Medicaid's Political Development since 1965: How a Fragmented and Unequal Program Has Expanded

Abstract

The Medicaid program has changed enormously over the past 60 years from a very restrictive program primarily attached to recipients on public assistance in 1965 to a much more expansive program today allowing coverage for persons regardless of marital, parental, or employment status. Incorporating the “medically needy”—an ambiguous concept from the start—allowed states to include many different groups in Medicaid who are not traditionally thought of as poor. In addition, three structural features illuminate why the program has expanded and changed dramatically over time: federalism and intergovernmental financing, the dominance of the private sector, and fragmentation. Unequal treatment among Medicaid-covered groups alongside partisan politics create a political discourse that often reveals Medicaid as a public subsidy for stigmatized groups while hiding Medicaid's reach into the middle class. This central political ideological tension collides with programmatic realities such that Medicaid strangely often suffers from a residual retrenchment politics while at the same time benefiting from embeddedness, making it extremely difficult to truly turn back the clock on Medicaid's expansion.

Medicaid's adoption in 1965 and its development in the subsequent 60 years must be understood in the broader context of America's “Grow and Hide” medical governance regime. As I have argued elsewhere (Grogan 2023), government funding of the American health care state has expanded dramatically over time, constituting at least 70% of national health expenditures in 2024 (Grogan 2023; Pauly 2019), but government's role is largely obscured from view. Government's influence is hidden because of the way we structure the US health care system—a heavy reliance on federalism, contracting out with the private sector, and extreme fragmentation—but also because of intentional public discourse that perpetuates a myth that we have a predominantly private health care system with relatively little government involvement.

This statement may seem surprising, because health policy experts will quickly point out—and most Americans know—that government is involved in health care today. They would point to public insurance—Medicare and Medicaid—as the main drivers of government spending. Indeed, they are the main drivers today, but government subsidies extend far beyond these two programs and did not start with the launch of Medicare and Medicaid in 1965.

For the past century plus there has been widespread support for government investment in four main areas: infrastructure development (e.g., building hospitals, nursing homes, and clinics), workforce development, public health promotion and prevention, and care for the poor and medically indigent. What is most important about care for the indigent is that support for “the poor” depends on setting boundaries, but where to demarcate these boundaries has always been—and continues to be—very ambiguous. Where to draw the line between the poor and nonpoor, and the medically needy and nonneedy, is a basic moral question that is logically contested in a democratic polis; but it is also ambiguous politically and in its implementation, which allows political actors to strategically hide or reveal growth depending on their position. Medicaid sits at the center of this political gamesmanship. States and the federal government are constantly shifting the boundaries of what and who Medicaid includes and excludes, and actors strategically hide and reveal expansions and retrenchment.

Discursive strategies are used to both hide and reveal growth according to political strategy and views about the appropriate role of government. There are three main structural features of the US health care system that explain why public subsidies have, by and large, continued to grow and yet are largely hidden: (1) reliance on federalism, (2) delegating authority to the private sector, and (3) fragmentation within government programs and across multiple agencies. Intergovernmental financing under federalism makes it easy to grow and hide health care expenditures because states can leverage funds from the federal government. But this leverage also hides total expenditure growth from those focusing on local governance (Rose 2013; Sparer 2015; Thompson 2012). Contracting out government-funded services to the private sector makes it easy to grow and hide because this activity is framed as investing in an innovative and efficient private sector while hiding the role of public financing behind private provision (Morgan and Campbell 2011). Fragmentation makes it easy to grow and hide because the expenditures of any one program or any one agency at one level of government is not very large, and the totality of expenditures across this fragmented system is very hard to see.

This same structure—federalism, contracting out, and fragmentation—is embedded within Medicaid, making the program so complicated today that it is difficult to see the totality of Medicaid's growth. Although the Medicaid expansion passed under the Affordable Care Act (ACA) has substantially expanded coverage, Medicaid was America's largest health insurance program a decade before the ACA was passed in 2010. By 2002, the number of individuals covered by Medicaid, a health care program framed as only for “the poor,” surpassed those covered by Medicare, a universal program for the elderly. In 2009, Medicaid had about 62 million enrollees, compared to Medicare's 45 million.

But Medicaid does substantially more than provide health insurance to millions of low-income families. It plays a crucial role in providing needed long-term care (LTC) services for the elderly and disabled as well. For example, it covers 70% of the elderly residing in nursing homes and 41% of total nursing home spending, helps more than eight million elderly pay their Medicare premiums and prescription drug costs, and covers nearly nine million nonelderly people with disabilities, including paying for the bulk of services (44%) provided to AIDS patients (KFF 2013). These statistics reveal that even before the passage of the ACA, we could no longer easily describe Medicaid—based on the program's actual statistics—as a health care program for “the poor.”

However, the framing of a program like Medicaid can depart from its reality, often for political reasons. Because delegating public money to private actors—e.g., hospitals, nursing homes, provider groups, and insurance and pharmaceutical companies, to name just a few—is central to the operation of the US health care system (the contracting-out mechanism described above), the private sector plays a central role in framing Medicaid and defining the boundaries of the health care safety net that Medicaid enrollees rely on for care. Private-sector discursive power existed before the launch of Medicare and Medicaid, but it became even more powerful over time as these programs continued to funnel enormous public expenditures into the private health care sector. Substantial subsidization, especially subsidization of capital expenditures from federal and state governments, created a powerful health care industry with vested interests in defining not only what is “private” but also what should be included in the “public” sector. Ironically, even though these private actors receive substantial funding from government, the demonization of government is a key tactic in their framing strategy.

While the term Grow and Hide might suggest that all government expenditures are hidden, the situation is far more complicated than that. Under the Grow and Hide regime, government expenditures for the poor are strategically revealed either as important for the truly needy or problematic because public expenditures are being misused on those not truly needy. It is this conspicuous description of government expenditures that allows the broader growth in government expenditures for middle- and upper-income Americans—and the systems that support them—to remain largely hidden (Grogan 2023; Hacker 2002; Mettler 2011). Yet, even though Medicaid is often explicitly and intentionally revealed for those groups fitting the American myth of minimal government and appropriate government involvement only for the poor and vulnerable, such as Medicaid expansions for pregnant women and young children in the late 1980s, the consequences of Grow and Hide—an enormously expensive system, inequitable access to and quality of care, and a highly fragmented system lacking rational planning—leave a growing number of middle-class Americans vulnerable and left out of the so-called private system that is supposed to serve them. As a result, Medicaid's middle-class reach keeps expanding. This has the potential to change Medicaid politics and at times creates a politics that reveals Medicaid's middle-class reach, such as President Clinton's reveal of middle-class families’ reliance on Medicaid-funded nursing home benefits, and more recently voter-led ballot initiatives in seven states that approved the Medicaid expansion. While these are important examples, the logics of Grow and Hide make revealing growth difficult, and more often than not they create a bipartisan consensus to hide the program's reach into the middle class. In this way, the Medicaid program is crucial to the larger workings of the Grow and Hide regime.

Clearly Medicaid politics can at times be quite salient and contentious. My core argument is not that Medicaid's remarkable expansion since 1965 has been entirely invisible or invariably consensual. Rather, the Grow and Hide regime has shaped how Medicaid's growth has unfolded and has influenced the reactions the program's evolution has generated among key stakeholders. Overall, Grow and Hide has made it harder for ordinary citizens to grasp Medicaid's cost, size, and programmatic scope; has promoted the bifurcation and fragmentation of health care financing and delivery; has reinforced narratives about the deserving and undeserving; and has made Medicaid politics conspicuous and contentious at crucial times and for select populations to serve a much larger hiding of a heavily government-subsidized US health care system.

How the Grow and Hide Regime Structured Medicaid's Development

Medicaid's Origins

Medicaid's origins and political development must be understood within the larger structure of the US health care system. After President Truman's failed attempt to pass national health insurance in 1948, liberal proponents of health care reform shifted their attention to senior citizens, a group considered deserving of government support and already tied to the Social Security system, a compulsory social insurance scheme. While the Democratic landslide in the 1964 elections provided a large enough congressional majority to pass Medicare and Medicaid, there was a divide at the start over what this dual adoption meant for the future trajectory of health care reform. Conservatives—such as Representative Wilbur Mills, who was chair of the House Ways and Means Committee and was crucial to the passage of Medicaid—viewed the program as a way to reduce demand for universal coverage; by providing for “truly needy” groups, Mills sought to stave off claims for broader health care coverage (Blumenthal and Morone 2009). In contrast, liberal proponents of national health insurance continued to view Medicare as an important first step toward universal coverage and Medicaid as a mere residual program that could be swiftly eliminated when national health insurance was adopted (Blumenthal and Morone 2009; Patashnik and Zelizer 2001).

The Importance of Federalism and Intergovernmental Financing

Although Medicaid was structured as a means-tested, targeted program, its extension of comprehensive benefits to the “medically indigent”—an ill-defined concept from the start—allowed for expansions to occur while also being strategically framed within the boundaries of Medicaid. Almost immediately after Medicaid's passage, some liberal states viewed Medicaid as an opportunity to expand coverage with federal funds. The state of New York was a pioneer in this regard. Using the medically needy provision, which included nonpoor people with medical needs, New York passed legislation in 1967 to set the Medicaid income eligibility requirements at a level high enough to encompass almost half its residents, thus including not only the poor but also working- and middle-class families. New York's actions called into question the fundamental purpose of the newly created Medicaid program: should it serve only as a safety net for the nation's neediest citizens, or should it extend coverage to the middle class? Federal legislators responded to this question unequivocally by passing an amendment in 1968 that capped income eligibility for Medicaid at 133% of the state-mandated Aid to Families with Dependent Children eligibility line (Grogan and Patashnik 2003a).

Yet, although conspicuous expansions to the middle class were off limits, other expansions to the deserving poor were acceptable. For example, the 1968 statute expanded a series of well-child benefits for poor children, creating the Early and Periodic Screening, Diagnostic, and Treatment program, making the Medicaid benefit package even more comprehensive (Rosenbaum and Sonosky 1999). The creation of the Supplementary Security Income (SSI) program in 1972 also produced an enormous expansion of Medicaid. It consolidated five separate state-run cash assistance programs for the aged, blind, and disabled into a single federal means-tested program (Quadagno 1988). Because SSI, unlike most means-tested benefits, is run as a nationally uniform program, a clear bifurcation among Medicaid beneficiaries was established. The elderly, blind, and disabled—who tended to be viewed sympathetically—gained Medicaid eligibility based on a federal eligibility standard. In contrast, with few exceptions, poor mothers and their children gained eligibility according to a (typically much lower) state eligibility standard (Watson 1995). This was the first major layer of fragmentation in the Medicaid program, which allowed for expansion for some groups and containment (or retrenchment) for other groups. This also allows for strategic revealing and hiding by group within Medicaid. As detailed below, while the elderly and disabled have always consumed the vast majority (about two-thirds) of Medicaid expenditures, expansions for these groups have moved well into the middle class but remain largely politically hidden. In contrast, expansions for poor and low-income nonelderly people are strategically revealed as deserving or undeserving.

Pregnant Women and Children

In the 1980s, when it became more widely recognized that most uninsured children resided in working families, there was a bipartisan effort led by Representative Henry Waxman (D-CA) to expand Medicaid to more “deserving groups” (Tanenbaum 1995). The political dynamics behind these expansions were a product of major interest group and partisan battles. The National Governors Association lobbied for expanded coverage for children, and the American Hospital Association lobbied for coverage for pregnant women, especially in light of the high and rising number of low-birthweight infants among uninsured women (Olson 2010; Sparer 2015). Consistent with the logic of expansionary federalism (along with intergovernmental financing), uptake was high among the states even when the federal expansion was optional. For example, when the 1988 mandate was passed to expand coverage to 100% of the federal poverty level (FPL) for pregnant women and infants, 76% of the states were already compliant. When this mandate was expanded again the following year to 133% of the FPL, 40% of states were still compliant. By at least initially allowing flexibility for state policy makers to shape their respective Medicaid programs, the federal government used financial inducements to encourage states to take the lead on transforming Medicaid eligibility and then later demanded that other states follow suit (Rose 2013).

The federal funding match meant that although states publicly complained about mandates, the vast majority of states began what came to be called a “Medicaid maximization” effort (Buck 2011). Quite simply, it became clear to states that it was cheaper to enroll many groups in Medicaid since the state was often paying for 100% of health care expenses in their safety net institutions, whereas even relatively wealthy states would receive a 50% federal match under Medicaid. All told, the targeted Medicaid expansions adopted from 1984 to 1990 increased the number of people receiving Medicaid benefits to 36 million in 1996, up from an average of 20 to 23 million between 1973 and 1989 (Melnick 1998).

LTC for the Middle Class

Medicaid's role in financing services for the elderly and disabled individuals also began to grow in the late 1980s and 1990s. Despite restricting the definition of “medically needy” early on, both the concept and the comprehensive benefits embedded in Medicaid's enabling legislation were sufficiently elastic that Medicaid continually filled the gaping long-term hole, as no other state or federal program covers these costs. Medicare has never covered the costs of long-term custodial nursing home care, and relatively few Americans have been able or willing to purchase private LTC insurance during their working years (Konetzka and Luo 2011).

As early as 1970, Medicaid had already emerged as the primary public purchaser of nursing home care. Just 10 years later, annual Medicaid spending on nursing home care reached $8.8 billion, equal to all other private and public sources for nursing home care combined (Olson 2010).

Senior advocacy groups believed that Medicaid's means test was stigmatizing and degrading to the elderly, and therefore they fought to expand Medicare—not to new groups as advocates of national health insurance had hoped, but by expanding the benefit package and reducing the out-of-pocket burdens on seniors. Congress responded to senior demands in 1988 by passing the Medicare Catastrophic Coverage Act (MCCA), which expanded Medicare's scope of services—prescription drugs, hospice, and long-term hospital care—but also required all Medicare beneficiaries to pay special premiums pegged to income. Medicaid was mandated to pay the premiums for beneficiaries with incomes below the FPL (Oberlander 2003). While the Medicare expansions under MCCA were repealed just one year after their enactment, MCCA provisions requiring Medicaid to pay Medicare premiums for low-income elderly remained intact (Himelfarb 1995).

When Medicare failed to respond to the needs of LTC for the elderly while Medicaid kept expanding in this regard, middle-class reliance on Medicaid for LTC became more politicized. In a 1990 congressional committee hearing titled “Medicaid Budget Initiatives,” Chairman Henry Waxman highlighted the middle-class aspect of Medicaid in his opening statement, but rather than presenting an endorsement, he argued—hearkening back to a residual Medicaid frame—that it was unacceptable for the deserving elderly to have to rely on a stigmatized Medicaid program: “Most people who need nursing home care eventually find themselves dependent on Medicaid. . . . It is absurd, and it is unacceptable that an individual could work hard for their entire life . . . then have to be impoverished and go on welfare” (US Congress, Waxman 1990).

The focus of this hearing was on Representative Barbara Kennelly's (R-CT) bill, which proposed demonstration projects to develop private-public partnerships to encourage the middle-class elderly to use public subsidies to purchase private LTC insurance. This was promoted as the solution to Medicaid's “inappropriate” reach. Under this bill, as Kennelly described it, “if and when an individual exhausts his or her insurance and applied for Medicaid, each dollar that the insurance policy has paid out in accord with state guidelines will be subtracted from the assets Medicaid considers in determining eligibility. In other words, coverage of LTC expenses by private insurance would count as asset spend-down for the purpose of Medicaid eligibility” (US Congress, Kennelly 1990a).

Although Waxman expressed general support for Kennelly's bill in his opening remarks, he raised concerns about whether the bill would create a new category of Medicaid eligibility that expanded eligibility to middle-class (or even upper-income) elderly people by allowing Medicaid to be used to protect assets and to finance the transfer of wealth. Kennelly's response to Waxman is revealing:

Kennelly's quote provides a perfect example of Grow and Hide's American myth: expanding public expenditures while claiming minimal government, and hiding the reach of her proposal under the banner of “private.” Indeed, if we decode her quote, it is quite shocking: “the ONLY [way of] protecting Medicaid . . . [as] its means-tested origin” is to expand public expenditures to the middle class through the private sector.

A few Democrats made some efforts to openly embrace a more expansive narrative revealing what the Medicaid program was already doing—its true reach—and who the program should cover. For example, President Clinton recast Medicaid as a middle-class entitlement during his budget showdown with the Gingrich-Dole Republicans in 1995–96 (Grogan and Patashnik 2003a). Yet this approach was not common, and no other Democratic president has done so since. More common (as detailed below) is Republican exploitation of Medicaid's fragmentation to pit one group against another.

Subsidies to the Private Sector Define the Boundaries of Medicaid

The government's approach to encouraging private capital investments is rarely discussed as health policy. Yet the amount of public expenditures committed to securing private capital is staggering. Moreover, capital requirements drive a particular type of behavior among health care institutions that defines the health care safety net, the boundaries of Medicaid, and ultimately health equity (Brown and Saltman 1985; Brown and Thomas 1988).

The competitive approach in health care that started in the 1980s assumed a level playing field among health care institutions, ignoring the biases of capital markets. In reality, health care institutions bifurcated dramatically after 1980 (Berenson 2015). Indeed, while public hospitals had long taken care of the poor, the “health care safety net” emerged as a new term and concept during this time period. Acknowledging the need for a health care safety net implicitly accepted the reality of a separate system of care for Medicaid and uninsured poor patients. Under the logic of Grow and Hide, even when the health care safety net—which Medicaid patients predominantly rely on for care—is explicitly funded by government, its composition is defined by the market.

In the early 1980s, US capital health policy consisted of two major drivers: tax-exempt bonds and Medicare reimbursement for interest and depreciation. By treating depreciation and interest expense as reimbursable costs, the federal government greatly expanded hospitals’ capacity to carry debt (Cohodes and Kinkead 1984). Investors in the tax-exempt bond market for hospitals viewed Medicare reimbursement as a stable revenue source for the nonprofit hospital industry and were therefore willing to finance much larger portions of capital projects with a lower pledge of revenue from the hospitals. As result, debt financing increased from 38% of total financing for hospital construction in 1968 to 69% in 1981 (Cohodes and Kinkead 1984: 22). The dollar volume of health care debt issues in the tax-exempt market went from $22 billion (5.7% of total issues) in 1974 to $75 billion (12.3%) in 1982 (Cohodes 1983; Cohodes and Kinkead 1984).

The increased reliance on capital markets created an even more bifurcated two-tiered system of care in the United States. While public subsidies made access to tax-exempt bonds easier for nonprofit hospitals after 1965, public hospitals were unable to take comparable advantage of these subsidies. Because public hospitals and small community-based hospitals in underserved areas had a higher proportion of Medicaid patients among their payer mix, even private investors in the tax-exempt bond market with more lax rules saw them as a greater risk (Cleverley and Nutt 1984). Several key factors were (and still are) taken into account to determine a hospital's creditworthiness, including hospital location, hospital market share, hospital management, reimbursement system (whether private, Medicare, or Medicaid), and financial performance (Cohodes and Kinkead 1984: 7–8). Capital health policy was clearly biased against public hospitals and small nonprofit community-based hospitals with a larger Black and Brown patient base and those in poorer Black and Brown neighborhoods in urban centers and rural areas, where many safety net hospitals were located (Brown and Saltman 1985).

Once a hospital that predominantly served Medicaid enrollees was labeled as a bad risk, it was very difficult for that facility to change its risk rating. Already by the mid-1980s, public and nonprofit hospitals that served a disproportionate number of Black, Latino, and poor residents were very unlikely to have access to capital markets, while those hospitals that served predominately white residents with private insurance were much more likely to have access to capital markets and the public subsidies that supported those investments (Cleverley and Nutt 1984; Kinney and Lefkowitz 1982).

As bifurication between the “have” and “have not” facilities continued to spread, and Medicaid enrollment and the number of uninsured increased, the need for an expanded health care safety net intensified. Public investment in community health centers increased enormously after 2000. In 2001, Congress endorsed President Bush's call for a doubling of the number of Federally Qualified Health Centers (FQHCs), creating 600 new and expanded FQHCs that served more than three million new patients (Mickey 2012). From 1999 to 2004, the total number of patients served by FQHCs increased by 45%, from nine million to more than 13 million (Hoffman and Sered 2005). Moreover, in 2008 Bush signed into law the Health Care Safety Net Act (Pub. L. 110–355), which provided substantial new funding and reauthorized the program for an additional four years.

Some scholars have argued that part of the reason Bush and many Republicans in Congress supported expanding FQHCs was because conservatives came to see them as weakening pressures to expand Medicaid or other larger reform efforts such as “Medicare for All” (Glenn and Teles 2009; Mickey 2012; Pierson and Skocpol 2007). While these are valid arguments, it is also important to note how safety net investments keep capital health policy intact and the private sector in control. Even as the Consolidated Health Center Program expanded FQHCs, the legislation put clear boundaries around its reach, “support[ing] the provision of health care services to the medically underserved—meaning those individuals living in rural and urban communities” (US Congress 2001). Similar to the bifurcated hospital system, because a large proportion of physicians refuse to treat Medicaid patients, FQHCs are a crucially important part of the delivery system for Medicaid enrollees. Nearly 50% of FQHC patients are funded by Medicaid. In this way, conspicuous funding of the health care safety net functions alongside hidden subsidies for the so-called private health care sector to maintain the segregated system of care for the poor and medically needy, many of whom are Medicaid patients. Notice how muddling the distinctions between public and private is useful politically because it allows the strategic use of “public” and “private” as rhetorical labels—to obscure government under private action, and to expose government under public action. In reality, the Grow and Hide regime grows and reveals a little for the deserving poor as a way to enable growing and hiding massive subsidies for the more privileged.

Medicaid Privatization: Contracting Out to Managed Care Organizations

In the early 1990s, policy makers readily characterized state Medicaid programs as failures. There was broad agreement among Republican and Democratic officials alike that Medicaid suffered from serious operational flaws. Numerous studies documented that Medicaid recipients were much less likely than Americans with private health insurance to have a relationship with a primary care doctor or to receive needed preventive care, and much more likely to receive their care in hospital emergency room settings or public clinics with long waiting lines (Davidson and Somers 1998). Despite targeted efforts to increase prenatal care and well-child-care coverage in the 1980s, a large proportion of Medicaid women received no or only minimal prenatal care services (Colburn 1991), and many children enrolled in Medicaid were failing to receive needed immunizations (Slovut 1991).

In perfect harmony with the rhetoric of private-sector superiority under the Grow and Hide regime, the diagnosis of Medicaid's failure did not place blame on inequitable access to subsidies and capital, or the creation of a segregated system of care. Instead, this diagnosis of Medicaid's ills under Grow and Hide fingers government as the culprit and encourages government contracting out to the private sector as the cure—and, in this specific case, the idea of contracting with managed care organizations (MCOs) as the solution. Proponents argued that Medicaid managed care (MMC), as state contracting with MCOs is called, would be more efficient and innovative than government-led Medicaid, promising to simultaneously reduce costs, improve access, and raise the quality of delivered services. As a result, adoption of MMC spread quickly across the states regardless of party control: total enrollment into managed care plans nearly doubled in 1994 and increased 51% in 1995 (Grogan 1997; Horvath and Kaye 1995).

With these initial adoptions came rhetorical promises from political actors that MMC would help to eliminate the two-tiered system of care and promote “mainstreaming,” as many called it at the time. They argued that MMC would increase physician participation in Medicaid because private plans could structure their existing contracts with physicians to provide strong incentives to participate in their Medicaid “product.” Although state MMC programs did not give Medicaid recipients equal access to mainstream providers or facilities, the prominent mainstreaming rhetoric attached to MMC continued, presenting the illusion that private plans provided expanded access for Medicaid recipients (Grogan and Gusmano 2007). Part of what perpetuated this illusion was the fact that under MMC, Medicaid recipients would receive a Medicaid managed care card (e.g., Blue Cross Blue Shield's Medicaid care), which often incorrectly suggested they had private coverage equal to that provided to state employees or provided by other employer-sponsored BCBS plans.

Rather than requiring equal access, federal policy moved in the opposite direction. In the early 2000s, Congress allowed states to contract with private plans that operated exclusively in the Medicaid market. This change allowed safety net providers (public and nonprofit hospitals, community health centers, and public health clinics)—who had long been the providers for the vast majority of Medicaid recipients—to form their own Medicaid MCOs. Soon, a growing number of for-profit health plans operated only in the Medicaid market and contracted primarily with the health care safety net (Sparer 2012).

Although some states realized cost savings through a reduction in inpatient utilization and improvements in quality of care through an increase in childhood immunizations, most states during the 1990s and 2000s were not able to realize substantial managed care cost savings (Davidson and Somers 1998; Sparer 2012). Despite the lack of evidence of MMC success (Sparer 2012), states continued to favor MMC over what politicians often referred to as the government-run system. In 2006, 65% of the total Medicaid population in the United States was enrolled in some form of managed care, and this increased to 70% by the decade's end (Ku et al. 2000; MACPAC 2011). By 2010, there was no longer any discussion of expanding access to “mainstream” providers, since it was clear that Medicaid managed care organizations offered much more restrictive provider networks (Bisgaier, Rhodes, and Polsky 2014; MACPAC 2011). Instead, MMC was repeatedly described as more efficient, innovative, and less stigmatizing, with no evidence to back that up.

Further Expansions to Hide and Reveal

The State Children's Health Insurance Program (SCHIP), 1997

While Medicaid enrollment continued to rise through 1996, employer-sponsored health insurance declined by a greater amount. By 1997, 43 million Americans lacked health insurance (Holahan and Pohl 2002). In 1997, Congress passed legislation creating the State Children's Health Insurance Program (SCHIP, now called CHIP), which helped make Medicaid coverage expansions—separate from cash assistance—possible. When the window of opportunity opened in 1997 for a targeted coverage expansion as a result of the growing economy, elected officials decided to focus, just as they had in the 1980s, on groups considered most deserving and cost-effective, namely uninsured children in working families.

The key debate in Congress was whether the coverage expansion should take place exclusively within Medicaid or be designed as a separate block grant (Sardell and Johnson 1998). Not surprisingly, the National Governors Association favored the new block-grant approach, in large part because it offered state policy makers more control over eligibility levels and the benefit package. Emphasizing federalism yet again, Congress allowed states to choose one of three options: a block grant, a Medicaid expansion approach, or a combined approach. While 16 states opted to use their SCHIP funds for Medicaid expansions, more than half either implemented a new separate (non-Medicaid) SCHIP program (16 states) or created combined programs (19 states) (Mann et al. 2002). As expected, the main reason states implemented a separate (non-Medicaid) program was to reduce costs by offering a restricted benefit package. However, many states also argued that they needed a separate program because working families would not enroll their children in Medicaid because of the welfare stigma associated with the program. There was a strong assumption that families without previous ties to cash assistance from the Temporary Assistance for Needy Families program (i.e., “welfare”) would be unwilling to sign up for Medicaid. Rhetorically, these notions of stigma were strongly associated with a government-run program.

The enactment of SCHIP coincided with the implementation of Medicaid managed care across the states, and separate SCHIP programs with restricted benefit packages were often added to existing contracts with private plans. Ironically, but quite intentionally, as Medicaid was moving away from its legacy of “welfare medicine,” conservatives began attaching the language of welfare medicine to certain Medicaid recipients. In particular, under expansions to low-income families, Republicans started to argue that Medicaid reform needed to encourage “personal responsibility,” a common welfare trope. In the Medicaid context, personal responsibility meant recipients should act like a private insurance consumer: choosing a health plan, becoming literate in the language of health insurance (i.e., copays, cost sharing, premiums, deductibles), and utilizing the health care system appropriately, which meant not using the emergency room (despite often not having adequate access to health care providers or urgent care centers).

To be fair, many policy makers across party lines had good intentions about wanting to destigmatize the Medicaid program. Many states streamlined the eligibility process by allowing people to mail in applications or sign up online or in places more accepting and welcoming than welfare offices, such as community centers or health care facilities. States also adopted 12-month continuous eligibility to eliminate the stigmatizing process of month-to-month redeterminations and to increase continuity of care. Finally, almost all states changed the name of their Medicaid programs, especially for programs associated with SCHIP, to “user-friendly” names, such as Denali KidCare in Alaska, AllKids in Illinois, HUSKY in Connecticut, and BadgerCare in Wisconsin (Snow 2003). These new names intentionally hid the connection to Medicaid and—along with their MCO contracts—presented a new frame that eligible recipients are deserving of private insurance.

In the early years of SCHIP implementation, from 1998 to 2001, Medicaid and SCHIP enrollment grew by an average of 30% across nearly every state, declining in only three states over this time period (Mann et al. 2002). By 2006, the average SCHIP eligibility level was 220% of the FPL, and 11 states set eligibility above 300% (Grogan and Rigby 2009). Through Medicaid and SCHIP combined, 47% of all US children were eligible, and 28 million poor and low-income children were covered in 2005 (Ross and Cox 2002).

Obviously, these expansions were crucially important for covering uninsured children. Yet the programmatic reforms under SCHIP were also a perfect example of how programs expand under the Grow and Hide regime: under the structure of federalism, contracting out, and fragmented into separate programs, along with intentional discourse to hide government's role. In this case, intergovernmental Medicaid funding was expanded for SCHIP recipients under a separate state-run program; their “deservingness” resulted in their being covered by the “private” Medicaid MCO prevision rather than government Medicaid. In this way, the government's role is hidden behind private MCO provision.

Federalism and More Fragmentation: Expanding Coverage through State Options

Since the program's inception (and as described above for pregnant women and children), Medicaid legislation has allowed states the option to cover particular groups and receive federal matching dollars. Today, states have the option to expand coverage to the following groups at higher eligibility levels: pregnant women and infants up to 185% of FPL, children up to 200% of FPL, working parents up to 250% of FPL, elderly and disabled people up to 300% of the SSI level, and “medically needy” persons. States can also expand Medicaid to uncovered groups through a large set of optional services, such as targeted case management and expansion of the psychosocial rehabilitation option. While these sound like wonky, minor add-ons, the technical terminology hides significant expansions. For example, the psychosocial rehabilitation option allows states to cover intensive community-based supports for people with serious mental health disorders. By 2005, almost all states had adopted these optional services; targeted case management accounted for $2.9 billion in Medicaid expenditures, and the rehabilitative option accounted for another $6.4 billion (Shirk 2008). People with mental health disorders constituted close to three quarters of recipients under the rehabilitative services option and accounted for almost 80% of expenditures (Crowley and O'Malley 2007). It is difficult to overstate the importance of these optional provisions for expanding coverage. Even before the ACA was passed, optional coverage consumed about two thirds of all Medicaid spending. This is important because while states complained about the federal Medicaid mandates, 60% of Medicaid expenditures was the result of states' willingly adopting optional Medicaid expansions.

Many of these groups did not fit the “deserving poor” image (e.g., those with mental illness), but their coverage under Medicaid fit the logic of Medicaid maximization under federalism, and fragmentation allowed states to hide their reach.

The ACA Embraces “Grow and Hide” and Its Shaping of Medicaid

The federal government passed major reforms under the ACA while leaving the main structure of the Grow and Hide regime untouched: relying primarily on Medicaid to expand coverage to the uninsured by using a set of complex and convoluted subsidies to bolster the position of private health care companies (e.g., insurance, pharmaceutical, and medical device companies as well as large hospital systems, to name but a few), and an unregulated capital health policy that perpetuates a bifurcated US health care system (Bruch, Roy, and Grogan 2024; Grogan 2023). The ACA's Medicaid expansion policy brought on a heated political debate about the use of a “government-run” Medicaid program to extend benefits to the uninsured. The political focus on Medicaid coverage expansions diverted attention from—and helped to hide—how the state props up private profit making in the US health care system. The test of whether Medicaid's political development fits the Grow and Hide logic is not whether Medicaid is always hidden. Under the logic of Grow and Hide, political actors strategically reveal public expenditures on groups framed as either deservingly poor or insufficiently vulnerable to be deserving of Medicaid's public expenditures. Although the bulk of political focus is on the government's role in caring for the poor and vulnerable, all the other enormous subsidies are left largely off the political agenda. Moreover, the ACA Medicaid expansion built on current inequities within Medicaid by continuing to bifurcate the program across enrollee groups, emphasizing questions of deservingness for some groups while hiding its reach to middle-class groups. It is in this way that Medicaid politics—often importantly conspicuous, but also strategically inconspicuous—serves the larger workings of the Grow and Hide regime.

The ACA expanded coverage in two primary ways: (1) by creating health insurance Marketplaces where individuals purchase health insurance from private plans during an annual open enrollment period, and (2) by expanding the Medicaid program to persons up to 138% of the FPL ($27,750 annual income for a family of four in 2022). Despite the importance of the ACA marketplace expansions for uninsured Americans, which covered 12 million people in 2021, it pales in comparison to the Medicaid program (Keith 2021). Because the ACA encouraged all uninsured Americans to sign up for health insurance, total Medicaid enrollment has increased 50% since enrollment for the ACA Medicaid expansion began in 2014. Medicaid has expanded so significantly that it is the dominant form of health insurance in many states. In 2020, one in five nonelderly residents relied on Medicaid for coverage in half of all states. In 10 of these states, one fourth of their residents relied on Medicaid, and nearly a third relied on the Medicaid program in New Mexico (34%) and Louisiana (31%) (KFF 2021).

Even though the ACA drafters were rhetorically clear that the Medicaid expansion was designed to help low-income populations and the Marketplace was designed to help middle-class individuals access private coverage, the reality is much more complex and not so clearly delineated. The trend of Medicaid enrollment reaching into the middle class that started before the ACA has continued and expanded further. The majority (65%) of nonelderly Medicaid enrollees had incomes above the FPL in 2019, up from 55% in 2008 (fig. 1). One third (32%) of nonelderly Medicaid enrollees had incomes above 200% of the FPL in 2019. Although definitions of “low income” and “middle class” are clearly ambiguous and politically contested, it is important to note that the US median income level in 2019 was slightly below 300% of the FPL, and about 20% of Medicaid enrollees had incomes above that level. While there is significant variation in the degree to which states cover Medicaid enrollees above the FPL, it is noteworthy that even in the most restrictive states, 54% of nonelderly Medicaid enrollees have incomes above the FPL (fig. 2). Conversely, even several conservative, nonexpansion states (e.g., Texas, Georgia, South Carolina, and Florida) cover individuals with incomes above $48,000 (above 400% FPL) for 9%–12% of their Medicaid enrollees (fig. 3).

Finally, another example of Medicaid's significant reach into America's middle class is the extent to which people have a connection to the program. In 2020, two thirds of all Americans (66%) receiving Medicaid coverage reported that either they (31%), their child (9%), or other friends or family members (26%) had received health insurance through the program (KFF 2020). Note that these are 2019 figures and the survey occurred before the COVID expansions; therefore, this is not a COVID blip but part of a larger trend in coverage.

Building on Federalism, Fragmentation, and Bifurcation

The ACA made the Medicaid program the main engine for expanded coverage yet hid the significant extent of these expansions through the institutional pillars of Grow and Hide—federalism, private contracting, and fragmentation—and a polarized conspicuous partisan political discourse that pits Medicaid-covered groups against one another, while a bipartisan discourse hides Medicaid's middle-class reach for the elderly and disabled.

Federalism

While the ACA Medicaid expansion allowed Democratic-controlled states to embrace a broader ideological vision of Medicaid as a program to prevent poverty and promote health equity for low- and even middle-income Americans, it also ushered in a strong Republican-led rejection of this vision. The National Federation of Independent Business v. Sebelius lawsuit, supported by 25 states, argued against the constitutionality of the federal ACA mandate to expand Medicaid (KFF 2019), and the Supreme Court ruled in favor of the plaintiff. This ruling emphasized the true significance of federalism under the Grow and Hide regime because the ACA expansions were built on an intergovernmental structure and allowed conservative states to bargain with the federal government over how to extend and frame expanded coverage.

Fragmentation

The ACA did not restructure the Medicaid program; rather, it added a new eligibility group on top of a complex eligibility system. Although the ACA expansion of Medicaid expanded coverage to all individuals up to 138% of FPL, states must still go through the arduous process of figuring out which people fit into which eligibility bucket. They do this because the federal government applies different matching rates for individuals eligible through different programs, whether it is disability, “regular Medicaid,” CHIP, the ACA expansion, or many other designations. As a result, states can and do design different benefits for different eligibility groups, which means individuals within the same state Medicaid program have access to a different set of covered services and medications, a different set of providers and health care facilities, and a different set of requirements for cost sharing and premium payments.

Partisan Discourse: Revealing “Public”

Under the ACA, Republican-led states have expanded Medicaid, but often under policy reforms that focus on mandating individual behaviors of particular Medicaid recipients with the intent to uplift “personal responsibility” (Grogan, Singer, and Jones 2017; Vulimiri et al. 2019). Under the Trump administration, by November 2019 seven states had approved work requirements, which require work as a condition of eligibility for Medicaid; 10 states required premium payments, including receipt of payment before coverage begins or a lockout period (e.g., six months in Indiana) if premiums are not paid; and seven states have received waivers to increase copays above previously allowed levels, and/or healthy behavior incentives tied to premiums or cost sharing (KFF 2019). For example, in Wisconsin, Governor Scott Walker requested permission to subject “the poor” who apply for Medicaid to a drug test (Hall 2015). All these reforms add enormous administrative burdens for individuals often already experiencing high levels of precarity in their lives (see Herd and Johnson, this issue). Although the Biden administration rolled back state work requirements, Republican-led discourse promoting work requirements and premium payments intentionally raise questions about whether certain low-income Americans deserve public Medicaid coverage in light of their current behavior, with the emphasis on public subsidies for these individuals (Grogan, Singer, and Jones 2017).

Partisan Discourse: Hiding the Medicaid Expansion under Private Reforms

When Republican-led states adopt the Medicaid expansion, they often portray their reform as not being Medicaid but rather a private initiative that demands consumer-driven personal responsibility. Under this frame, the state's role is intentionally hidden, and public funds received from the federal government are strategically obscured (Grogan, Singer, and Jones 2017). For example, Iowa's governor, Terry Branstad, described the Medicaid program as broken and his Medicaid waiver as providing a “commercial-like benefits package” (Noble 2013). Similarly, Seema Verma, who became the director of the Centers for Medicare and Medicaid Services under the Trump administration, and was the architect of Indiana's Medicaid reforms—the Healthy Indiana Plan (HIP 2.0)—remarked that the structure of Indiana's reforms was meant to “promote the notion of consumerism,” and argued that it “transforms Medicaid beneficiaries into consumers” (Roob and Verma 2008).

Conclusion: Consequences for Medicaid Politics

Medicaid's political evolution illustrates the power of the Grow and Hide regime to entrench an expanded Medicaid as fragmented and unequal, institutionally hidden through federalism with a heavy reliance on the private sector, and strategically made visible for low-income groups whose deservingness is affirmed or questioned while hiding expansions to the middle class.

Some have argued that Medicaid growth over time, and its reach into the middle class, might create a more favorable politics if cross-constituency political mobilization can occur (Grogan 2013). The mobilization efforts in seven conservative states (Idaho, Maine, Missouri, Nebraska, Oklahoma, South Dakota, and Utah), which successfully bypassed their state legislatures and passed the Medicaid expansion via ballot initiative, suggests Medicaid mobilization for a new benefit among those not currently covered can occur.

Yet, if Medicaid is entrenched as fragmented and unequal, this has serious political implications for mobilizing Medicaid constituents across enrollment groups (see Michener, this issue). First, when states eliminate Medicaid from specific program names, as in the examples discussed of the Medicaid expansion programs, and as they did with CHIP programs (e.g., AllKids in Illinois, Peach Care in Georgia), it may be (as the rhetoric and name change intended) that the people enrolled in these programs do not know they are (or were) on Medicaid. A recent study by McIntyre, McCrain, and Pavliv (2024) confirms that changing the name creates confusion among the American public, and surprisingly, it does not improve the popularity of the program. It seems that this is caused in part by increased confusion. The authors conclude that state-specific names serve to mainly “obscure the role of government without generating material increases in public support, regardless of partisanship” (McIntyre, McCrain, and Pavliv 2024: 467).

Second, because states have also intentionally created multiple types of Medicaid programs, each with a different name and connected to a private plan with different benefit packages, provider networks, and delivery system, constituents may feel connected to their particular program but not to Medicaid as a whole. For example, people connected to Medicaid through a “private” Medicaid expansion program may only be motivated to fight against retrenchment to their program and not against retrenchment for what is now called “regular Medicaid” for poor families (Grogan and Park 2018).

Finally, recent empirical studies suggest that public opinion about different Medicaid constituencies aligns with a conspicuous politics focused on deservingness for some groups. Groups that have historically been framed as deserving—for example, pregnant women and children—receive high levels of support for Medicaid coverage across partisan affliations; however, where partisan framings differ, such as with regard to work requirements and citizenship requirements, public opinion varies along partisan lines (Haeder, Sylvester, and Callaghan 2021, 2023). Those groups connected to the middle class that are rarely invoked—the elderly and disabled—receive support across party affiliation, suggesting that a hidden politics works in their favor. Notably, politicians have not engaged in significant cross-group mobilizing work within Medicaid, or across Medicaid and the Marketplace, and severe fragmentation may make it difficult even for advocacy and constituency groups to see the potential for a broad-based Medicaid coalition (Grogan and Park 2018; see Michener, this issue).

This central partisan ideological tension helps shed light on the politics of Medicaid's most recent retrenchment: the unwinding of Medicaid COVID expansions starting in 2023. While estimates of the total number of Medicaid enrollees losing coverage during the unwinding period (May 2023 to May 2024) vary, it is clearly significant, with a minimum of eight million to a maximum of 24 million people losing Medicaid coverage (Tolbert and Ammula 2023). Yet, similar to previous periods of retrenchment, state approaches to the unwinding vary, with some states allowing significant disenrollment while others attempt to maintain coverage as best they can. Even as the federal government rolls back its COVID funding, it approved 188 waivers for 47 states to allow states to pursue a whole host of strategies to maintain coverage for its Medicaid enrollees. The broad political patterns persist: partisan discourse frames the deservingness of the unwinding differently, and states contract with the private sector (MCOs and Community Health Centers) to conduct outreach to increase continuity of care for groups still eligible during the unwinding. Thus, Medicaid's paradox under Grow and Hide continues: a conspicuous residual retrenchment politics while at the same time contracting out to the private sector to maintain enrollments, making it extremely difficult to truly turn back the clock on Medicaid's fragmented and unequal expansions.

Acknowledgments

I would like to thank Eric Patashnik, Jon Oberlander, Michael Sparer, special issue participants, and an anonymous reviewer for their extremely helpful comments and suggestions for revision.

References