurance product with premiums that varied by income. Federal accounts counted New York among the small group of state programs grandfathered into the new federal structure. New York used the federal enactment to expand Child Health Plus to children through age 18, and the program operated as a separate CHIP program alongside Medicaid, with income related premiums and no Medicaid entitlement for the Child Health Plus group. Subsidized eligibility later reached 400 percent of the federal poverty level in 2008. In Florida, the legislature established the Florida Healthy Kids Corporation in 1990. A pilot in Volusia County began in 1992, and the program combined public subsidies with family premiums on a sliding scale to purchase coverage for school age children through private insurers and managed care organizations. National accounts treated Florida Healthy Kids as a model for the federal program enacted in 1997, and the program was grandfathered into Title XXI. Healthy Kids continued as a separate program component within Florida KidCare, the umbrella structure that coordinated Medicaid for children, Healthy Kids, and related components under a common application and public identity, so that a family encountered a single doorway while the legal character of coverage depended on the component behind the doorway.

For a family, the state plan choice therefore decided far more than the letterhead on an enrollment notice. A parent who applied for coverage in a Medicaid expansion state, or for the Medicaid component of a combination state, received a determination under Medicaid rules, a benefit package anchored by EPSDT, minimal cost sharing, and a right to coverage that persisted through an allotment shortfall at a higher state fiscal cost. A parent who applied in a separate program received a determination under Title XXI plan rules, a benchmark or benchmark equivalent package, premiums or fees where the plan imposed them, a cost sharing ceiling of 5 percent of family income with Medicaid level protection below 150 percent of the poverty level, a possible waiting period after a loss or drop of private coverage, and no individual claim when the state closed enrollment for lack of available allotment funds. A move across a state line could thus change a child’s legal protection without any change in family income or medical need. Title XXI held both positions inside one statute, funded both with an enhanced match, and left the balance between security and control to the plan that each state filed and defended.

The three-option table

The table below summarizes the three structural options a state could choose under Title XXI and the consequences that followed from the choice for entitlement status, cost sharing, federal matching, and families.

Structural option	Entitlement status	Cost sharing permitted	Federal match	Consequences for families
Medicaid expansion	Individual entitlement under Medicaid	Nominal charges under Medicaid limits, with many children’s services exempt	Enhanced match within the allotment; regular Medicaid match after allotment exhaustion	Coverage continues after allotment exhaustion; Medicaid benefits including EPSDT; minimal financial barriers
Separate program	No individual entitlement; a state may cap enrollment and keep waiting lists	Premiums and cost sharing permitted; aggregate capped at 5 percent of family income; Medicaid level limits below 150 percent of the poverty level	Enhanced match within the allotment only	Benchmark benefits; possible waiting periods; enrollment may close when funds run short
Combination	Follows the component; entitlement in the Medicaid part and no entitlement in the separate part	Follows the component; Medicaid limits in one part and the Title XXI schedule in the other part	Enhanced match within the allotment; regular match protects the Medicaid part after exhaustion	Consequences depend on the band a child sits in; lower income children usually hold stronger protection

Reauthorization and Extension: How Congress Rebuilt the Funding Three Times

The first authorization of the Children’s Health Insurance Program carried a built in expiration date. Because Title XXI of the Social Security Act rested on capped allotments rather than on open ended appropriations, the program depended on Congress returning, voting new money, and restating the terms on which states drew federal funds. A reauthorization fight therefore involved how much the federal government would spend, which children the money would reach, how generously the treasury would match state spending, and how aggressively states would enroll children who qualified but had never signed up. By 2007, administrators had learned where the original allotment formula fit state experience and where the formula strained against that experience. The debate exposed a durable disagreement about the boundary between public coverage and private insurance, a disagreement that delayed renewal in 2007 and produced a far more detailed statute in early 2009.

The road to the 2009 law ran through a confrontation in the autumn of 2007 between a Democratic controlled Congress and President George W. Bush. Lawmakers assembled an expansion bill that would have reauthorized the program and added substantial new funding over five years, financed in large part through an increase in the federal cigarette tax. Supporters presented the measure as a means of covering several million additional uninsured children. The administration objected that the bill reached too far up the income scale, that public coverage would displace private coverage, and that the scope departed from the original focus on lower income families. On October 3, 2007, President Bush vetoed H.R. 976, the Children’s Health Insurance Program Reauthorization Act of 2007, stating that the legislation would move health care in the wrong direction. The House attempted to override the veto on October 18, 2007, and the attempt failed, leaving the program to operate under short term extensions.

A second bill followed the failed override, revised to answer objections from the first round while preserving the commitment to a large expansion. The revisions tightened the income ceiling at the center of the dispute, addressed the treatment of adults in a program written for children, and addressed eligibility documentation. The White House signaled that the second measure retained flaws serious enough to warrant another veto. On December 12, 2007, President Bush vetoed the second expansion bill as well. Two vetoes in a single autumn made the political geometry plain. A bipartisan majority in Congress favored a considerably larger program, and the House had fallen short of the two thirds threshold required for an override, so enactment required either a change in the White House or a compromise scaled closer to the administration proposal for a far smaller increase. Congress kept the program funded through temporary measures and carried the argument into the 2008 election year. Much of the architecture debated in the vetoed bills reappeared in the bill that a new president signed within weeks of taking office.

What did the Children’s Health Insurance Program Reauthorization Act change?

The Children’s Health Insurance Program Reauthorization Act of 2009 renewed federal funding through fiscal 2013, rebased state allotments on actual spending, added enrollment bonuses and outreach grants, created a contingency fund, guaranteed dental coverage, strengthened premium assistance, limited enhanced matching above 300 percent of poverty, and offered states a Section 214 coverage option.

The statute behind that summary, Public Law 111-3, was signed on February 4, 2009, by President Barack Obama, less than a month after inauguration. Children’s coverage was treated as settled enough in purpose to move first, on a separate track, with funding that ran from fiscal 2009 through fiscal 2013. The 2009 act supplied predictability after a period of stopgap extensions and redesigned how the money was divided, how extra money could be earned, and what states owed children in return for federal support. Public Law 111-3 kept the 1997 framework of state choice among a Medicaid expansion, a separate program, or a combination of the two, while rewriting the financial incentives inside that framework.

The most consequential mechanical change concerned allotments, the capped amounts that defined the maximum federal payment available to each state. Under the approach inherited from 1997, allotments rested heavily on estimates of low income children and low income uninsured children in each state, adjusted for health care costs, a method that depended on projections made before state programs had matured. By 2009, Congress had a decade of spending records, and the reauthorization rebased allotments on actual state spending. Rebasing shifted the formula from an estimate of need in the abstract toward demonstrated use, so a state that had built a large program entered the new period with an allotment anchored in that experience. Supporters argued that the change reduced the mismatch between paper allocations and real program costs. The change did not convert the program into an entitlement to federal funds without limit, and the cap remained the defining difference between Title XXI and Medicaid. Rebasing instead recalibrated the cap.

Performance bonus payments added a second financial layer, aimed at eligible children who remained unenrolled. Administrators had long identified a gap between eligibility and participation, driven by application burdens, renewal procedures that cut off coverage after paperwork failures, uneven outreach, and lack of awareness that a program existed. The 2009 act attacked that gap with money rather than instruction alone. A state became eligible for a bonus only by satisfying two conditions together. First, the state had to exceed a child enrollment target, expressed as a baseline derived from earlier Medicaid child enrollment and adjusted for growth in the child population and for additional percentage points specified in the statute. Second, the state had to adopt specified enrollment simplification measures from a statutory list intended to make enrollment and retention easier for families. Neither condition sufficed on its own. Growth without procedural reform earned no bonus, and reform without growth beyond the target earned no bonus.

The simplification measures reflected a detailed judgment about why eligible children missed coverage. Practices in the statutory family included eliminating face to face interview requirements, relying on information already held by other public programs, easing asset tests and documentation demands, permitting presumptive eligibility so a child who appeared eligible could receive coverage while a final determination proceeded, and adopting renewal procedures that reduced procedural terminations for children who remained eligible. Federal administrators later described qualifying states as having adopted at least five of eight enrollment and retention provisions specified under the program, a threshold that left states some choice while requiring breadth rather than a single token reform. The bonus was calculated by reference to the state share of costs for additional children above the baseline, with a higher share paid for growth at or beyond ten percent above the baseline. A guardrail limited the incentive to children who could have met eligibility criteria in place on July 1, 2008, so a state could not manufacture a bonus by broadening eligibility and counting the newly eligible population.

Outreach and enrollment grants formed the companion investment to the bonuses. Bonuses rewarded success after the fact, while grants financed the work that made success possible. The 2009 act funded grants for outreach and enrollment activities directed at children who qualified for Medicaid or for the Children’s Health Insurance Program but lacked coverage. Eligible recipients extended beyond state agencies to include community organizations, schools, health care providers, and other entities positioned to reach families in places families trusted. That breadth reflected experience from the first decade, in which enrollment gains often followed from school based campaigns, clinic based application assistance, and partnerships with organizations serving immigrant and rural communities. A dedicated federal grant stream treated locating uninsured eligible children as a national objective.

The Child Enrollment Contingency Fund addressed a different risk, the risk that success itself would strain a state budget. Under a capped allotment system, a state that enrolled children faster than projected, whether because of an economic downturn, an effective outreach campaign, or a change in employer coverage, could exhaust an allotment before the fiscal year ended. Such a state faced the choice of freezing enrollment, cutting benefits, or covering costs at a less favorable match. The contingency fund created a federal reserve for states whose enrollment outran allotments, available under conditions that tied access to genuine enrollment driven need rather than to general fiscal distress. The fund softened the hardest edge of the cap without removing the cap. Congress preserved the discipline of a fixed national appropriation while acknowledging that child enrollment responds to economic conditions outside a state budget office’s control. Rebasing set a more accurate starting line, redistribution moved idle money toward states in need, and the contingency fund supplied a buffer when surges exceeded what redistribution alone could cover.

Coverage policy changed alongside finance, and the dental guarantee illustrated the shift. Dental disease ranked among the most common chronic conditions of childhood, and untreated decay interfered with eating, speaking, school attendance, and overall health. Coverage for dental services in separate state programs had varied in ways that left gaps between a formal promise of child health coverage and the services a child received. The 2009 act guaranteed dental coverage, requiring state programs to include dental services necessary to prevent disease and promote oral health, restore oral structures to health and function, and treat emergency conditions. The guarantee applied as a defined expectation rather than as an optional enhancement that a state could drop when budgets tightened. The act also paired the guarantee with attention to information for families about participating dentists, because a benefit on paper delivered little value if parents could not locate a provider willing to treat a child under the program.

Premium assistance received renewed emphasis for a related reason, the long standing debate over the boundary between public programs and employer sponsored insurance. Premium assistance allowed a state to use program funds to help a family purchase private coverage, typically employer sponsored coverage that included the child, by subsidizing the family’s share of the premium. Where required, the state wrapped around that coverage with supplemental benefits and cost sharing protections so the child received protection comparable to direct program coverage. Advocates presented the approach as a way to keep a family in a single plan, to support rather than supplant an employer’s contribution, and to answer the crowd out concern that public expansion would mainly pay for children whose families or employers would otherwise have paid. Skeptics asked whether private plans offered benefits as complete for children as public coverage and whether wraparound administration justified the complexity. The 2009 act strengthened premium assistance within guardrails, without making the approach mandatory.

The treatment of higher income coverage supplied the clearest example of Congress trimming at one margin while expanding at others. Eligibility above 300 percent of the federal poverty level had drawn scrutiny from the earliest debates, because coverage at that range intensified questions about substitution for private insurance and about the proper target population for a program written for targeted low income children. The 2009 act reduced the federal match to the regular rate for children above 300 percent of the poverty level, replacing the enhanced matching rate with the ordinary Medicaid matching rate for expenditures in that band. The change lowered the federal subsidy for the highest income segment of the caseload and raised the state price of maintaining eligibility at that level, while leaving enhanced support intact below the threshold. Exceptions softened the transition for states that already held approved plans or enacted laws covering children above that level. States had made commitments to families under earlier federal approvals, and an abrupt change would have disrupted coverage and budgets planned in reliance on those approvals.

Section 214 of the 2009 act requires precise description, because shorthand accounts often overstated the change. The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 imposed a five year waiting period before many lawfully residing immigrants became eligible for federal means tested benefits, and that waiting period carried into children’s coverage and into coverage for pregnant women in many circumstances. Section 214 gave states the option to cover lawfully residing children and pregnant women without the five year waiting period. The provision concerned lawfully residing persons, present in the United States under immigration law, and no part of the option extended coverage to persons without lawful status. The provision operated as a state option. A state could elect coverage for the affected children, for pregnant women, or for both groups, and a state could also decline the option and retain the waiting period within the federal framework. No nationwide removal of the bar occurred on February 4, 2009. Congress instead transferred a decision that federal law had previously made uniformly to the states individually. For families in a state that adopted the option, the effect could be immediate coverage for a lawfully residing child or for prenatal care during a pregnancy that fell inside the former waiting period. For families in a state that declined, the former waiting period remained the governing rule.

That settlement lasted only a year before Congress folded children’s coverage into a far larger statute. The Patient Protection and Affordable Care Act, Public Law 111-148, signed on March 23, 2010, reorganized much of the American health financing system, creating new marketplaces, new subsidies, and a major expansion of Medicaid eligibility for adults. The act treated the Children’s Health Insurance Program as a component to preserve through the transition rather than as a program to absorb immediately into the new structure. For funding, the 2010 law extended the program through fiscal 2015, adding two years beyond the horizon set in 2009 and bridging the period during which the new coverage systems were built and launched. Continuity mattered because families, providers, and state agencies faced enough change from the new law without a simultaneous lapse in the children’s program.

The maintenance of effort requirement supplied the legal lock that accompanied the funding extension. States were required to maintain the Medicaid and Children’s Health Insurance Program eligibility standards, methodologies, and procedures in effect on March 23, 2010, for children through September 30, 2019. By keying the obligation to the standards in force on the day the 2010 law was signed, Congress froze a baseline that no state could lower for children during the specified period without risking federal consequences. The requirement covered standards, methodologies, and procedures alike, so a state could not accomplish indirectly through paperwork burdens what the requirement barred directly through a lower income threshold. The obligation ran for children through September 30, 2019, well beyond the fiscal 2015 funding horizon and well beyond the opening of the new marketplaces. The span reflected a judgment that children’s coverage should not become the release valve for state budgets during the first years of the reformed system. A separate provision of the 2010 law raised the enhanced matching rate by 23 percentage points for a period beginning in fiscal 2016, and that increase made the maintenance obligation less expensive for states to honor. Together, the funding extension and the maintenance of effort converted the 2009 settlement into a protected bridge, financed through fiscal 2015 and legally guarded through the end of fiscal 2019.

The third rebuilding arrived in a statute whose center of gravity lay outside children’s health. The Medicare Access and CHIP Reauthorization Act of 2015, Public Law 114-10, signed on April 16, 2015, permanently replaced the sustainable growth rate formula that had governed updates to Medicare physician payment and had generated a recurring cycle of scheduled cuts and temporary reprieves. Children’s coverage rode in that vehicle as a widely supported companion provision. Public Law 114-10 funded the Children’s Health Insurance Program for fiscal 2016 and fiscal 2017, maintaining the allotment approach and carrying forward the 23 percentage point increase in the enhanced matching rate scheduled under the 2010 law for the period that included those years. The extension arrived months before the September 30, 2015, expiration of the funding supplied in 2010. An extension enacted in April for a deadline in September allowed states to write budgets, renew contracts, and communicate with families without contingency warnings, and that calm was itself a policy outcome.

What distinguished the 2015 extension from the 2009 reauthorization was deliberate restraint. Public Law 114-10 left the program’s policies largely untouched. Congress did not rewrite the allotment formula in 2015, did not replace the benefit framework, did not reopen the settlement on higher income matching, and did not revisit the state option structure created by Section 214. Legislative energy concentrated instead on the Medicare payment question, on the offsets chosen to pay for parts of the package, and on the length of the children’s funding extension, with proposals for a longer extension considered before the two year period prevailed. For the Children’s Health Insurance Program, the absence of major policy change functioned as an affirmative signal. A program that had been contested terrain in 2007, vetoed twice over the proper reach of public coverage for children, was treated in 2015 as established infrastructure whose rules required funding rather than redesign. The argument had shifted from whether the federal government should finance coverage for children in working families above Medicaid levels to how long each extension should run.

That signal had limits, visible in the structure Congress left in place in 2015. A two year extension preserved the program and preserved the expiration mechanism that had defined the program since 1997. Every protection associated with the 2009 and 2010 settlements, including the maintenance of effort obligation for children running through September 30, 2019, remained anchored to dates, and every dollar of allotment authority depended on Congress acting again before the next deadline. The continuity of 2015 therefore differed in kind from permanence. Medicaid, resting on an open ended appropriation and an individual entitlement, required no comparable rescue on a fixed calendar. The Children’s Health Insurance Program, even after three rebuildings of the funding in 2009, in 2010, and in 2015, remained a program that Congress had to choose again at intervals. Each interval offered an occasion for delay when children’s coverage became entangled with unrelated disputes. The years that followed the 2015 enactment tested whether broad support would reliably arrive before funding expired, a test whose results belong to the account of the lapse and restoration that closed the decade.

The Lapse: What a Funding Deadline Does to a Capped Program

Federal funding authority for the Children’s Health Insurance Program expired at the end of September 30, 2017. The Medicare Access and CHIP Reauthorization Act of 2015, Public Law 114-10, had funded the program for fiscal 2016 and fiscal 2017 with the program’s basic policies unchanged, and no successor law took effect before the fiscal 2018 year opened on October 1, 2017. The expiry therefore arrived on a date fixed years in advance and visible on every state budget calendar, rather than through a sudden administrative action or a court order. A program built on capped federal allotments met the precise condition that the cap structure made consequential, a lapse in the authority to provide new federal money, and the months that followed tested how a capped program behaves when the calendar, rather than enrollment or medical cost in a single state, becomes the binding constraint.

The design explains both the delay before any closure and the certainty that delay had a limit. Under Title XXI of the Social Security Act, federal payments to a state cannot exceed the state’s allotment, and allotments rest principally on the state’s share of low-income children and low-income uninsured children, with an adjustment for health care costs. In most years before the lapse, many states did not spend full allotments, and unspent amounts fed a redistribution mechanism that directed money toward states facing shortfalls. That history mattered in October 2017. The absence of new funding authority did not erase balances already allotted, and it did not halt spending from balances that states held when the deadline passed. The program entered the lapse with a cushion accumulated from the same feature that defined its ordinary operation, underspending in some states relative to allotments large enough to leave money available for reallocation.

What kept state programs running after federal funding authority expired?

Unspent state allotments from earlier years and federal redistribution of remaining balances toward states projected to exhaust funds first kept programs operating after September 30, 2017. States drew down carryover balances while the Centers for Medicare and Medicaid Services reallocated available money, postponing closures without restoring long term funding authority.

The cushion operated through two channels that reinforced each other for a time. The first channel sat in state accounts, in the form of allotment balances carried into fiscal 2018. A state that had spent less than its allotment in a prior period retained amounts available for program spending, subject to the limited period of years during which allotments remained available, and administrators drew on those amounts to pay for coverage in October, November, and December 2017. The second channel sat at the federal level. The Centers for Medicare and Medicaid Services redistributed remaining funds toward states projected to run short first, so that the scarce resource moved in the direction of earliest need rather than remaining locked in the accounts of states with slower spending. Health Affairs accounts of the period described twenty states needing redistribution funds to cover program expenditures in the first quarter of fiscal 2018, from October 1 to December 31, 2017, a figure that measured both the breadth of reliance on the mechanism and the speed at which ordinary balances proved insufficient in a substantial minority of states.

Redistribution softened the deadline without removing it, because redistribution moves a finite stock rather than creating a flow. Every dollar sent to a state projected to exhaust funds early was a dollar drawn from a pool that received no replenishment after September 30, 2017, apart from the December 2017 stopgap described below. States also spent at different rates, for reasons rooted in program structure. A state running a separate child health program carried the full cost of that design within the allotment system, while a state using Title XXI funds to expand Medicaid held a different position, because children in a Medicaid expansion held an individual entitlement and, if the Title XXI allotment was exhausted, coverage for those children continued with federal Medicaid matching at the regular federal medical assistance percentage. Children in a separate program held no individual entitlement, and a state administering a separate program could cap enrollment, keep waiting lists, and impose waiting periods within the governing rules. The lapse therefore pressed unevenly across the map, hardest on separate programs dependent entirely on a wasting pool and less directly on expansion children whose coverage had a Medicaid fallback at a lower federal match.

The wasting character of the cushion shaped state planning from the opening weeks. Administrators could not treat carryover balances as a substitute appropriation, because balances declined with each month of claims, each premium payment to a managed care organization, and each administrative cost charged within the statutory limits. Federal estimates turned that arithmetic into dates. Even after Congress added money in December 2017, federal officials estimated that some states could begin exhausting funds around January 19, 2018, the date attached in federal communications to the earliest projected exhaustion. The Congressional Budget Office’s state-by-state estimates pointed in the same direction, indicating that the added money would carry some states only weeks. A date of that kind carries operational weight well before it arrives, because a state preparing to close enrollment or terminate coverage must reprogram eligibility systems, notify managed care organizations and providers, and give families notice in time to seek other coverage, and each of those steps requires a decision made on projections rather than on an empty account confirmed after the fact.

Congress supplied the first relief in a stopgap enacted on December 22, 2017, Public Law 115-96, which added 2.85 billion dollars for the program and adjusted redistribution. The amount extended operation without settling duration. The law functioned as a bridge sized to keep programs running into early 2018, and contemporary federal and analyst assessments agreed that the bridge had uneven length across states. States with larger balances and slower spending could project coverage farther into the winter, while states that had required redistribution in the first quarter faced exhaustion dates clustered near the January 19, 2018 estimate. The stopgap also changed state calculations in a second way, by demonstrating that a further law was possible without demonstrating that a further law was assured. Alabama suspended planned restrictive actions after the stopgap, as detailed below, while Connecticut maintained an enrollment freeze adopted in the last week of December 2017, and the contrast between those responses captured the range of judgments administrators made about money that extended a deadline by weeks rather than by years.

The state contingency record gives the lapse a concrete administrative history. Eight states posted announcements or sent notices warning families that coverage could end on a stated date, according to a Health Affairs account of the period. The eight states were Alabama, Colorado, Connecticut, Oklahoma, Utah, Virginia, Washington, and West Virginia. Within that group, Colorado, Connecticut, and Virginia sent letters to families, a step reported by Kaiser Health News in January 2018, and letters of that kind converted a congressional funding dispute into a household document that named a date on which coverage for a child might end. Announcements and letters served distinct functions. A posted announcement warned the broader public, providers, and enrollment assisters that a program faced a funding limit, while a letter to an enrolled family triggered practical decisions about appointments, prescriptions, and applications for other coverage through Medicaid or through the health insurance exchanges established under the Affordable Care Act.

Alabama supplied the most fully dated sequence among the states that announced restrictive steps. Before the December 2017 stopgap, Alabama announced that enrollment would freeze on January 1, 2018 and that the program would shut down on January 31, 2018. After Congress enacted the stopgap on December 22, 2017, Alabama suspended both plans. The sequence matters because it shows contingency planning operating in both directions, toward closure when projections showed exhaustion within weeks and away from closure when new federal money lengthened the projection, even though the added money did not restore long term authority. Connecticut followed a different course in the same window. Connecticut froze enrollment in the last week of December 2017, after the stopgap had passed, a decision that reflected a state judgment that the added funds did not remove the risk of exhaustion in early 2018. An enrollment freeze protects funds for children already enrolled at the cost of turning away newly eligible children, and in a separate program design the freeze rested on the absence of an individual entitlement for children in that route.

The other states in the group of eight, Oklahoma, Utah, Washington, and West Virginia, together with Colorado and Virginia, completed a record that was national in spread without being uniform in form. Notices differed because programs differed, in eligibility levels, in the balance between Medicaid expansion and separate program enrollment, in allotment balances entering October 2017, and in state law governing notice and termination. The shared element was the source of pressure. None of the eight actions responded to a state decision to reduce eligibility in ordinary course, and none responded to a federal finding that a state plan failed to comply with Title XXI. Each action responded to a projection that federal money available to the state would reach a limit on a date that administrators could estimate and families could read. The lapse thereby revealed a property of capped finance that ordinary years concealed, because in ordinary years the cap disciplined spending mainly through the appropriations calendar and periodic expiry, funded generously relative to need and repeatedly increased, rather than through mid-year exhaustion in most states.

Restoration stalled through the autumn of 2017 over financing, after agreement on duration proved easier to reach in committee than agreement on offsets proved to reach across chambers. In October 2017 the Senate Finance Committee approved a five year extension of program funding. The committee action established that a bipartisan extension text existed in the Senate, and the same action left the central dispute unresolved, because the committee approved the extension without agreement on offsets to pay for it. A five year extension without an agreed financing path could advance as a statement of policy and stall as legislation, since floor consideration in the Senate required a resolution of the cost question that the committee had deferred. The House moved a financed bill and, in doing so, relocated the dispute rather than settling it.

The House passed an extension in November 2017, the CHAMPIONING HEALTHY KIDS Act, on November 3, 2017. The House bill paired the extension with financing that drew from the Affordable Care Act’s Prevention and Public Health Fund and from higher Medicare premiums for high income beneficiaries. House supporters presented that financing as the offsets that the Senate committee bill lacked, while Senate Democrats opposed the House financing, objecting to the use of the Prevention and Public Health Fund and to the Medicare premium provision as sources for a children’s coverage extension. Each position attached to named holders in the record of the period. The Senate Finance Committee majority and minority had joined in approving an extension without offsets in October 2017, the House majority passed an extension with the stated offsets in November 2017, and Senate Democrats withheld support from the House approach because of those offsets. Between November 2017 and January 2018, the disagreement was therefore not over whether the program would receive funding in principle, a point on which committee and floor votes showed broad support, but over which other federal commitments would absorb the budgetary cost of a decision that both chambers had recorded support for in some form.

A contrast with the fixed welfare block grant clarifies what kind of failure the lapse represented. The welfare block grant created by the 1996 welfare law pays states a fixed amount that does not rise automatically when a recession increases need, so strain in a downturn appears as stretched dollars, longer waits, and state choices about rationing a grant whose nominal size holds while caseload pressure grows. The grant’s weakness in a downturn is rigidity in level. The children’s program at expiry displayed a different weakness, fragility in duration. During the lapse, funding did not merely fail to grow with need, it approached a legal end, with a finite pool of prior balances standing between ordinary operation and closure dates that states announced in writing. A fixed block grant in a recession keeps flowing at the set level and forces accommodation inside a continuing structure, while a capped allotment program past a funding deadline consumes a stock and forces preparation for termination of the structure unless Congress acts. The two designs share a cap and differ in the event that exposes it, rising need in one case and a passing date in the other, and the winter of 2017 to 2018 supplied the date driven version in full.

Budget estimates in January 2018 altered the terms on which the financing dispute had rested. The Congressional Budget Office estimated in January 2018 that extending the program would reduce federal deficits by roughly 6 billion dollars over the decade under a ten year extension, because the alternative coverage cost the federal government more. The reasoning ran through the exchanges. Children who lost program coverage and gained subsidized exchange coverage would draw premium tax credits, and after repeal of the individual mandate penalty the Congressional Budget Office projected higher exchange premiums and therefore higher federal subsidy costs per subsidized enrollee, alongside additional enrollment by family members in exchange plans. Earlier in January 2018 the same office had estimated that a five year extension would increase deficits by about 800 million dollars, a sharp reduction from an October 2017 estimate above 8 billion dollars, and the ten year estimate crossed into deficit reduction because savings in the later years, when program funding at the standard enhanced match replaced more expensive subsidized coverage, outweighed costs in the earlier years. The estimates did not enact a law, and the office presented them as projections resting on stated assumptions about premiums, enrollment, and mandate repeal taking effect in 2019, but the figures removed the premise that a long extension required offsets of matching size.

Restoration arrived on January 22, 2018, in Public Law 115-120, a continuing appropriations law that carried the HEALTHY KIDS Act. The law funded the program for six years, fiscal 2018 through fiscal 2023, and the funding applied retroactive to October 1, 2017, so that the fiscal year that had opened under a lapse closed the gap in authority back to its first day. Allotments under the restoration began at 21.5 billion dollars for fiscal 2018 and rose to 25.9 billion dollars for fiscal 2023. The lapse in long term funding had lasted 114 days, counted from October 1, 2017 to the January 22, 2018 enactment, a period that included the December 2017 stopgap and the January 19, 2018 exhaustion estimate for the earliest states, both of which fell inside the interval rather than ending it. Six years of funding represented the longest extension enacted for the program to that date, and retroactivity addressed a practical problem created by the interval, the status of state spending undertaken from balances and redistributed funds while no long term authority existed.

A second law completed the extension less than three weeks later. The Bipartisan Budget Act of 2018, Public Law 115-123, signed on February 9, 2018, added four more years of funding, fiscal 2024 through fiscal 2027, for a ten year extension in total when combined with the six years in Public Law 115-120. The February 2018 law converted a six year restoration into a decade of allotment authority, and the Congressional Budget Office’s January 2018 finding that a ten year extension would reduce deficits by roughly 6 billion dollars supplied the budgetary context in which that addition moved. States that had written closure letters, frozen enrollment, or announced shutdown dates in December 2017 and January 2018 administered programs in February 2018 under authority extending to September 30, 2027, and the distance between those two positions measured the effect of the two enactments more plainly than any statement of relief issued after the fact.

The restoration laws also redrew the maintenance of effort obligation that the Affordable Care Act had imposed. The Affordable Care Act had required states to maintain the Medicaid and children’s program eligibility standards, methodologies, and procedures in effect on March 23, 2010, for children through September 30, 2019. The 2018 extensions prolonged the maintenance of effort through fiscal 2027 and narrowed its reach on a stated date. From October 1, 2019, the obligation protected eligibility only for children in families at or below 300 percent of the federal poverty level. A state could, from that date forward, roll back eligibility above that level and continue to draw federal funds, while eligibility at or below that level remained protected by the maintenance requirement through fiscal 2027. The narrowing aligned the obligation with a threshold that had appeared in program law before, in the Children’s Health Insurance Program Reauthorization Act of 2009, which had reduced the federal match to the regular rate for children in families above 300 percent of the poverty level, with exceptions for states that already held approved plans or enacted laws, and the 2018 change left higher income eligibility as a state option financed without the protection of a federal maintenance mandate.

The lapse left a record that later funding debates could consult in concrete form. Eight states had posted warnings or sent notices, three states had sent letters to families, one state had announced and then suspended a dated freeze and shutdown, and one state had frozen enrollment during the final week of December 2017. Federal officials had placed the earliest projected exhaustion around January 19, 2018, and Congress had answered with a 2.85 billion dollar stopgap on December 22, 2017, a six year restoration on January 22, 2018, and a four year addition on February 9, 2018. No state program closed on October 1, 2017, and no closure was required to prove that the deadline had force, because the administrative steps taken in anticipation of exhaustion, the letters, freezes, and announced dates, were themselves the mechanism by which a funding deadline acted on a capped program. The cap had operated quietly in years when allotments exceeded spending in many states, and it operated audibly in the winter of 2017 to 2018, when expiry converted undistributed balances into a countdown that each state read at a different speed and that only new law could stop.

What the Coverage Record Shows and What the Program Never Promised

The Children’s Health Insurance Program belonged to a longer sequence of federal health coverage statutes, and that sequence defined the work the program could plausibly claim. Medicare and Medicaid came first in 1965, Medicaid expansions for children followed across later decades, the children’s program arrived in 1997 as a capped grant for children above Medicaid income thresholds, and the Patient Protection and Affordable Care Act of 2010 then reshaped the system around the program while extending its funding and raising its matching rate for a defined period. The program was an addition to a layered system, built for a group the earlier statutes had left exposed, and the record after 1997 reflected the combined movement of all of those layers.

The measured decline in child uninsurance was steep, and each figure belongs to the survey that produced it and the years that survey covered. The National Health Interview Survey, conducted for the National Center for Health Statistics, recorded 13.9 percent of children under age 18 as uninsured at the time of interview in 1997, 10.8 percent in 2001, and 8.9 percent in 2005. Reporting in September 2017 summarized figures from the Medicaid and CHIP Payment and Access Commission as a fall from 13.9 percent in 1997 to 4.5 percent in 2015. The Bipartisan Policy Center, in its published summary of survey data, expressed the change in counts and stated that uninsured children fell from about 9.9 million in 1997 to about 3.5 million in 2015. Neither presentation supported crediting one program acting alone.

Scale accompanied the rate. During the funding lapse of late 2017 and early 2018, reporting and advocacy statements placed enrollment at roughly nine million children, and an October 2017 report from the Center on Budget and Policy Priorities, drawing on the Georgetown University Center for Children and Families, put enrollment at 8.9 million children. A joint statement in August 2017 by the American Academy of Pediatrics and seven other child health organizations, including the Children’s Defense Fund and the March of Dimes, credited the program together with Medicaid with helping to reduce uninsured children by 68 percent since 1997. The joint attribution in a statement from the program’s strongest advocates mattered, because the organizations described the gain as a shared product rather than the isolated achievement of Title XXI.

Honest attribution follows from eligibility itself. Children moved between Medicaid and the children’s program as family income rose and fell, as states set different thresholds, and as states chose Medicaid expansion coverage, separate coverage, or a combination. Medicaid expansions widened the entitlement base underneath the capped program. Outreach grants and performance bonus payments created by the Children’s Health Insurance Program Reauthorization Act of 2009 raised participation among children eligible for either program. The 2010 health law held state Medicaid and children’s program eligibility standards in place through September 30, 2019, expanded Medicaid for adults in participating states, and added subsidized exchange coverage. No serious account assigned the fall from the 1997 baseline to one statute. The decline belonged jointly to the children’s program, to Medicaid expansions for children, and later to the 2010 health law.

Coverage counts did not establish that enrolled children received care, so the access evidence requires careful statement. The evaluation mandated by Congress in the 2009 reauthorization law, conducted by Mathematica Policy Research with partners and reported in a supplement to Academic Pediatrics in 2015, found that for nearly every measure of access, use, and cost examined, enrolled children fared better than uninsured children. The same evaluation found access similar to private coverage for most measures, financial burdens substantially lower than under private coverage, and access to weekend and nighttime care weaker than under private coverage. The Medicaid and CHIP Payment and Access Commission, in a March 2017 report, stated that children covered through the program were more likely to have a usual source of care, including dental care, and more likely to have had a well-child visit in the past year than children without insurance. Stated at the strength the sources supported, enrolled children gained a usual source of care at rates approaching privately insured children, while uninsured children trailed both groups.

The same evaluation tradition recorded limits. The interim report of the earlier congressionally mandated evaluation, prepared for the Office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services, described access in the study states as generally good while identifying dental care, specialist care, and rural areas with provider shortages as more problematic. A study by David M. Rubin and colleagues at the Children’s Hospital of Philadelphia, published in JAMA Pediatrics in 2015 and based on the National Survey of Children’s Health for 2003, 2007, and 2011 to 2012, found preventive visits at higher adjusted rates for children in Medicaid or the children’s program than for privately insured children in low-income and moderate-income households, while caregivers across insurance types reported difficulty reaching specialty care. Enrollment moved children close to private coverage on primary and preventive measures, with specialty, dental, and after-hours weaknesses documented through the middle of the 2010s.

The sharpest empirical dispute concerned substitution, often called crowd-out, and the dispute turned in part on definition. The Congressional Budget Office, in a May 2007 report, concluded under a broad definition that between 25 percent and 50 percent of enrolled children had previously held private health insurance, a range the Center on Budget and Policy Priorities cited in its analysis of the literature. A broad definition counted children whose families had lost private coverage or had found private coverage unaffordable before enrollment. Analysts at the Heritage Foundation argued in 2007 that substitution grew as eligibility rose into income groups where private coverage was more common, and critics relying on the Budget Office range treated substitution as public spending replacing private spending. Supporters answered that a child whose parent had lost employer coverage, or whose premium had become unaffordable, had not been crowded out of a functioning private option in any practical sense.

State-level and ten-state studies pointed toward smaller direct substitution, and the contrast was methodological as much as numerical. A ten-state analysis from the congressionally mandated evaluation, published by Anna Sommers, Stephen Zuckerman, Lisa Dubay, and Genevieve Kenney in Health Affairs in 2007 under the title Substitution of SCHIP for Private Coverage, formed the basis for the finding, cited by the Center on Budget and Policy Priorities, that 93 percent of children who had held private coverage before enrolling had moved after losing that coverage or after that coverage had become unaffordable. State surveys of enrollees, summarized in the scholarly literature, likewise suggested that relatively few enrollees had transferred directly from employer coverage and that most longer-term enrollees had no access to an employer plan through a parent. Other researchers reached larger estimates by different methods, and no settled substitution figure gained acceptance across the Congressional Budget Office, the academic journals, and the state evaluations.

Did the capped structure fail the children it was meant to cover?

No, on the record through 2018. The cap produced a funding lapse, warnings, and enrollment freezes in late 2017, yet no state program closed outright, balances and redistribution carried programs forward, and Congress restored funding in January 2018. The structure created risk without, in that episode, ending coverage nationally.

The fuller record behind that answer began with dates. Federal funding authority expired at the end of September 30, 2017, and the lapse in long-term funding lasted 114 days before a continuing resolution signed on January 22, 2018 restored funding for six years, retroactive to October 1, 2017. No state program closed at expiry, because states held unspent allotments and because the Centers for Medicare and Medicaid Services redistributed remaining funds toward states projected to run short first. A December 2017 stopgap added 2.85 billion dollars and adjusted redistribution, while Congressional Budget Office state-by-state estimates indicated that the stopgap would carry some states only weeks. The Georgetown University Center for Children and Families reported in January 2018 that nearly 1.7 million children in separate programs in 21 states faced possible loss of coverage by the end of February 2018 without long-term funding in January. The capped structure had exposed children to a genuine risk an entitlement would not have produced in the same form, and accumulated balances had absorbed the shock for more than three months.

The flexibility case, stated at full strength, rested on state knowledge and state control. The National Governors Association, in a bipartisan letter adopted at its summer meeting in 2007, demanded reauthorization before the September 30, 2007 expiry, asked for a guarantee of flexibility for states in designing state programs, and warned that a series of temporary extensions would be untenable for states and for the millions of children who relied on the program. Before the 2015 funding decision, responses from 39 governors supported extending funding, and the Senate Finance Committee summary reported that most governors’ suggestions centered on simplifying enrollment, increasing state flexibility, and creating incentive payments. State officials argued in those submissions that a defined allotment, combined with state authority over benefits within federal benchmarks, premiums and cost sharing within federal limits, delivery systems, and outreach, let a state fit coverage to its insurance market, provider supply, and fiscal choices in ways a uniform entitlement could not. The Bipartisan Policy Center characterized the program as a capped entitlement to states designed to provide maximum flexibility to states. On that account, the cap was the disciplining half of a bargain in which the federal government paid an enhanced match, averaging about 71 percent before the 2010 law and about 93 percent during the 23 percentage point increase from October 1, 2015 through September 30, 2019, and states accepted a defined federal commitment in exchange for design control.

The entitlement-security case, also at full strength, rested on the claim that expiry risk reached families long before any fund reached zero. The American Academy of Pediatrics, in a September 2017 toolkit for pediatricians, warned that without certainty from Congress, states could be forced into drastic cuts, including slashing enrollment, reducing benefits, and imposing higher costs on families. After the December 2017 stopgap, an Academy statement reported by Kaiser Health News held that Congress had fallen far short of the reassurance and relief families deserved, and a coalition including the Academy described short-term funding as causing chaos and confusion on the ground. The Georgetown University Center for Children and Families documented the mechanism in state-by-state tracking through the autumn of 2017 and into January 2018, reporting states facing exhaustion of federal funds, states facing shortfalls in the first quarter of 2018, and states preparing to freeze enrollment, scale back eligibility, or end a separate program. Matt Salo, executive director of the National Association of Medicaid Directors, stated in May 2017 that certainty and predictability mattered because states unable to count on federal money had to plan to dismantle programs early, with a real human toll. In a 2017 policy statement on child health care financing, the Academy also argued that block grant and capped allotment reforms applied to Medicaid would compel states to reduce eligibility, trim benefits, or cut provider payments. Advocates in that camp treated entitlement financing as the form that removed expiry from family life, because an entitlement did not depend on a renewal vote arriving before a stated date.

The counter-reading began with a warning about analogy, and the counter-reading forms the analytical core of the coverage record. Critiques of capped grants from the welfare policy cluster, where a fixed federal grant combined with inflation, caseload change, and state fiscal pressure could erode real assistance over long periods, described a real mechanism in the programs where erosion had been measured. That mechanism did not transfer mechanically to Title XXI, because the financial facts differed at each point where the welfare critique required a tightening constraint. Congress funded the children’s allotments generously relative to measured need in ordinary years, repeatedly increased funding at reauthorization, rebased allotments on actual spending in the 2009 law rather than leaving allotments frozen at a historical base, created a Child Enrollment Contingency Fund in the same law for states whose enrollment outran allotments, and raised the matching rate by 23 percentage points from October 1, 2015 through September 30, 2019. A rebased grant with a contingency fund and redistribution of unspent balances occupied a different financial position from a grant left to erode against a fixed base, and treating both as a single category concealed the variables that decided whether a cap bound in practice.

Allotment balances supplied the clearest evidence that the cap rarely bound in ordinary operation. In most years many states did not spend full allotments, and the redistribution pool presupposed unspent funds available to move toward states facing shortfalls. The Medicaid and CHIP Payment and Access Commission, in 2014 technical presentations on program financing, described each state holding access to an allotment for a limited period of years, after which unspent funds became available for redistribution, with contingency funds behind states whose enrollment had risen sharply. Discipline therefore operated mainly through the appropriations calendar rather than through mid-year exhaustion. States planned against a known total, slow-spending states generated balances that covered fast-spending states, and the binding decision arrived at reauthorization dates, as in 2009, 2010, and 2015, before Congress failed to act on time in 2017. The structure imposed real discipline through a deadline while imposing little discipline through scarcity in ordinary years, a combination that fit the financing record through 2019 more closely than either chronic underfunding or an open-ended commitment under another name.

The events of 2017 and early 2018 marked the boundary of the counter-reading rather than refuting the counter-reading. Alabama announced an enrollment freeze from January 1, 2018 and a shutdown on January 31, 2018, then suspended those plans after the December 2017 stopgap. Connecticut froze enrollment in the last week of December 2017. Colorado, Connecticut, and Virginia sent letters to families warning that coverage could end, as Kaiser Health News reported in January 2018, and a Health Affairs account counted Alabama, Colorado, Connecticut, Oklahoma, Utah, Virginia, Washington, and West Virginia among states that posted announcements or sent notices warning families. Arizona, Colorado, Texas, Utah, and Virginia had laws requiring closure of a separate program without federal funding, according to an October 2017 Center on Budget and Policy Priorities summary of the Georgetown tracking. The structural risk was real and demonstrated in 2017, families in separate programs faced the sharpest exposure, and the same facts did not establish a binding scarcity constraint during the two decades before the lapse.

The precise lesson concerned design variables rather than financing labels. A capped grant behaved differently depending on allotment size relative to measured need, rebasing method, redistribution and contingency mechanisms, matching rate, maintenance of effort rules, and the constituency defending the program at each expiry date. The children’s program scored high on each variable through most of its history to 2019, and badly on exactly one, periodic congressional action, which converted a budget negotiation in the autumn of 2017 into warning letters and a 114 day lapse. Generalizing from the welfare cluster to every capped grant missed the generosity, rebasing, and balance evidence. Generalizing from two decades of adequate funding to the safety of expiring financing missed the Alabama and Connecticut freezes. Capped financing for children’s coverage had been workable in ordinary years because Congress had funded the cap generously and repaired the cap at each reauthorization, and workability had never amounted to security, because repair depended on congressional action arriving on time. That lesson was narrower, and more useful, than a general verdict on block grants.

What the program never promised followed from the statute’s text, and the limits were visible from 1997 rather than discovered in 2017. In a separate program, Title XXI created no individual entitlement, and a state could cap enrollment, maintain waiting lists, and impose waiting periods when funds ran short. Children in a Medicaid expansion financed by Title XXI held the Medicaid entitlement, and coverage continued at the regular Medicaid matching rate if a Title XXI allotment was exhausted, so entitlement status inside one federal program depended on the route a state had chosen for a child’s eligibility band. Benefits in a separate program followed benchmark and benchmark-equivalent standards rather than the Medicaid package with its Early and Periodic Screening, Diagnostic, and Treatment requirement, and premiums and cost sharing were permitted subject to an aggregate limit of 5 percent of family income. A guarantee of Medicaid-level benefits, Medicaid-level cost sharing, and an enforceable right to enroll had never formed part of the separate program route.

Parents stood largely outside the program’s scope, and permanence stood outside the program’s design. Title XXI defined its population as targeted low-income children, generally under age 19, in families with income above state Medicaid thresholds, and the definition did not extend to parents whose own uninsurance affected family finances and whether eligible children enrolled. A limited number of states covered some parents or pregnant women through demonstration waivers, but waiver coverage at state option differed in kind from a statutory eligibility category, and a child could hold coverage while a parent in the same household held none. Funding likewise arrived in defined periods, from roughly 24 billion dollars over the first five years under the Balanced Budget Act of 1997, through reauthorizations and extensions in 2009, 2010, and 2015, each ending on a date when authority expired without congressional action. The January 2018 restoration funded six years, fiscal 2018 through fiscal 2023, and the Bipartisan Budget Act of 2018 added fiscal 2024 through fiscal 2027, extending the horizon without removing expiry. The Congressional Budget Office estimated in January 2018 that extension would reduce federal deficits by roughly 6 billion dollars over a decade, because alternative exchange coverage would have cost more after repeal of the individual mandate penalty, and that finding argued for extension rather than permanence. Coverage for roughly nine million children rested, at every point in that history, on a future vote.

The coverage record and the withheld promises belonged in one account, because each disciplined the reading of the other. The National Health Interview Survey spine, from 13.9 percent of children under age 18 uninsured in 1997 to 8.9 percent in 2005, and the Bipartisan Policy Center count, from about 9.9 million uninsured children in 1997 to about 3.5 million in 2015, documented a transformation to which the program contributed jointly with Medicaid and, after 2010, with the health law of that year. Evaluations documented access gains against the uninsured baseline and access roughly comparable to private coverage on most measures, with specialty, dental, and after-hours weaknesses in the same literature. The crowd-out literature documented an unresolved dispute between the Congressional Budget Office range and smaller direct substitution in state and ten-state studies. The financing record documented a cap binding mainly at expiry dates, balances cushioning ordinary years and one extraordinary lapse, and a structural risk that Alabama and Connecticut families encountered in December 2017 and January 2018. The program had done a large and measurable part of the work assigned in 1997, under a structure that had never promised the security that supporters at the American Academy of Pediatrics and the Georgetown University Center for Children and Families told Congress during the lapse that families needed.

Closing Assessment: The Lesson of a Cap That Rarely Bound

The profile opened with a claim short enough to carry a verdict: A block grant for children’s health. Two decades of operation under Title XXI supply the material for judging that claim, and the material resists the two easiest judgments. Condemnation borrowed from welfare policy fails on the funding facts, and celebration borrowed from enrollment totals fails on the entitlement facts. Between those failures stands a narrower finding about a capped instrument that Congress funded generously relative to need, increased repeatedly, repaired at reauthorization, and allowed to expire on schedule in 2017, with consequences that states and families absorbed in the winter that followed.

The counter-reading deserves statement at full strength. In ordinary years the cap rarely bound through scarcity, because many states spent less than full allotments, unspent balances fed a redistribution pool, a contingency fund stood behind enrollment surges after 2009, and the matching rate rose far above Medicaid levels, reaching an average federal share of about 93 percent during the period from October 2015 through September 2019. Risk under Title XXI was real, documented in warning letters, enrollment freezes, and a shutdown date announced and then suspended in Alabama, and a freeze imposed in Connecticut during the final week of December 2017. That risk bound mainly through the appropriations calendar, at expiry dates fixed years in advance, rather than through mid year exhaustion of funds in a typical state. A structure can therefore be generous in level and fragile in duration at the same time, and the children’s program displayed both properties across the same two decades.

A profile of a statute settles the structure of the statute and the documented record of the structure in operation, and several questions remain open after that work is done. The magnitude of crowd out, the share of enrollment that replaced private coverage that families would otherwise have kept, stayed disputed between the Congressional Budget Office range and smaller estimates from state and multi state studies, and the evidence assembled here identifies the dispute without resolving the dispute. The contribution of the children’s program to the decline in child uninsurance, relative to Medicaid expansions and the health law of 2010, resists exact division for the same reason the decline itself was joint. Whether an entitlement design would have produced better coverage at acceptable cost, and whether generosity toward a sympathetic population of children predicts generosity toward less sympathetic populations, are questions of political judgment that a statutory profile can inform and cannot decide. The record also ends at a fixed point, with funding enacted through fiscal 2027 and a matching schedule written into law, and later congressional choices lie outside the evidence a profile dated 2019 can weigh.

A reader moving outward from this profile will find the surrounding architecture treated in companion profiles across the series. The structure of Medicaid, the open ended entitlement against which Title XXI was drafted, receives a dedicated account of Title XIX financing and eligibility. The welfare reform statute of 1996, the source of the capped instrument and of the five year waiting period that later children’s legislation addressed, receives a full guide to the replacement of Aid to Families with Dependent Children by Temporary Assistance for Needy Families. The health law of 2010, which extended children’s funding, imposed the maintenance of effort, and raised the matching rate, receives a provision by provision account. A comparison of the old cash assistance entitlement and the block grant that replaced that entitlement examines the two welfare models side by side, and a survey of federal health legislation since 1950 places Medicare, Medicaid, the children’s program, and the 2010 law in a single sequence. Each companion treats one element that this profile could only touch in passing, and together the set describes the system inside which a capped grant for children operated.

The lesson that survives is a lesson about variables rather than labels. Allotment size relative to need, rebasing on actual spending, redistribution of unspent balances, a contingency reserve, a matching rate generous enough to recruit states, maintenance of effort rules that held eligibility steady during a system transition, and a constituency prepared to defend funding at each expiry decided whether a cap bound in practice. Title XXI scored favorably on every variable except permanence, and the score on permanence was written into the statute on the day of enactment in August 1997. A block grant for children’s health worked, on the evidence assembled here, for as long as Congress kept choosing the grant, and the condition inside that sentence is the whole of the lesson.

Frequently Asked Questions

Q: Who is covered by the Children’s Health Insurance Program?

The program covers uninsured children described in the statute as targeted low-income children, generally under the age of 19, living in families with income above the thresholds that the state applies under Medicaid. A child must therefore fall outside Medicaid eligibility and inside the income band that the state sets for the children’s program, and the child must lack other coverage. States set the upper income limits for that band, and many states set limits well above 200 percent of the federal poverty level, reflecting how far up the wage scale the absence of employer dependent coverage extends. Coverage reaches a child through a state plan approved at the federal level, administered by the Centers for Medicare and Medicaid Services.

Q: Who created the Children’s Health Insurance Program?

Congress created the program in the Balanced Budget Act of 1997, Public Law 105-33, which established the program as Title XXI of the Social Security Act, codified at 42 U.S.C. 1397aa and following. President Bill Clinton signed the act on August 5, 1997. The Senate architects of the children’s proposal were Senator Orrin Hatch of Utah, a Republican, and Senator Edward Kennedy of Massachusetts, a Democrat, whose joint bill, the Child Health Insurance and Lower Deficit Act, announced in March 1997, supplied much of the design that the enacted compromise carried. First Lady Hillary Clinton served as a prominent administration advocate for children’s coverage. At the federal level, the Centers for Medicare and Medicaid Services administers the program, and states operate the program under approved state plans.

Q: Is the Children’s Health Insurance Program an entitlement?

The answer depends on the route a state chose, because Title XXI created a capped grant to states rather than a uniform individual entitlement for every eligible child. A child covered through a separate program holds no individual entitlement, and a state may cap enrollment, keep waiting lists, and impose waiting periods when funds run short. A child covered through a Medicaid expansion holds the Medicaid entitlement, because the coverage delivered is Medicaid coverage in substance and in law. In a combination program, entitlement status follows the component in which a child sits. The capped character of the federal funding therefore coexisted with an enforceable right to coverage for one group of children and no such right for another group inside the same federal statute.

Q: How is the Children’s Health Insurance Program matched by the federal government?

Federal payment uses the enhanced federal medical assistance percentage, set in Section 2105(b) of the Social Security Act. The formula starts with the regular Medicaid matching rate for a state and reduces the state share of cost by 30 percent, which raises the federal share by the same amount. A state with a 50 percent Medicaid rate receives a 65 percent enhanced rate, and a state with a 60 percent Medicaid rate receives a 72 percent enhanced rate. A statutory floor of 65 percent and a ceiling of 85 percent bound the result. Before the increase enacted in the health law of 2010 took effect, the average enhanced rate was about 71 percent, roughly 15 points above the average Medicaid rate. Federal payment remains subject to the state allotment ceiling.

Q: What happened when Children’s Health Insurance Program funding lapsed?

Federal funding authority expired at the end of September 30, 2017, and no state program closed immediately, because states held unspent allotments and the federal agency redistributed remaining funds toward states projected to run short first. A stopgap enacted on December 22, 2017 added 2.85 billion dollars and adjusted redistribution, and federal officials estimated that some states could begin exhausting funds around January 19, 2018. Eight states posted announcements or sent notices warning families that coverage could end on a stated date. Alabama announced an enrollment freeze and a shutdown, then suspended both plans after the stopgap, and Connecticut froze enrollment in the last week of December 2017. A law signed on January 22, 2018 restored funding for six years, retroactive to October 1, 2017, after a lapse of 114 days.

Q: Did the Children’s Health Insurance Program cut the uninsured rate for kids?

The measured rate fell steeply during the years after enactment, and each figure belongs to the survey that produced the figure. The National Health Interview Survey recorded 13.9 percent of children under age 18 as uninsured in 1997, 10.8 percent in 2001, and 8.9 percent in 2005. The Bipartisan Policy Center, in a published summary of survey data, expressed the change in counts and stated that uninsured children fell from about 9.9 million in 1997 to about 3.5 million in 2015. Attribution for the decline is shared among the children’s program, Medicaid expansions for children, and later the health law of 2010, because children moved between programs as income and state rules changed. No serious account credits Title XXI alone with the whole decline.

Q: What are the three state options in the Children’s Health Insurance Program?

A state plan may expand Medicaid using Title XXI funds, create a separate child health program outside Medicaid, or combine both approaches for different groups of children. Under a Medicaid expansion, a child receives Medicaid coverage and holds the Medicaid entitlement. Under a separate program, a child receives benchmark or benchmark equivalent coverage, premiums and cost sharing are permitted within federal limits, and no individual entitlement attaches. Under a combination, a state typically uses a Medicaid expansion for younger or lower income children and a separate program above that band, and entitlement status follows the component in which a child sits. Every route draws the same enhanced federal match within the state allotment, and federal approval of the state plan governs each route.

Q: How was the Children’s Health Insurance Program paid for?

The enacted compromise rode inside the Balanced Budget Act of 1997 and funded the children’s initiative at roughly 24 billion dollars over the first five years. The federal cigarette tax supplied the signature revenue device associated with the initiative. In place of the 43 cent increase that Senators Hatch and Kennedy had proposed in March 1997, the enacted law raised the federal cigarette tax by 15 cents per pack in two steps, to 34 cents on January 1, 2000 and to 39 cents on January 1, 2002. The tobacco connection carried a policy argument alongside the revenue, because sponsors in both parties argued that a higher cigarette price would deter teenagers from starting to smoke. Later funding came through reauthorization and extension laws in 2009, 2010, 2015, and 2018.

Q: What is the difference between the Children’s Health Insurance Program and Medicaid?

Medicaid is Title XIX of the Social Security Act, enacted in 1965, and rests on open ended federal matching with an individual entitlement for eligible persons. The children’s program is Title XXI, enacted in 1997, and rests on capped state allotments with an enhanced matching rate. A person who meets Medicaid criteria holds a claim to coverage that federal payment follows without a state ceiling of the Title XXI kind. A child in a separate children’s program holds no such claim. Benefits differ as well, because Medicaid coverage for children includes early and periodic screening, diagnostic, and treatment services, while a separate children’s program follows benchmark and benchmark equivalent standards. Cost sharing in Medicaid is confined to nominal amounts in most circumstances, while a separate program may charge premiums and copayments within federal ceilings.

Q: What family income qualifies a child for the Children’s Health Insurance Program?

A qualifying family earns too much for the child to qualify under the state Medicaid thresholds and no more than the upper income limit that the state sets for the children’s program. Federal law defines the population as targeted low-income children and leaves the upper limit to the state plan, so no single national income figure governs eligibility. Many states set limits well above 200 percent of the federal poverty level. The reauthorization law of 2009 reduced the federal match to the regular Medicaid rate for children in families above 300 percent of the federal poverty level, with exceptions for states that already held approved plans or enacted laws, a change that raised the state price of eligibility at the highest band without imposing a national cutoff.

Q: Can the Children’s Health Insurance Program charge premiums or copayments?

A separate program may charge premiums, enrollment fees, deductibles, copayments, and coinsurance, subject to protections in Title XXI. Total cost sharing for a family, including premiums, may not exceed 5 percent of family income for the period to which the cost sharing applies. For a child in a family with income below 150 percent of the federal poverty level, cost sharing may not exceed the amounts permitted under Medicaid, which means nominal charges. Cost sharing may not be imposed for well baby and well child services, including age appropriate immunizations. A Medicaid expansion follows Medicaid limits instead, which confine charges for children to nominal amounts in most circumstances and bar premiums for most low income children.

Q: What benefits does the Children’s Health Insurance Program cover?

Benefits follow the route a state chose. A child in a Medicaid expansion receives the Medicaid benefit package, including early and periodic screening, diagnostic, and treatment services, under which screening leads to treatment services necessary to correct or ameliorate conditions discovered through screening. A child in a separate program receives benchmark coverage, modeled on the standard Blue Cross and Blue Shield option for federal employees, a state employee plan, or the largest commercial health maintenance organization in the state, or benchmark equivalent coverage with an aggregate actuarial value at least equal to a benchmark. Benchmark equivalent coverage must include basic categories such as inpatient and outpatient hospital services, physician services, laboratory and x ray services, and well baby and well child care. The reauthorization law of 2009 guaranteed dental coverage across state programs.

Q: What happens when a state runs out of its Children’s Health Insurance Program allotment?

The consequence depends on the route and on the relief mechanisms that Congress built around the cap. Unspent allotments from other states flow toward states facing shortfalls through redistribution, and the Child Enrollment Contingency Fund created in 2009 supplies a federal reserve for states whose enrollment outran allotments. A state administering a separate program may cap enrollment, freeze enrollment, and keep waiting lists, because children in that route hold no individual entitlement. A state using Title XXI funds for a Medicaid expansion faces a different result, because coverage for expansion children continues after the allotment is exhausted, with federal payment continuing under Medicaid at the regular matching rate. Exhaustion therefore produces a fiscal problem for every state and a coverage problem only where the separate program route governs.

Q: What did the Children’s Health Insurance Program Reauthorization Act change?

The Children’s Health Insurance Program Reauthorization Act of 2009, Public Law 111-3, signed on February 4, 2009, funded the program for fiscal 2009 through fiscal 2013 and redesigned the incentives inside the 1997 framework. The act rebased state allotments on actual spending, created performance bonus payments for states that exceeded child enrollment targets and adopted specified enrollment simplification measures, and funded outreach and enrollment grants. The act created the Child Enrollment Contingency Fund for states whose enrollment outran allotments, guaranteed dental coverage, and strengthened premium assistance. The act reduced the federal match to the regular rate for children in families above 300 percent of the federal poverty level, with exceptions for states that already held approved plans or enacted laws, and Section 214 gave states an option described in full elsewhere in these answers.

Q: How did the Affordable Care Act change the Children’s Health Insurance Program?

The Patient Protection and Affordable Care Act, Public Law 111-148, signed on March 23, 2010, preserved the children’s program through the transition to a reorganized coverage system. The act extended program funding through fiscal 2015, beyond the fiscal 2013 horizon set in 2009. The act required states to maintain Medicaid and children’s program eligibility standards, methodologies, and procedures in effect on March 23, 2010 for children through September 30, 2019. The act also raised the enhanced matching rate by 23 percentage points for expenditures from October 1, 2015 through September 30, 2019, subject to a ceiling of 100 percent, and the average federal share during that period was about 93 percent. The combination converted the settlement of 2009 into a protected bridge during the first years of the reformed system.

Q: Can lawfully residing immigrant children receive Children’s Health Insurance Program coverage?

Section 214 of the reauthorization law of 2009 gave states the option to cover lawfully residing children and pregnant women without the five year waiting period that the welfare law of 1996 imposed on many immigrants for federal means tested benefits. The provision operated as a state option rather than as a nationwide removal of the waiting period. A state could elect coverage for lawfully residing children, for pregnant women, or for both groups, and a state could also decline the option and retain the waiting period within the federal framework. The option concerned persons lawfully residing in the United States, and no part of the option extended coverage to persons without lawful status. Families in a state that adopted the option could gain coverage inside the former waiting period.

Q: Does the Children’s Health Insurance Program cover pregnant women?

The statute defines the population as targeted low-income children, generally under age 19, and pregnant women stand largely outside the core eligibility category. A limited number of states covered some parents or pregnant women through demonstration waivers, and waiver coverage at state option differs in kind from a statutory eligibility category. Section 214 of the reauthorization law of 2009 gave states a further option to cover lawfully residing pregnant women without the five year waiting period that applied to many immigrants, alongside the parallel option for lawfully residing children. A child could therefore hold coverage while a parent in the same household held none, and coverage for a pregnant woman depended on a waiver or on a state election under Section 214 rather than on the basic design of Title XXI.

Q: For how long was the Children’s Health Insurance Program funded after the lapse ended?

Restoration arrived in two laws. The continuing resolution signed on January 22, 2018, Public Law 115-120, carrying the HEALTHY KIDS Act, funded the program for six years, fiscal 2018 through fiscal 2023, retroactive to October 1, 2017, with allotments of 21.5 billion dollars for fiscal 2018 rising to 25.9 billion dollars for fiscal 2023. The Bipartisan Budget Act of 2018, Public Law 115-123, signed on February 9, 2018, added four more years, fiscal 2024 through fiscal 2027, for a ten year extension in total. The Congressional Budget Office estimated in January 2018 that extending the program would reduce federal deficits by roughly 6 billion dollars over the decade, because alternative coverage in the subsidized exchanges would cost the federal government more after repeal of the individual mandate penalty.

Q: Why was the Children’s Health Insurance Program built as a capped grant rather than an entitlement?

The capped form answered the political conditions of 1997. After the failure of comprehensive health reform in 1994 and the Republican victories in the elections of that year, a new federal health commitment carrying an open ended entitlement faced deep resistance in Congress, and budget leaders insisted that new spending fit inside the budget agreement negotiated with the White House. A fixed allotment let supporters name the full federal cost in advance, and state flexibility over design supplied the stated compensation for the federal ceiling. Senator Edward Kennedy accepted a capped grant, a state contribution, and state discretion that an earlier Kennedy proposal for a federal entitlement had rejected, and Senator Orrin Hatch defended the discipline of the cap as the condition that made federal help compatible with a balanced budget.

Q: What is the maintenance of effort requirement in the Children’s Health Insurance Program?

The health law of 2010 required states to maintain the Medicaid and children’s program eligibility standards, methodologies, and procedures in effect on March 23, 2010 for children through September 30, 2019. The requirement covered standards, methodologies, and procedures alike, so a state could not accomplish indirectly through paperwork burdens what the requirement barred directly through a lower income threshold. The extension laws of 2018 prolonged the requirement through fiscal 2027 and narrowed the reach of the requirement on a stated date. From October 1, 2019, the obligation protects eligibility only for children in families at or below 300 percent of the federal poverty level, and a state may roll back higher eligibility and continue to draw federal funds.