## Abstract

Seven states have used Section 1115 waivers to expand Medicaid as part of the Affordable Care Act (ACA). While each state pursued a unique plan, there are similarities in the types of changes each state desired to make. Equally important to how a state modified their Medicaid programs is how a state talked about Medicaid and reform. We investigate whether the rhetoric that emerged in waiver states is unique, analyze whether the rhetoric is associated with particular waiver reforms, and consider the implications of our findings for the future of Medicaid policy making. We find that proponents in waiver states have convinced a conservative legislature that their reform is sufficiently innovative that they are not doing a Medicaid expansion, and not building on the traditional Medicaid program. Particularly striking is that none of these reforms are entirely new to the Medicaid program. While not new, the way in which waiver states have been allowed to implement many of the reforms is new and has become stricter. We find an emerging consensus utilized by conservative policy makers in framing the Medicaid expansion. Expansion efforts by conservative policy makers in other states have subsequently pushed this framing far to the right.

## Introduction

Three major principles were embedded in the Affordable Care Act (ACA) when it passed in 2010: (1) federalism and encouraging state-level innovation; (2) universal coverage; and (3) incentives to bend the cost curve through delivery model innovations. When the Supreme Court ruled on the constitutionality of the ACA in 2012, upholding the individual mandate but allowing states the option to expand Medicaid, the Court consequentially privileged the first principle over the second (with no comment or action on the third). Allowing states the option to expand coverage necessarily means universal coverage is no guarantee.1 By the end of 2015, nearly 33 million people remained uninsured, primarily those residing in the nineteen non-expansion states (KFF 2016b). Although the decision to expand is largely driven by party control in the states, some Republican governors in conservative states have been able to pass their own version of a Medicaid expansion through the use of waivers. Indeed, Republican governors and legislators have used the first principle—federalism and states' rights—to claim that they have expanded coverage to low-income families, thereby allowing them to pull down substantial federal funds, but have taken a path uniquely suited to their state's conservative values.

Seven states have used Section 1115 waivers as a key element of their expansion of Medicaid as part of the Affordable Care Act (ACA): Arkansas, Indiana, Iowa, Michigan, Montana, New Hampshire, and Pennsylvania.2 This is in addition to the other twenty-four states (including Washington, DC) that have expanded Medicaid. These waivers were negotiated between the Centers for Medicare and Medicaid Services (CMS) and state administrations to make substantial changes to the Medicaid program in these states. Essentially, proponents in these seven states have convinced a conservative legislature that their reform is sufficiently innovative and different that they are not doing a Medicaid expansion as called for in the ACA, and not building on the traditional Medicaid program, which conservatives view as faulty and dysfunctional. These leaders need to convince CMS that their reform is a legitimate version of a Medicaid expansion and they need to convince their own state legislature that this is not Medicaid at all. While certain reform elements appear crucial to gain conservative support—charging premiums, imposing forms of cost sharing, incentives to modify lifestyle behaviors—equally important (if not more so) is the use of rhetorical devices to sell these reforms to a conservative constituency.

In an effort to distinguish between the diffusion of policy elements versus political rhetoric, we document not only how the elements of waiver reforms have developed in these first adopter waiver states, but also how the reforms have been framed, and whether there is an interaction between reform elements and political discourse. For example, when cost-sharing and premium components are discussed as important waiver components, what arguments (or political frames) appear in political discourse most frequently: Cost containment? Individual responsibility? Similarly, when the private option or Medicaid managed care designs are discussed, what arguments in political discourse appear most frequently: Individual responsibility in a marketplace? Efficiency of private insurance? Cost containment?

Very few scholars studying policy diffusion have looked at the influence of political discourse (Boushey 2016). This study will help illuminate the extent to which political discourse is an important mechanism to nudge policy adoption, and subsequent research can ascertain its import on policy diffusion of waivers in conservative states. Moreover, by looking at both policy reforms and rhetoric, we can examine whether framing becomes more important than the reform itself. In other words, are cost-sharing elements crucial for passage in a conservative legislature or is it the framing of personal responsibility that matters most?

We argue that these waiver states are important because they may act as a harbinger for how far Medicaid may be allowed to move to meet a conservative ideology, and, as such, has the potential to put Medicaid on a distinctive path in these states. Before detailing this argument, we provide a brief history of how Medicaid's past reforms have been framed to set the context to understand how these conservative ACA waiver reforms are similar and different from past reform efforts. Next, we provide a brief review of the policy diffusion literature to locate our contribution. After detailing our methodological approach, we then present our findings. We find a chronological pattern developing where each grouping of conservative states pushed for reforms further to the right of their predecessors and its associated rhetoric remained in sync—questioning the deservingness of the newly eligible, and seeking to return Medicaid to its original intent of only serving the truly needy.

## Background: The Framing of Medicaid's Past Reforms

Since 1962, Section 1115 waivers have allowed states to modify, or waive certain requirements associated with entitlement programs, including Medicaid when it was originally passed in 1965. Before 1993, the federal government observed strict budget neutrality requirements and demanded fully developed research designs, which limited the use of states applying for 1115 waivers (Thompson 2012). However, since then, the federal government has loosened requirements and states have used 1115 waivers to expand coverage, reform delivery systems, adjust payment models, and revise benefits and cost-sharing requirements (KFF 2011). As of November 2015, thirty-nine states had a currently approved or pending waiver with the federal government, and all but five states (Alaska, Nebraska, North Dakota, South Carolina, and South Dakota) had applied for a Section 1115 waiver (Medicaid.gov2016a).

These changes come at a time when Medicaid is already undergoing its most substantial shift in its fifty-year history. States expanding Medicaid as part of the ACA will cover all individuals below 138 percent of the federal poverty level (FPL). Prior to the ACA, Medicaid would only cover individuals who met specific categorical eligibility rules—for example, the elderly and disabled, and children and pregnant women—and were below federally established income thresholds. States were allowed flexibility in expanding the categories of eligibility and for modifying the income threshold to cover more people. However, the ACA Medicaid expansion is the first time the federal government has allowed states to only consider income and completely disregard categorical eligibility (with the important exception of federal policy regarding immigrants). Since children, pregnant women, and parents were already allowed Medicaid coverage, what this reform essentially allows is Medicaid coverage for healthy, single adults.

Questions about whether healthy, single adults are deserving of subsidized public health insurance is an old and long-standing debate in America. When Medicaid and Medicare were passed in 1965, the question of deservingness was central to the ideological wedge between conservatives and liberals. Most liberals at the time were in favor of universal health care coverage and viewed Medicare as the stepping-stone to achieve such coverage. In contrast, most conservatives viewed the dual passage of Medicare and Medicaid as having solved the problem of the uninsured. In particular, they argued that Medicare was now available for the elderly, Medicaid was available for the “truly deserving”—poor mothers and children, and poor aged and disabled persons—and affordable private insurance was available for the remainder of Americans, including able-bodied working men (Grogan 2008; Grogan and Smith 2008).

However, even by the mid-1980s, as health care costs continued to increase, and it became increasingly difficult to argue that private insurance was affordable for lower- and even middle-class Americans, the contours of this debate changed. More and more conservatives were comfortable with expanding the notion of deservingness—acknowledging that, for many working Americans, private insurance was unaffordable. A number of voluntary expansions occurred during the 1980s, and by 1990 Medicaid was required to cover additional groups: children and pregnant women and the low-income elderly up to the FPL (Tanenbaum 1995; Grogan 2008). In response to the inclusion of these sympathetic expansion groups, the program was no longer overlooked, but was rather seen as an essential component of the American safety net.

During this time, both Republicans and Democrats argued that Medicaid was essential, though they differed on the details of how the program should operate. Republican rhetoric after they gained majority control of Congress in 1995 focused on strengthening Medicaid through block-granting the program, rather than retrenchment. At the same time, an analysis of Democratic Party platforms found that Medicaid was now talked about as a broad social entitlement that provided assistance to the middle class (Grogan and Patashnik 2003). In a complete reversal from the views of policy makers at the implementation of Medicaid thirty years earlier, the failure of the Clinton health care plan and unsuccessful state-level health reforms during the early 1990s meant that many policy makers—across political parties—turned willingly to the Medicaid program to expand coverage (Grogan 2008; Thompson 2012).

The State Children's Health Insurance Program (SCHIP) passed in 1997 and expansions to working parents in 2003 were a response of this shift in opinion. Both passed with bipartisan support and allowed states to use Medicaid to expand coverage. However, a backlash against an expanded Medicaid program began to emerge as states took advantage of these options to expand Medicaid and the number of enrollees—as well as Medicaid expenditures—increased dramatically. The backlash started first under the 2008 SCHIP reauthorization debate where conservatives fought strongly against reauthorizing SCHIP, which by that time had expanded coverage to children in working families in some states as high as 300 percent of the FPL. Conservatives argued that states had gone too far in expanding coverage, especially since many of the “truly deserving poor” were still not enrolled (Grogan and Rigby 2009; Grogan and Andrews 2011).

This debate primed the discussion that followed after passage of the ACA. Although most of the disagreement in 2009–10 over the passage of the ACA at the federal level focused on how to design health care reform and not on questions of deservingness,3 moving the decision to expand Medicaid to the states allowed the deservingness question to emerge front and center again in conservative states (Grogan 2013). Yet, while the rhetoric of reform is unique in conservative states, many of the policy design elements requested and granted under the waivers are characteristic of broader trends undertaken by many states—conservative and liberal alike. Thus, although proponents of waivers in the conservative states appear to all claim their reforms to be dramatically different from traditional Medicaid, we interrogate the reform elements of waivers separate from their associated rhetoric and framing to determine if it is the policy design, its rhetoric, or aspects of both which makes these waivers unique.

### Policy Diffusion

The literature on policy diffusion provides helpful guidance and context on how to examine learning between states. Scholars have studied a variety of mechanisms contributing to the likelihood of a specific policy diffusing across states, including geographic proximity (Case, Rosen, and Hines 1993), shared similar political ideologies (Grossback, Nicholson-Crotty, and Peterson 2004), participation in extra-governmental organizations (Skocpol et al. 1993; Mintrom 1997; Balla 2001), similar institutional factors shared by different states (Boehmke 2005), the success of a policy (Volden 2007), and the expertise and experience of the policy maker within a state (Shipan and Volden 2014). Much of this research has focused on the horizontal spread of policies, for example, among various states within a federated system. However, another strain of research has highlighted the vertical nature of policy diffusion and learning. Vertical diffusion highlights when policy makers at different levels of government—federal, state, and local—learn from and are influenced by the implementation of policies at other levels of government (Shipan and Volden 2006; Karch 2007).

## Methods

We focus on nine states: the six that have used a waiver to expand Medicaid (Arkansas, Indiana, Iowa, Michigan, Montana, and New Hampshire), the one that submitted but subsequently rescinded its waiver application (Pennsylvania), and two in which governors pushed for an expansion waiver but were rebuffed by their legislatures (Tennessee and Utah). These states made for particularly good comparisons given their interesting variation in partisan control. Five of these states (Michigan, Indiana, Iowa, Tennessee, and Utah) were controlled by Republican governors during the entire period of our study (2012–15). Of these, only Iowa did not also have a unified Republican-led legislature. Two states (Montana and New Hampshire) were led by Democratic governors during the entire four-year period. Republicans controlled the legislature in both states except for a two-year period when Democrats controlled the New Hampshire House. Partisan control of the governorship changed in two states, with Arkansas moving from Democrat Mike Beebe to Republican Asa Hutchinson in 2015, and Pennsylvania moving from Republican Tom Corbett to Democrat Tom Wolf in 2015. Both legislatures were controlled entirely by Republicans during this period, except for one year in which Democrats controlled the Arkansas House and Senate (see table 1). This variation in party control, especially in the governorship because the state executive primarily shapes the framing of the waiver proposal, is methodologically helpful because it allows us to observe whether and how rhetoric changes when party control shifts.

Our methodological approach involved two main data collection efforts. First, we examined each Section 1115 waiver application submitted by the seven states that connected a waiver to the ACA's expansion. We used these applications as primary source material to focus on the structure of the reforms in each of these states. We collected the following seven variables of interest: design of expansion plan, premiums, cost sharing, healthy behaviors, employment regulations, health savings accounts, and employer-sponsored insurance premium supports.

Design of expansion plan includes the main mechanism which the state implemented for their expanded Medicaid program; examples of this variable include premium assistance plans in Arkansas and expansion of Medicaid managed care in Michigan. States vary at the FPL in which they require premiums and cost-sharing from new enrollees. with premiums in Indiana start at 0 percent of the FPL, and the other states require premiums starting at 50 percent and 100 percent of the FPL. Our healthy behaviors variable includes any requirement from the state which demands that enrollees complete health risk assessments, wellness activities, or preventative health activities in return for reduced cost sharing or premiums.

The US Department of Health and Human Services (HHS) has been adamant about not accepting waiver applications which required new enrollees to be employed to receive benefits. In an effort to claim success in negotiations with the federal government though, waiver states have pushed for certain employment regulations in a “holistic” approach to poverty reduction, requiring the newly enrolled to be engaged in employment programs or job-seeking efforts as part of their Medicaid benefits. Health savings accounts have an interesting history, introduced into Medicaid programs by Indiana in an earlier 2008 reform effort (Commonwealth Fund 2008). These accounts operate as a repository for enrollee premiums, subsidized by state contributions, and are used by the new enrollees to finance the costs associated with covered Medicaid benefits (i.e., premiums, co-pays and deductibles). The last variable collected, Employer-sponsored insurance premium supports, involve states providing subsidies to newly enrolled Medicaid beneficiaries who have access to employer insurance.

The second main data collection effort involved gathering text from policy and political discourse related to the waiver adoption process. The key sources of data which we collected were speeches and press releases produced by the governor of each of the waiver states; legislator statements, including press releases and floor speeches; and presentations, research briefs, and reports produced by relevant state agencies. We focused on key policy makers in each state, including the governor, state house and senate leadership, chairs of relevant committees, and heads of relevant agencies in each state.

To further support our detailed content analysis of the waiver applications, we also relied on two forms of media data to provide further documentation of the public debate around these waivers. The first was a search of the capital city-based National Public Radio subsidiary in each state. We employed uniform search criteria for each of these news websites of stakeholder first name, stakeholder last name, and Medicaid, with a date range of June 29, 2012, to July 15, 2015. We also completed the same search for the newspaper with the highest circulation in each state in our study: Arkansas Democrat-Gazette, Indianapolis Star, Des Moines Register, Detroit Free Press, Billings Gazette, New Hampshire Union Leader, Philadelphia Inquirer, Tennessean, and Salt Lake Tribune.4

In total, we amassed 1,227 documents as part of our data collection efforts.

Using these data, we conducted detailed content analysis of the frames, rhetoric, and arguments provided by the stakeholders in each of the waiver states. To ensure reliability among the three researchers, we individually coded a subset of press releases, speeches, and newspaper accounts related to Arkansas's Medicaid expansion. Independently, the three researchers identified themes related to the data. The three researchers then cross-checked their results to ensure similar identification of themes as well as developed coding strategies used for the rest of the analysis.

Each coauthor was assigned three states to code: an early adopter state, and two states which successfully expanded Medicaid through a waiver. Two coauthors were assigned states where governors sought a waiver but failed to implement an expansion. We coded several themes for each article included in our content analysis, including who was speaking, the intended audience, the tone of the speech toward the waiver, requested policy design of the waiver, how the deservingness of the expansion population was framed, and the use of evidence by the speaker. The data collected included direct quotations from key policy makers as well as our own notes on the use of frames and phrases changed over time. After finishing each state, the coder wrote a short summary of arguments and framing used to talk about Medicaid expansion within the state.

## Findings: Medicaid Waiver Reforms—Rhetoric and Reality

The most striking pattern about the reform elements passed in the seven waiver states is that the most popular reforms are also those most likely to be passed in non-waiver states. All seven waiver states have implemented some form of cost sharing and Medicaid managed care, five states have included various premium charges, and four states have included incentive policies to encourage healthy behaviors (see table 2). The most novel reform, which also received the most national press coverage, is the creation of the so-called private option which allows the state to draw down federal Medicaid funds for the new Medicaid enrollees to purchase private insurance on the state's ACA health insurance exchange. This is also called premium assistance since the state pays premiums to qualified health plans (QHPs) on the exchange for newly eligible Medicaid recipients. Arkansas was not only the first state to implement an expansion waiver, but also the first state to adopt a private option. Iowa quickly followed with their version of the private option and, a few years later, New Hampshire followed suit.

Much of the rhetoric supporting the uniqueness of the private option reform was that it was reforming Medicaid away from a state-run program since it contracts with private plans. However, as others have pointed out, state Medicaid programs have contracted with private managed care plans since the early 1970s, and today the vast majority of states utilize some form of managed care contracting (Rosenbaum and Sommers 2013). What makes the private option different from Medicaid managed care plans is that under the private option, insurers are not able to distinguish who in their population is a Medicaid beneficiary (Kliff 2013). In contrast, under Medicaid managed care, states contract with private plans who offer a separate Medicaid product, which means that Medicaid enrollees have access to a separate, and often much more limited, network of providers. Nevertheless, while this integration with other non-Medicaid enrollees is new, the reliance on the private sector is not new to the Medicaid program (Rosenbaum and Sommers 2013). When the ACA was signed into law in 2010, nearly 70 percent of the 60 million Medicaid beneficiaries were enrolled in a managed care program (Sparer 2012). This number continues to increase, with 80 percent of Medicaid beneficiaries currently enrolled in managed care plans—most under regular Medicaid expansions, not waivers.

The two other reform elements which were each adopted by three states are health savings accounts (in Indiana, Michigan, and later Arkansas), and voluntary work incentives (adopted by Pennsylvania, Indiana, and New Hampshire). Again, these two reform ideas have been around a long time. Under health savings accounts, individuals are encouraged to save for financial costs associated with out-of-pocket health care expenditures. Many states have utilized the idea of health savings accounts—most notably Indiana, even for its Medicaid program—before the ACA was passed. Job training and counseling have long been offered to recipients of public programs.

So, are these expansion waivers much ado about nothing?5 First, although most of these reform ideas are not new, the way in which waiver states have been allowed to implement many of the reforms is new and much more aggressive. Second, it is this stronger approach—combined with its associated rhetoric—that arguably helped allow for a Medicaid expansion in these conservative states during this particular time period. While other conservative states passed a “regular” Medicaid expansion, these waiver states are important because they may act as a harbinger for how far Medicaid may be allowed to move to meet a conservative ideology, and, as such, has the potential to put Medicaid on a distinctive path in these states. To detail this argument, we start by discussing the reforms adopted and the framing used in the first three states—Arkansas, Iowa, and Michigan. We then illustrate how Indiana and Pennsylvania adopted additional reforms and pushed the framing further to the right. Third, the more focused reforms adopted under New Hampshire's and Montana's waivers, and Arkansas's reauthorization, suggest an emerging consensus among conservatives that even very poor people must have skin in the private sector game. Finally, local debates surrounding Utah's and Tennessee's failed waiver proposals might suggest an even further push for additional conservative reforms. Conservative states continue to push for a Medicaid work requirement and are looking for ways to impose an expenditure cap (like a Medicaid block grant long desired by Republicans), and states—like Tennessee and Utah—may not be willing to compromise on these points.

## How Do States Expand Medicaid without Doing a Medicaid Expansion?

Political rhetoric around Medicaid has shifted over the lifespan of the program, often mirroring the social construction of the individuals eligible for the program (Olson 2010; Grogan 2011). Because earlier Medicaid reforms focused on expanding coverage to groups broadly considered deserving—children and pregnant women and infants—there was little debate about the recipients themselves (Tanenbaum 1995). Instead, the debate tended to focus on how state governments could redesign the Medicaid program to improve quality and access to care, and control (or decrease) Medicaid expenditures.

Actually, there was not much debate about redesigning the Medicaid program, since a broad consensus emerged by the mid-1990s that Medicaid managed care was the desired reform option (Grogan 1997). There was bipartisan agreement that contracting with private managed care organizations would be more efficient than relying on the traditional state-run fee-for-service Medicaid program and hopefully improve access and quality. As mentioned above, when the conservative backlash toward Medicaid expansions took hold during the SCHIP reauthorization period in 2008, the argument was twofold: first, that states had been allowed to expand Medicaid too far, so that conservatives questioned whether recipients were deserving of subsidized coverage; and second, an inefficient publicly run program was an inappropriate vehicle to use as a platform for expanding health coverage in America. Given these two concerns, it is not surprising that many Republican leaders spoke out strongly after 2010 against relying on Medicaid as the centerpiece for expanding coverage for America's uninsured.

Nonetheless, despite these major ideological concerns, the federal funding attached to the Medicaid expansion is a huge inducement, making outright rejection of expansion difficult even for Republican governors. Thus, as mentioned, the waivers present an opportunity for conservative leaders to reform the Medicaid program substantially enough to make a convincing argument that they are not utilizing a state-run public program to expand coverage to a group with questionable deservingness. How do they do this?

### It's a Private Sector Approach, Not State-Run Medicaid—Arkansas, Iowa, and Michigan

Arkansas started the fight for a Medicaid expansion waiver in ways similar to states adopting Medicaid managed care in earlier years. The key argument initially put forward by Democratic governor Beebe in conservative Arkansas focused on the need to control Medicaid costs. He never questioned the deservingness of the uninsured to receive Medicaid coverage but argued that the private sector could do it better. In 2011 the Beebe administration did not shy away from arguing that Medicaid should remain intact, but agreed that it needed to be significantly reformed. However, as opposition in the Republican-controlled legislature increased in 2013, the administration strategically eliminated the term Medicaid from public statements, and only focused on the private option.

The intent of this framing was clearly to portray the waiver as distinct from Medicaid and not building on the traditional program. Republican representative Justin Harris's question to Arkansas's Medicaid Director, Andrew Allison, is illustrative of conservative attempts to link reform back to Medicaid: “You also made a comment earlier about . . . the private option being, private insurance. But is it not true . . . that private insurance still has to act and resemble Medicaid, is that not correct?” Allison's response insisted that incorporating private plans into the reform will make it distinct from Medicaid: “Neither the Health Care Independence Act, nor the Affordable Care Act, nor guidance that the Insurance Commissioners put out, nor this draft waiver says private insurance must now mimic Medicaid. It just doesn't” (Kauffman 2013).

Interestingly, many proponents of the private option in Arkansas argued that not only would the reform control health care costs—a claim that many supporters and opponents questioned because premiums on the exchange were typically more expensive than what states would pay Medicaid managed care plans—but it would improve access to care because more providers would participate in the exchange plans and the state did not have a robust Medicaid managed care market (Rosenbaum and Sommers 2013; Allison 2014). This argumentation is important, because desire to not only expand coverage but also improve access implicitly acknowledged the new enrollees' deservingness. When deservingness was explicitly acknowledged, the newly eligible were always referred to as the “working poor.” For example, Governor Beebe said the following in his January 2013 State of the State Address:

There is another important discussion to be had this session about a very different group of Arkansans than the elderly, the disabled, and the children who we currently insure under Medicaid. There are thousands of Arkansas families living in homes where one or both parents work, but where health insurance is not affordable. Very rarely do adults of working age qualify for Medicaid, and rising costs have led more companies to drop insurance coverage for their employees. These families and individuals are often referred to as “the working poor,” and we have a real chance to provide them better access to health care. (Beebe 2013)

Although Beebe was clearly making a moral argument in support of their deservingness to receive publicly subsidized health insurance, it is important that deservingness is specifically attached to the “working poor.” We see a very similar rhetoric take hold in the next two states that passed Medicaid waivers—Iowa and Michigan.

Similar to Arkansas, Iowa's Governor Branstad described the Medicaid program as broken and also pursued a private option approach where the state contracts with qualified health plans on the ACA exchange to provide coverage to new Medicaid enrollees (see table 2). However, although the Branstad administration described the waiver as providing a “commercial-like benefits package,” surprisingly little rhetoric focused on this aspect of Iowa's reform. Instead, the discourse shifted in Iowa to a focus on “shared responsibility.” As Branstad described it, “If you have no skin in the game, you spend more. . . . We want to give people incentives to make the right choice” (Noble 2013).

Iowa's waiver reforms go along with this rhetoric. Although Arkansas imposed cost-sharing mechanisms for its newly eligible at very low income levels (50–138 percent FPL), the state capped it at 2 percent of income. In contrast, Iowa's reform incorporated cost sharing, premiums, and health behavior incentives (see table 2). As mentioned before, these mechanisms are not new to the Medicaid program, but they had never been imposed on persons below the FPL prior to Iowa's waiver. Iowa asks for very little in premiums ($5 monthly premium for those at 50–100 percent FPL, and a$10 monthly premium for those at 101–138 percent FPL); however, the state imposes cost sharing on all those newly eligible (0–138 percent FPL) at up to 5 percent of quarterly income. This is quite a significant change. These are the “skin in the game” reforms that the Branstad administration emphasized.

However, the Branstad administration put equal emphasis on the healthy behavior reforms that allowed a more malleable discourse to emerge. As Branstad put it: “The carrots and sticks in the Iowa plan will not produce miracles. At the same time, there is a real ‘declarative value’ for promoting healthy behaviors” (Bruner 2013). This language allowed Democrats in Iowa to be supportive even though they were concerned with the premium and cost-sharing reforms being too harsh. Although they recognized that changing health behaviors, such as quitting smoking and losing weight, is difficult, they still believed it was the right direction to go in. At the same time, conservatives who demanded greater responsibility from the newly eligible could point to his calls for increased financial skin-in-the-game for the newly enrolled.

In sum, Branstad emphasized how the two sides of the responsibility coin in the Iowa plan—cost sharing and premiums on one side, and healthy behavior incentives on the other—would encourage shared responsibility.

Just a few months later, Michigan's waiver was approved, meaning the debate about their waiver proposals was happening at about the same time period. Despite important differences in their waiver proposals, the discourse over reform was almost identical. Specifically, Michigan did not include a “private option” approach. However, because the rhetoric focused almost entirely on personal responsibility in ways very similar to Iowa, one would be hard-pressed to know that this reform element was missing from Michigan but present in Iowa.

While Michigan's leaders used the term “personal responsibility” instead of “shared responsibility,” the emphasis was primarily focused on healthy behaviors as evidenced by the reform's name: the “Healthy Michigan Plan.” Similar to Iowa, Michigan included monthly premiums and cost sharing, as well as healthy behavior incentives. However, distinct from Iowa, it also included a health savings account component (see table 2). Although Michigan imposes “skin in the game” and also adopted a health savings account requirement to be used to pay for monthly premiums, its rhetorical framing of “personal responsibility” focused on responsibility for the individual to take care of themselves, to exercise, eat right, and ensure they are seeing their physicians regularly for checkups (Michigan Health & Wellness 2016). Governor Snyder's press release asserts that personal responsibility is the hallmark of the Healthy Michigan Plan (Governor Rick Snyder 2013), but also argues that “the Healthy Michigan Plan is providing hard-working Michiganders with the health care coverage they need to lead healthy, productive lives” (Governor Rick Snyder 2014a).

It is important to note that, similar to the discourse in Iowa and Arkansas, Governor Snyder and other proponents repeatedly call the new enrollees “hard-working Michiganders” (Governor Rick Snyder 2015). Snyder argued that most of the enrollees who would benefit from the program were already working, but have low earnings, putting them in danger of bankruptcy if they needed to receive any care. This group would be protected from bankruptcy and have healthier outcomes if they had health insurance. Snyder very explicitly framed the debate around expanding Medicaid as an economic argument, both for the individual and for the state of Michigan (Governor Rick Snyder 2014b).

By implementing cost sharing, premiums, and healthy behavior incentives, and by focusing on shared or personal responsibility, Iowa and Michigan moved the personal responsibility lever further to the right, both in rhetoric and in the actual reforms. At the same time, the overall tone from both the Branstad and Snyder administrations was not overtly aggressive or harsh toward new enrollees. Governor Branstad would refer to the waiver's ability to provide needed services, such as mental health and disability services for many who were not previously eligible, and Governor Snyder highlighted the improvement in health and the quality of life for all of the new enrollees in the program (Governor Rick Snyder 2013; Governor Rick Snyder 2015). Nonetheless, by bringing the term responsibility to the fore, and emphasizing the working poor, they set the stage for more conservative states to move even further to the right.

### It's Not a Medicaid Handout Because Only Paying Consumers are Deserving—Pennsylvania, Indiana, and Arkansas Again

As mentioned above, questions about whether “able-bodied” adults are deserving of subsidized public health insurance is an old and long-standing debate in America. Allowing states the option to expand to this very specific group, whose deservingness has always been questioned, meant that the cultural trope of ‘individual responsibility’ (Harvey 2005: 76; Wacquant 2010) could reemerge at any time. It did with a vengeance in the next two waiver states.

Pennsylvania under Republican governor Tom Corbett was the first state to actively pursue a work requirement in their waiver proposal. Requiring the newly enrolled to either be employed or seeking employment was a point of contention within the state and in negotiations with federal officials. The rhetoric in Pennsylvania pivoted on the work requirement even though the federal government ultimately did not approve this reform element, and the remaining reforms are almost identical to Michigan's reforms (cost sharing for 0–138 percent FPL up to 5 percent of income; monthly premiums for 101–138 percent FPL up to 2 percent of income; and healthy behavior incentives), but with no health savings account.

Initially, the discourse in favor of Pennsylvania's waiver was focused on the benefits provided by the private sector and the virtues of the free market. For example, Governor Tom Corbett (R) argued that the “Healthy Pennsylvania” plan would provide “high-quality, private sector health insurance within reach of all citizens, regardless of their means” (Governor Tom Corbett 2014). This part of Pennsylvania's rhetorical frame was surprisingly similar to Arkansas's discourse around the private option, and ironically so, given that Pennsylvania did not pursue a private option, but a Medicaid managed care approach instead (see table 2).

However, Corbett was also a firm and outspoken believer in connecting employment and eligibility for the Medicaid expansion waiver. In the Healthy Pennsylvania waiver proposal, any individual who was working less than twenty hours a week needed to demonstrate that they had completed twelve job search activities a month in order to remain eligible for coverage. For Corbett, the work requirement was essential to the entire waiver proposal. Jennifer Bransetter, Corbett's policy director, remarked that, for the governor, removing the employment criteria “breaks the plan as a whole” (Associated Press 2014).

Although some of the rhetoric of personal responsibility was similar to Iowa and Michigan because it also focused on encouraging healthy behaviors, there was often an oblique reference to work as well: “[Our plan] provides incentives for healthy behaviors and increased independence through greater access to employment opportunities” (Governor Tom Corbett 2013). In Pennsylvania, there was a more overt attempt to define who among the newly eligible would be deserving of Medicaid benefits—those that are working or trying to work.

Note that this moral claim is so important that even after significant pushback from the federal government, legislators, and interest groups in Pennsylvania, Corbett did not drop the work requirement but modified it instead to a voluntary work search under a one-year pilot program.

Indiana was having its Medicaid expansion waiver debate at the same time as Pennsylvania and the discourse was very similar. Much of the rhetorical focus of Indiana policy makers was in couching the discussion of deservingness in terms of those individuals who acted like good health care consumers rather than individuals receiving a handout. The components of the Healthy Indiana Plan (HIP 2.0) were patterned after earlier reforms the state implemented in 2009. Seema Verma, architect to both reforms, remarked that the structure of Indiana's reforms was meant to “promote the notion of consumerism,” and that it “transforms Medicaid beneficiaries into consumers” (Roob and Verma 2008). This transformation led to the most complex structure of the Medicaid expansion in any waiver state. Indiana implemented four different benefits plans, premiums for all of the newly insured regardless of income, cost sharing for those below the poverty line, and an emphasis on healthy behaviors for all the newly enrolled. The central component to Indiana's plan—transforming the newly eligible into health care consumers—was the Personal Wellness and Responsibility (POWER) health savings account.

Fundamental to this belief in transformation is the logic that the newly eligible have to work to become consumers of health care. Although Indiana did not propose a work requirement as Pennsylvania did, it focused instead on imposing severe consequences for failure to pay monthly premiums. In particular, individuals above the poverty line who neglect to make monthly premium payments into their POWER account are dis-enrolled from the program for six months. Individuals below the poverty line who cannot afford or choose not to make monthly premium payments are shunted from the HIP+ plan, with its extended benefit package and cost-sharing protections, to the HIP Basic plan with fewer benefits and more mandatory cost sharing for services. The monthly premium payment requirements, along with severe consequences for nonpayment, implicitly impose a tie to work as it is difficult to imagine how one could pay their monthly premium without earnings. While the policy design imposes an implicit work requirement, the policy goal was made explicit: Ryan Streeter, the policy director for Governor Mike Pence, remarked, “We want to make sure that the program is consistent with our efforts to get people to work” (Wall 2014).

The design of Indiana's expansion plan affirms a traditional conservative view of structuring public benefit programs. This view holds that public benefits, which are purely a handout, should by design be miserly, so as to not encourage dependency. Indiana policy makers argued that their Medicaid expansion plan sought to short-circuit dependency with a system designed to encourage consumerism and consumer behavior by increasing benefits for individuals who are willing and able to pay into the system, and by putting money into a health savings account to pay for premiums and cost sharing. Without these elements, proponents argued, their program just “turns into a regular Medicaid program” (Groppe 2013).

Arkansas's experience after the election of a new Republican governor offers additional insight into this shifting conservative frame of the Medicaid expansion. As a Republican, Asa Hutchinson was immediately in opposition to the private option that previous Democratic governor Beebe had created and implemented. Whereas Beebe avoided the term Medicaid, claiming that the private option was an entirely new program, Hutchinson strategically linked the private option with Medicaid and argued that expansion built on and grew the Medicaid program. In all of his public speeches and statements on the Medicaid expansion, Hutchinson never referred to the waiver as the “private option” without also calling it “Medicaid.” It is noteworthy that in this context—a conservative state, which had just elected a Republican majority in both chambers of the General Assembly together with a Republican governor for the first time since Reconstruction—one central attack against private option reform was to simply call it “Medicaid.” Presumably, Medicaid is so poorly perceived in the state that just using the word Medicaid is sufficient to suggest the reform is bad.

While Hutchinson's discourse denotes a dramatic shift from that employed by Beebe, it is remarkably consistent with the rhetorical patterns we observed above in Pennsylvania and Indiana. Following directly on the heels of more conservative reforms in Pennsylvania and Indiana, Hutchinson also raised questions about the deservingness of the newly enrolled in the private option and the need for improving personal responsibility and work as part of the program (Governor Asa Hutchinson 2015). John Selig, director of the Arkansas Department of Human Services under the Beebe and Hutchinson administrations, almost seemed to be stealing a script from Indiana when he said after Hutchinson took office, “We believe in consumerism, we think they'll (the newly eligible) use care more appropriately and get a sense of how insurance works” (Andrews 2014).

In a speech on healthcare and Medicaid reform in Arkansas, Hutchinson specifically said he “wanted to talk about the profile of those on the Private Option.” In describing their profile, he began by stating that “there are unintended consequences to the Private Option. I don't know that anybody anticipated that parolees coming out of prison are put on the Private Option.” After noting this “unintended” characteristic, he then discussed recipients' employment status:

About 40 percent of the enrollees, at the time of application for the Private Option, showed no income. That means they were unemployed. Seventy percent of those on the Private Option were employed at some point in time, which tells me they were trying to get a job. That tells us that most are working but cannot find the steady work that is needed. Young people were more likely to have work than those who were over forty-five. Women were more likely to have work than men. This tells us that the older male population should be targeted for work. These might be men who've been laid off or who need to learn new skills to transition into another career. It's interesting that 10 percent of those on the Private Option are considered medically frail. And that population seems to me, if the Private Option were to end, would qualify for traditional Medicaid. This is all helpful information because it's the data that guides our debate. . . . I believe that there are some principles that should frame the debate. One of them is work and responsibility. I want our social programs in Arkansas to be an incentive for people to work as opposed to an incentive for people not to work. (Governor Asa Hutchinson 2015)

This long quote detailing the profile of the private option waiver recipient in Arkansas is significant because it reveals many important aspects of the conservative frame. First, not all recipients should be treated equally: ex-prisoners on parole should be treated differently from working families, for example. Second, and most explicitly stated, access to Medicaid should be tied to work. And third, Medicaid should return to its original intent—the medically frail should be kept separate in the traditional program as intended.

### Without a Work Requirement or the Block Grant, You Don't Get a Medicaid Waiver—Utah and Tennessee

Republican governors in Utah and Tennessee both tried and failed to get a Medicaid waiver approved by their Republican-controlled legislatures. Many of the same elements of reform were included in their proposed plans and much of the same rhetoric was used. Obviously, many political factors help to explain why waivers pass in some states and not others (Hertel-Fernandez et al., 2016), and we do not mean to imply in this article that the framing is directly related to passage or failure. However, in documenting the shifts in discourse over time among the waiver states, we note both what seems to become the expected conservative framing by 2015, and what may be new demands—even if not passed—and their associated rhetoric. The rhetoric used to explain why the waiver was not enough in Tennessee and Utah suggests that the needle on what conservatives say they require to comply with a Medicaid expansion may be shifting further to the right.

Following the lead of other Republican-dominated states, they sought to use the Medicaid expansion waiver to convince conservative legislators that their proposed reform was not a Medicaid expansion. Indeed, Tennessee governor Bill Haslam was explicit in describing this strategy: “What we have to come up with is a plan that says this is not really expanding Medicaid as contemplated by the Obama administration” (Farmer 2014). Explaining the political dilemma facing Republican legislators, he explained further, “I think the concern they have is ‘if I get in a primary race and somebody says he voted for Obamacare . . . ’ they want to be able to show the distinction and I think that's one of the things we'll work really hard over the summer to show here's why this is different” (Daniels 2015).

Governor Haslam ultimately released a waiver proposal called “Insure Tennessee,” which contained many of the same elements used by other states, notably premiums, co-pays, and incentives for healthy behaviors, and employed the same rhetorical device emphasizing greater personal responsibility. Also, similar to Indiana and Arkansas, Haslam tied his reform to the importance of work through the use of required payments into health savings accounts.

However, the Haslam administration also included two new elements. First, people below 138 percent FPL could choose to either receive a defined contribution voucher to apply to their employer's health insurance plan, or receive vouchers through a redesigned component of the state's existing TennCare program (Governor Bill Haslam 2014). Proponents developed a slightly new frame to explain this reform element, arguing that Haslam's plan better “prepared participants for eventual transition to commercial health coverage” (Governor Bill Haslam 2014).

Second, Haslam's plan also involved an agreement from Tennessee hospitals to make up the difference in funding when the federal match phases from 100 percent to 90 percent. As a result, the state budget would not be directly affected. Tennessee attempted to up the ante in emphasizing an increased role for the private sector and limiting the role of the state to a defined contribution under which hospitals would pay the state share. This latter reform element is especially related to long-standing Republican concerns about fiscal sustainability.

Initially, the plan seemed to win over previously skeptical legislators, and these two new elements seemed to make a difference in how the plan was perceived. One Republican member of the state house of representatives commended the governor that “this is not some cut-and-paste plan that other states have tried to sell as unique” (Daniels 2015). Another Republican senator said he decided to support the governor's plan once he became convinced that Haslam won conservative compromises from the Obama administration (Boucher 2015). Ultimately, Insure Tennessee did not receive enough support to advance through key legislative committees, despite the concessions in the waiver and a similar conservative framing.

And even Haslam's strong moral claim that the state has an obligation to help the poor, rather than moving other legislators, was largely countered with a very old conservative argument that people have an obligation to increase charity care, and that charity care is available for people in need. For example, a local news source reported that state representative Sheila Butt said that “the existence of facilities like the Hope Clinic show government intervention may not be needed. ‘I can tell you from experience, that when constituents have called our office, we have found places like Hope Clinic,’ she said. ‘We can't depend on the government for every single answer to every single question’” (Sisk 2015).

Nonetheless, in Tennessee, Republican arguments that a Medicaid expansion—even a conservative waiver reform—would not be fiscally sustainable with no escape valve became a focal point. Despite Tennessee hospitals agreeing to pay the state share, Senator Brian Kelsey, chair of a key committee, explained, “I'm concerned that this will be like the Eagles' ‘Hotel California,’ where you can check in but never check out” (Sisk 2015).

Fiscal sustainability was a similar battle cry and sticking point in Utah as well. However, they were also rhetorically unwilling to compromise on the work requirement reform as Pennsylvania had, for example. Similar to governors in other waiver states, Utah governor Gary Herbert asserted that personal responsibility should be the guiding framework for any attempts to reform the health care system (Herbert 2012).

However, Utah house Republicans also wanted assurances that Medicaid would be returned to its “original intent” as specified under their “Utah Cares” plan, which called for using state general funds and the less generous federal matching rate to fund the program (Moulton 2015). Ironically, Utah Cares would have cost the state more money while only extending coverage to 100 percent of FPL with a more limited benefit package. But, proponents of Utah Cares argued that their program was focused on the truly vulnerable and deserving in their state. Speaker of the Utah House of Representatives Greg Hughes (R), a backer of the Utah Cares plan, stated that Governor Herbert's plan broadens Medicaid to include populations which were never intended to be covered by the original legislation, referring to the mostly single, able-bodied adults without dependent children which would be covered under the “Healthy Utah” plan. Policy makers aligned with Speaker Hughes against Governor Herbert argued that physicians would no longer accept the needy Medicaid patients because of low reimbursement rates, and that potential cost overrun in the governor's plan could lower the amount of funding for programs that served the truly vulnerable in society, including the developmentally disabled, and those in children's programs and nursing homes (Gehrke 2015; Davidson 2015).

Initially, Governor Gary Herbert (R) of Utah had identified a work requirement as an essential component of any waiver application from the state. During a meeting with President Obama, Herbert referenced the work requirement, remarking that he had not gotten everything that he wanted in his negotiations over expanding Medicaid. Obama stopped the meeting and called Secretary Burwell to see if there was any flexibility on this point. When Herbert returned from his trip to Washington, DC, he reported that federal officials agreed to a work effort reform, which would require that newly eligible Medicaid enrollees would be connected with job training and job enhancement services, but not require employment to be eligible for Medicaid (Moulton 2014).

But this was not enough for Utah house Republicans. As Speaker Hughes wrote to explain his opposition to the governor's waiver proposal, there were two main sticking points for which they were unwilling to compromise. “Our governor, Gary Herbert, sought to receive a federal waiver that would allow Utah to implement a work requirement [that tied eligibility to work instead of just job training and job enhancement services] for Medicaid expansion recipients. He was turned down. He also tried to structure a program with a cap in order to protect the state from serious cost overruns that could put our budget, and the ability to fund other needed programs, in jeopardy. Again, the federal government said no” (Hughes 2015).

A cap on expenditures (or what is more commonly referred to as a Medicaid block grant) is not a new Republican idea. It was first requested by Reagan in 1981 and pushed hard by the Gingrich Republican-controlled Congress in 1994, and was a Republican rallying cry again during the SCHIP reauthorization debate in 2008. However, until Utah Cares, this was not part of the waiver requests. Time will tell whether this becomes a common waiver request as the work requirement now is. Note, as of this writing, Arizona has a waiver proposal pending with the federal government that also includes a work requirement.

## Summary: Patterns around Rhetoric and Reform

In an effort to distinguish between the diffusion of policy elements versus political rhetoric, we documented not only how the elements of waiver reforms have developed in these first adopter waiver states, but also how the reforms have been framed, and whether there is an interaction between reform elements and political discourse.

We presented a chronological evolution of framing in these waiver states, though we acknowledged that our timeline was short and the number of states was small. Our data are exploratory but suggest the possibility of some emerging rhetorical-reform patterns. Although the most common reform elements are similar to elements in non-waiver states, the rhetoric around and the use of the other reform elements make the waiver states distinct. In general, we found a strong link between overall framing and reforms proposed. In each stage, when the discourse significantly changes, it is associated with a specific demand for new reform elements. The first three states—Arkansas, Iowa, and Michigan—focused largely on the importance of embedding reform within the private sector and encouraging the “right healthy choices” through “shared responsibility”; whereas, the next set of states—Pennsylvania and Indiana—were steadfast that personal responsibility must encourage work behaviors where one is expected to pay for benefits received. This became the first real push to tie Medicaid reforms to work through the use of health savings accounts and a proposed work requirement (see table 3). This framing was carried forward by the next wave of states—New Hampshire and Montana—where it seemed commonplace for conservatives to claim that poor people must look for work and must have skin in the private sector game. And, the debate in the failed waiver states—Utah and Tennessee—largely hinged on their lack of ability to secure a work requirement.

We find a chronological pattern developing where each grouping of conservative states pushed for reforms further to the right of their predecessors, and its associated rhetoric remained in sync—questioning the deservingness of the newly eligible and seeking to return Medicaid to its original intent of only serving the truly needy. The framing of deservingness in the first three adopter states—Arkansas, Iowa, and Michigan—focused on the “working poor” and assumed most new recipients are “hard-working,” and the frame was fairly similar to that used in non-waiver expansion states. However, the next set of adopters put a dagger in this assumption of deservingness. Pennsylvania and Indiana set the stage, and all remaining waiver states followed, by claiming that Medicaid deservingness should be tied to work. Under this frame, only the “truly needy,” who are unable to work, should be given Medicaid benefits with no requirements attached (see last row of table 3).

Proponents of waivers in these conservative states all claim their reforms to be dramatically different from a traditional Medicaid expansion. While their most common reform elements are more similar than different from non-waiver states, our analysis of all the elements of reform, as well as the associated rhetoric, suggests that both the rhetoric and reforms are distinct in important and potentially consequential ways. The framing has resurfaced questions—some thought settled under the passage of the ACA—about who is deserving of Medicaid and whether so-called able-bodied adults should be left out. And, the elements of reform assert this deservingness frame: enrollees are required to pay for the benefits (e.g., premiums, co-pays), and they must have earnings to do so (e.g., health savings accounts, a work requirement or incentive).

## Conclusion

The evolution of state 1115 waivers connected to the expansion of Medicaid is a fascinating example of reinvention throughout the policy diffusion process. The first states that expanded Medicaid with a waiver—Arkansas and Iowa—can either be thought of as late adopters who reinvented the innovation, or first adopters of a new innovation. Either way, this is a clear example that adopting a policy is not necessarily a simple dichotomous choice for a state. The terms of the debate changed with each subsequent waiver, influencing the range of options being considered and the way the reform was talked about.

While our focus has been on Medicaid policy, waivers have also played an important role in the development of social policy in the United States, particularly with education policy in the wake of “No Child Left Behind” and welfare policy in the 1990s. States have few options in the face of federally established rules and regulations. State policy makers can attempt to persuade a majority of members of Congress and the executive branch to modify existing rules and regulations, or they can withdraw from participating in the program. Waivers allow states to find a middle-ground approach between those two options and it is in a federal system where waivers are an important instrument for policy diffusion. Waivers allow modifications to federal policy to diffuse to interested states.

These waiver states are important because they may act as predictors of how far Medicaid may shift to meet conservative ideology, thus having the potential to put Medicaid on a distinctive path in waiver states. This is a crucial moment in the evolution of the Medicaid program. One would have thought that the ACA Medicaid expansion would further consolidate the Medicaid program, creating more equal treatment within states because categorical requirements were abandoned in favor of a simple means test,6 and more equity across states because of the initial federal mandate to expand to 138 percent FPL. Of course, the Supreme Court upended this latter goal, but the waivers may upend the former as well. A handful of Republican-controlled states have used intense criticism of Medicaid to convince fellow conservatives to support their distinct version of a Medicaid expansion.

While the Obama administration and most state Democrats support a straight expansion, they have been willing to compromise and support waiver proposals under the logic that even restrictive coverage is better than no coverage. It appears as a win-win, since this approach has also enabled conservative legislators to say that they did not support Obamacare or a Medicaid expansion but an entirely new reform. Our tour through these conservative states with adopted or proposed waivers provides a window into the language used to thread this delicate needle. However, documenting the reality of the reforms adopted alongside this framing also provides a window into what is at stake.

While proponents both in Indiana and for Arkansas's reauthorization were explicit that “skin in the game” meant an implicit tie to work—since earnings are needed to meet the health savings account requirements used to allow recipients to act like “consumers” who pay the co-pay and premium share requirements—an implicit tie may not be enough for conservative states going forward. Debates surrounding Utah and Tennessee's failed waiver proposals, and Arizona's pending waiver, suggest conservative states will continue to push for a Medicaid work requirement and are also looking for ways to impose an expenditure cap (like a Medicaid block grant long desired by Republicans). And, if the wider Republican Party looks like it does in Tennessee and Utah (i.e., a strong Tea Party component, see Hertel-Fernandez et al., 2016), states may not be willing to compromise on these points.

It is not only conservative rhetoric around Medicaid that has shifted back toward a focus on the personal responsibility and deservingness of Medicaid enrollees. Conservative efforts to return the program to its so-called original intent—to reserve the program only for those who are “truly needy”—may be successful in these waiver states. Indiana's reform, for example, creates a very fragmented Medicaid program where each group is treated to a different set of rules and benefits according to group characteristics (e.g., single adults) and ability to pay. In this sense, the reform begins to look a lot like a very old Medicaid program where the program contributed toward keeping the poor impoverished. Moreover, new cost-sharing reforms allowed for the very poor and enabled states to dis-enroll individuals who are unable to pay, which were never allowed previously, moving the program even further to the right of Medicaid's original design (and intent) because it arguably seeks to punish the poor.

In response to conservative claims that Medicaid has moved too far toward covering undeserving people of adequate means, Democrats have shied away from offering a robust defense of the virtues of expanding Medicaid to people of some means (or what some might call that completely ambiguous term, the “middle class”). Instead, Democrats have argued that Medicaid is vitally important for the “vulnerable” and for “hard-working” American families. Note, in both cases, their defense for expanding Medicaid ironically plays into a larger Republican frame of returning Medicaid to its original intent. First, using the term “vulnerable” primes one for a debate about who is truly vulnerable, and second, claiming that the newly eligible under a Medicaid expansion are deserving because they're “hard-working” suggests an underlying agreement about tying Medicaid deservingness to work. Although one would not expect a robust defense of Medicaid among Democrats in a predominantly conservative state, it is noteworthy that there is no readily apparent distinctive liberal frame for a transformative Medicaid expansion even in liberal expansion states.7 Given the importance of political discourse to move a vision for larger reform forward, the silence of an alternative frame is deafening.

## Notes

1. Of course, even in its original form the ACA excluded undocumented immigrants and legal immigrants in residence for less than five years, which compromised the universality claim.

2. Pennsylvania subsequently rescinded its grant application when a new governor was elected and decided to implement a straight expansion consistent with the ACA.

3. This is obviously not entirely true given the debates about whether immigrants should have access to the benefits of health reform and the decision to exclude those here lawfully only after a five-year waiting period; however, the older rhetoric related to work and health care benefits was largely absent.

4. Inclusion of National Public Radio in our content analysis was due to subsidiaries located in every state and their emphasis on in-depth state and local news. One challenge we faced was ensuring an ideological balance of media outlet accounts of Medicaid expansion. There is no analogous conservative-leaning media outlet similar in scope to National Public Radio, with a network of state-based subsidiaries. However, the nine newspapers included in our content analysis did exhibit conservative leanings, with five of the newspapers endorsing Mitt Romney for president in 2012, with three supporting Barack Obama, and one abstaining. Additionally, our data collection efforts included press releases and texts of speeches by key stakeholders in each state, ensuring that we also included these primary documents in our content analysis.

5. We acknowledge the Rosenbaum and Sommers (2013) article which asked this same question regarding Arkansas's Private Option.

6. There are also federal requirements for states to create streamlined enrollment processes.

7. We emphasize “readily apparent” in this sentence, since we have not done the discourse analysis in predominantly liberal expansion states to back up this claim.

## References

References
Allison
Andrew
.
2014
. “
Arkansas's Alternative to Medicaid Expansion Raises Important Questions about How HHS Will Implement New ACA Waiver Authority in 2017
.”
Journal of Health Politics, Policy and Law
39
, no.
5
:
1089
98
.
Andrews
Michelle
.
2014
. “
Arkansas Weighs Plan to Make Some Medicaid Enrollees Fund Savings Accounts
.”
Kaiser Health News
,
July
22
. khn.org/news/michelle-andrews-on-arkansas-plan-for-medicaid-savings-accounts/.
Associated Press
. “
Pennsylvania Awaits Ruling on Proposal to Link Work Requirements to Medicaid Benefits
.”
PBS NewsHour
,
PBS
,
April
27
,
2014
.
Balla
Steven
.
2001
. “
Interstate Professional Associations and the Diffusion of Policy Innovations
.”
American Politics Research
29
, no.
3
:
221
45
.
Beebe
Mike
. (
2013
). Quoted in Project Vote Smart. “
2013 Arkansas State of the State Address
Boehmke
Frederick
.
2005
.
The Indirect Effect of Direct Legislation: How Institutions Shape Interest Group Systems
.
Columbus
:
Ohio State University Press
.
Bookman
Todd
.
2013
. “
Medicaid Expansion Plan Approved by Commission
.”
,
October
8
. nhpr.org/post/medicaid-expansion-plan-approved-commission.
Boucher
Dave
.
2015
. “
Q&A: Insure Tennessee Legislative Supporters
.”
Tennessean
,
April
20
Boushey
Graeme
.
2016
. “
Targeted for Diffusion: How the Use and Acceptance of Stereotypes Shape the Diffusion of Criminal Justice Policy Innovations in the American States
.”
American Political Science Review
110
, no.
1
:
198
214
.
Bruner
Charles
.
2013
. “
Iowa View: Health Care Comes with Responsibility
.”
Des Moines Register
,
December
15
. www.desmoinesregister.com/story/opinion/columnists/2013/12/14/iowa-view-health-care-comes-with-responsibility/4025123/.
Case
Anne
,
Rosen
Harvey
, and
Hines
James
Jr.
1993
. “
Budget Spillovers and Fiscal Policy Interdependence: Evidence from the States
.”
Journal of Public Economics
52
, no.
3
:
285
307
.
Corbett
Tom
(
2014
). Quoted in Project Vote Smart. “
Gov. Tom Corbett's 2014–15 Budget Speech
Daniels
Frank
.
2015
. “
Governor, Time to Spend That Political Capital
.”
Tennessean
,
April
22
.
Davidson
Lee
.
2015
. “
House Unveils Counter to Healthy Utah, Though with Few Details
.”
Salt Lake Tribune
,
February
27
. www.sltrib.com/home/2231335-155/house-unveils-counter-to-healthy-utah.
Dennison
Mike
.
2013a
. “
Broad, Deep Support for Medicaid Expansion Shows Muscle at Capitol
.”
Billings Gazette
,
March
25
Dennison
Mike
.
2013b
. “
Montana Senate Dems, Handful of Republicans Roll New Medicaid Program into ‘Compromise’ Bill
.”
Billings Gazette
,
April
16
. billingsgazette.com/news/state-and-regional/montana/montana-senate-dems-handful-of-republicans-roll-new-medicaid-program/article_dc799a9c-7ceb-5113-8eee-c7bde4c4c7ce.html.
Dennison
Mike
.
2014
. “
Bullock on Medicaid: ‘More or Less a New Way, than Any Other State Has Taken.’
Billings Gazette
,
November
17
. billingsgazette.com/news/state-and-regional/montana/bullock-on-medicaid-more-or-less-a-new-way-than/article_5e878c89-4289-55df-81cf-d375d1b5f82f.html.
Farmer
Blake
.
2014
. “
Interview: Haslam on ‘Threading the Needle’ with Medicaid Expansion
.”
,
December
3
Gehrke
Robert
.
2015
. “
Herbert Leaves Health Reform to a Legislature That Has Shunned Earlier Attempts
.”
Salt Lake Tribune
,
October
15
. www.sltrib.com/home/3064148-155/herbert-now-up-to-legislature-to.
Glick
Henry R.
, and
Hays
Scott P.
1991
. “
Innovation and Reinvention in State Policymaking: Theory and the Evolution of Living Will Laws
.”
Journal of Politics
53
, no.
3
:
835
50
.
Governor Asa Hutchinson
.
2015
.
Transcript of Governor Asa Hutchinson's Healthcare Speech
. governor.arkansas.gov/press-releases/detail/transcript-of-governor-asa-hutchinsons-healthcare-speech.
Governor Bill Haslam
.
2014
.
Haslam Announces Insure Tennessee Plan
. news.tn.gov/node/13336.
Governor Maggie Hassan
.
2014
.
State of the State Speech
. governor.nh.gov/media/speeches/state-of-state-2014.htm.
Governor Rick Snyder
.
2013
.
Gov. Snyder Signs Healthy Michigan into Law, Bringing Health Care to 470,000 Michiganders
. www.michigan.gov/snyder/0,4668,7-277-57577_57657-312514–,00.html.
Governor Rick Snyder
.
2014a
.
Gov. Rick Snyder Says Healthy Michigan Plan Surpasses First-Year Enrollment Goal of 322,000
. www.michigan.gov/snyder/0,4668,7-277-57577_57657-332691–,00.html.
Governor Rick Snyder
.
2014b
.
Gov. Snyder: Healthy Michigan Plan Reaches 85,761 Enrollees After One Week
. www.michigan.gov/snyder/0,4668,7-277-57577-325714–,00.html.
Governor Rick Snyder
.
2015
.
Gov. Rick Snyder: Healthy Michigan Plan Has Enrolled More Than 600,000 People in First Year
. www.michigan.gov/snyder/0,4668,7-277-57577_57657-351257–,00.html.
Governor Steve Bullock
.
2015
.
Governor Steve Bullock Signs Montana HELP Act into Law
. governor.mt.gov/Newsroom/ArtMID/28487/ArticleID/1313.
Grogan
Colleen M.
1997
. “
The Medicaid Managed Care Policy Consensus for Welfare Recipients: A Reflection of Traditional Welfare Concerns
.”
Journal of Health Politics, Policy and Law
22
, no.
3
:
813
36
.
Grogan
Colleen M.
2008
. “
Medicaid: Health Care for You and Me?
” In
Health Politics and Policy
,
4th
ed., edited by
Morone
James
,
Litman
Theodor
, and
Robins
Leonard
,
329
54
.
New York
:
Delmar Thompson
.
Grogan
Colleen M.
2011
. “
You Call It Public, I Call It Private, Let's Call the Whole Thing Off?
Journal of Health Politics, Policy and Law
36
, no.
3
:
401
11
.
Grogan
Colleen M.
2013
. “
Medicaid: Designed to Grow
.” In
Health Politics and Policy
,
5th
ed., edited by
Morone
James
and
Ehlke
Daniel
,
142
63
.
Stamford, CT
:
Cengage Learning
.
Grogan
Colleen M.
, and
Andrews
Christina M.
2011
. “
The Politics of Aging within Medicaid
.” In
The New Politics of Old Age Policy
, edited by
Hudson
Robert B.
.
Baltimore
:
Johns Hopkins University Press
.
Grogan
Colleen M.
, and
Rigby
Elizabeth
.
2009
. “
Federalism, Partisan Politics, and Shifting Support for State Flexibility: The Case of the U.S. State Children's Health Insurance Program
.”
Publius: The Journal of Federalism
39
, no.
1
:
47
69
.
Grogan
Colleen M.
, and
Patashnik
Eric
.
2003
. “
Between Welfare Medicine and Mainstream Program: Medicaid at the Political Crossroads
.”
Journal of Health Politics, Policy and Law
28
, no.
5
:
821
58
.
Grogan
Colleen M.
, and
Smith
Vernon K.
2008
. “
From Charity Care to Medicaid: Governors, States, and the Transformation of American Health Care
.” In
A More Perfect Union
, edited by
Sribnick
Ethan G.
.
:
University of Pennsylvania Press
.
Groppe
Maureen
.
2013
. “
Expanding Medicaid Tied to ‘Personal Responsibility.’
Indianapolis Star
,
December
5
.
Grossback
Lawrence J.
,
Nicholson-Crotty
Sean
, and
Peterson
David A. M.
2004
. “
Ideology and Learning in Policy Diffusion
.”
American Politics Research
32
, no.
5
:
521
45
.
Harvey
David
.
2005
.
A Brief History of Neoliberalism
.
New York
:
Oxford University Press
.
Hays
Scott P.
1996
. “
Patterns of Reinvention: The Nature of Evolution during Policy Diffusion
.”
Policy Studies Journal
24
(
1996
):
551
66
.
Herbert
Gary
.
2012
. “
HERBERT: The Statehouse, Not White House, Should Lead on Health Reform
.”
Washington Times
,
July
31
Hertel-Fernandez
Alexander Warren
,
Skocpol
Theda
, and
Lynch
Daniel
.
2016
. “
Business Associations, Conservative Networks, and the Ongoing Republican War over Medicaid Expansion
.”
Journal of Health Politics, Policy and Law
,
41
, no.
2
:
239
86
.
Hughes
Greg
,
Speaker of the Utah House of Representatives
.
2015
. “
The Case for Reform and against Healthy Utah
.” Posted in Policy Buzz.
February
25
. utahpolicy.com/index.php/features/policy-buzz/4914-the-case-for-reform-and-against-healthy-utah.
Karch
Andrew
.
2007
.
Democratic Laboratories: Policy Diffusion among the American States
.
Ann Arbor
:
University of Michigan Press
.
Karch
Andrew
, and
Cravens
Matthew
.
2014
. “
Rapid Diffusion and Policy Reform: The Adoption and Modification of Three Strikes Laws
.”
State Politics and Policy Quarterly
14
, no.
4
:
461
91
.
Kauffman
Jacob
.
2013
. “
Arkansas Committee Discusses Differences between Medicaid and Private Option
.”
,
June
27
KFF (Kaiser Family Foundation)
.
2011
. “
.”
Washington, DC
:
Kaiser Family Foundation
.
KFF (Kaiser Family Foundation)
.
2016a
. “
Average Annual Growth in Medicaid Spending
.”
Washington, DC
:
Kaiser Family Foundation
.
KFF (Kaiser Family Foundation)
.
2016b
. “
Health Insurance Coverage of the Total Population
.”
Washington, DC
:
Kaiser Family Foundation
.
Kliff
Sarah
.
2013
. “
Privatizing the Medicaid Expansion: ‘Every State Will Be Eying This.’
Washington Post Wonkblog
,
March
8
. www.washingtonpost.com/news/wonk/wp/2013/03/08/privatizing-the-medicaid-expansion-every-state-will-be-eying-this/.
Michigan Health & Wellness
.
2016
. “
A Healthier Michigan Will Be a Stronger Michigan
.” www.michigan.gov/healthymichigan.
Mintrom
Michael
.
1997
. “
Policy Entrepreneurs and the Diffusion of Innovation
.”
American Journal of Political Science
41
, no.
3
:
738
70
.
Moulton
Kristen
.
2014
. “
Utah Gov. Gary Herbert: We Have a ‘Conceptual’ Deal on Medicaid
.”
Salt Lake Tribune
,
September
9
. archive.sltrib.com/story.php?ref=/sltrib/news/58395005-78/utah-health-herbert-medicaid.html.csp.
Moulton
Kristen
.
2015
. “
Panel Rejects Healthy Utah; Opts for House Alternative, Utah Cares
.”
Salt Lake Tribune
,
January
28
. www.sltrib.com/home/2249356-155/panel-rejects-healthy-utah-opts-for.
Noble
Jason
.
2013
. “
Branstad Promises Changes to IowaCare Program to Improve Health-Care Delivery
.”
Des Moines Register
,
February
26
Olson
Laura Katz
.
2010
.
The Politics of Medicaid
.
New York
:
Columbia University Press
.
Plazas
David
.
2015
. “
Haslam Counters Critics, Makes Case for Insure Tennessee
.”
Tennessean
,
February
3
. www.tennessean.com/story/opinion/2015/02/02/haslam-counters-critics-makes-case-insure-tn/22774731/.
Rogers
Everett M.
1983
.
Diffusion of Innovations
.
3rd
ed.
New York
:
Free Press
.
Roob
Mitchell
, and
Verma
Seema
.
2008
. “
Indiana: Health Care Reform amidst Colliding Values
.”
Health Affairs Blog
,
May
1
. healthaffairs.org/blog/2008/05/01/indiana-health-care-reform-amidst-colliding-values/.
Rosenbaum
Sara
, and
Sommers
Benjamin D.
2013
. “
Using Medicaid to Buy Private Health Insurance – The Great New Experiment
.”
New England Journal of Medicine
369
, no.
1
:
7
9
.
Shipan
Charles R.
, and
Volden
Craig
.
2006
. “
Bottom-up Federalism: The Diffusion of Antismoking Policies from U.S. Cities to States
.”
American Journal of Political Science
50
, no.
4
:
825
43
.
Shipan
Charles R.
, and
Volden
Craig
.
2014
. “
When the Smoke Clears: Expertise, Learning, and Policy Diffusion
.”
Journal of Public Policy
34
, no.
3
:
357
87
.
Sisk
Chas
.
2015
. “
Lawmakers Give Insure Tennessee Cold Shoulder
.”
,
January
27
.
Skocpol
Theda
,
Howard
Christopher
,
Lehmann
Susan Goodrich
, and
Abend-Wein
Marjorie
.
1993
. “
Women's Associations and the Enactment of Mothers' Pensions in the United States
.”
American Political Science Review
87
, no.
3
:
686
701
.
Sparer
Michael
.
2012
. “
Medicaid Managed Care: Costs, Access, and Quality of Care
.”
Princeton, NJ
:
Robert Wood Johnson Foundation
.
Tanenbaum
Sandra J.
1995
. “
Medicaid Eligibility Policy in the 1980s: Medical Utilitarianism and the ‘Deserving’ Poor
.”
Journal of Health Politics, Policy and Law
20
, no.
4
:
933
54
.
Thompson
Frank J.
2012
.
Medicaid Politics: Federalism, Policy Durability, and Health Reform
.
Washington, DC
:
Georgetown University Press
.
Volden
Craig
.
2007
. “
States as Policy Laboratories: Emulating Successes in the Children's Health Insurance Program
.”
American Journal of Political Science
50
, no.
2
:
294
312
.
Wacquant
Loic
.
2010
. “
Crafting the Neoliberal State: Workfare, Prisonfare, and Social Insecurity
.”
Sociological Forum
25
, no.
2
:
197
220
.
Wall
J. K.
2014
. “
The Most Likely Conflict Zone between Pence, Obama
.”
,
August
7
. www.ibj.com/blogs/12-the-dose/post/48956-the-most-likely-conflict-zone-between-pence-obama.
Whitney
Eric
.
2015a
. “
Medicaid's Western Push Hits Montana
.”
Kaiser Health News
,
January
6
. khn.org/news/medicaids-western-push-hits-montana.
Whitney
Eric
.
2015b
. “
Why Montana Hospitals Back Bullock's Medicaid Expansion Plan
.”