With increasing ideological polarization both within states and across states, policy makers face new challenges in developing and refining policies. This essay explores these challenges in the context of the spread of health policies across the states under the Affordable Care Act, highlighting key arguments and findings from the authors in this Special Issue. I discuss how common mechanisms of policy diffusion, the attributes of policies themselves, and the conditional nature of policy diffusion all play somewhat different roles during polarized times. In addition to new challenges to policy makers, polarization offers new opportunities for experimentation and learning that may be valuable to scholars and practitioners alike.
After decades of state experiments with varying policies of health insurance coverage, the Obama administration and Democratic Congress took a major step in 2010 toward nationwide universal coverage in the Affordable Care Act (ACA). In so doing, they changed the balance of state and national policy making in the area of health care, and brought to light the degree of ideological and partisan polarization that has arisen in American politics in recent years. Many Republican-controlled states refused to set up their own health care exchange websites and turned down highly subsidized grants for Medicaid expansion. When Republicans gained control of Congress, they voted repeatedly to repeal “Obamacare.” And yet, as time has passed, and the ACA policies have spread across the states and taken greater effect, such polarization has presented both opportunities and challenges for politicians and the public alike, as they work through the complexities of the new health care landscape.
Many view the ACA as a substantial centralization of health insurance policy in the United States. The degree to which policies are centralized or decentralized in federal systems is a crucial choice that impacts how well those policies address public needs. Policies that are determined at the state or local level tend to feature benefits such as horizontal competition, with policy makers seeking to innovate and hold down costs in order to attract and retain residents and businesses (Tiebout 1956). States and localities can experiment with various policy models, learn from one another, and spread successes to others (Walker 1969). Policies can be tailored to local circumstances and can be formulated by policy makers attuned to local knowledge. And the heterogeneous preferences of different populations can be reflected in diverse state and local policies (Volden 2005).
On the other hand, centralized policy making helps limit the spillovers that occur beyond local or state jurisdictions, attenuating harmful externalities. “Races to the bottom” that undercut the preferred policies and shred the social safety net can be halted before they start (Peterson and Rom 1990). And economies of scale can be realized, without all of the redundancies in policy making and implementation that decentralized policies bring about.
Peterson (1995) succinctly captured many of these benefits in his articulation of the argument that developmental policies are best left at the state and local level where they can be tailored to local needs, whereas redistributive policies are best handled nationally to avoid adverse state and local competition. Health care is a complicated policy area, fitting neatly in neither the developmental nor redistributive categories. Certainly, healthy citizens and a healthy workforce are critical to a robust economy. But providing such health care, especially during an era of high income inequality, involves some redistribution to those who cannot afford health insurance and health care services on their own.
In advancing the ACA, policy makers attempted to thread the needle between centralized and decentralized approaches, hoping to capture all of the benefits of each. Universal health insurance coverage could only be accommodated with some degree of redistribution, with subsidized premiums for low-income Americans. Thus, a national program was deemed essential to avoid the race-to-the-bottom incentives in state or local provision. Yet, the benefits of state experimentation might still be possible if there were some flexibility in how states developed the policy's specifics, and therefore the ACA encouraged state-run health exchanges and allowed for waivers to various parts of the policy to best meet local circumstances.
However, in this article, I argue that this attempt to capture all the best features of American federalism has been dramatically influenced by the highly polarized environment in which American politics is operating today. Observers of American politics cannot miss the partisan bickering and ideological polarization found in the US Congress. And such divisions are finding their way into numerous state legislatures (Shor and McCarty 2011). Not only are there strong ideological and partisan divides between “red states” and “blue states,” but those divides are also evident in the policy-making processes within states. These polarized environments are broadly affecting how American federalism functions today (e.g., Conlan and Posner 2016), and polarization influences how the ACA has been and will continue to be implemented.
This special issue of the Journal of Health Politics, Policy and Law has brought together influential scholars to analyze and comment on one aspect of health policy making under the ACA. Specifically, the essays herein each speak to the diffusion of health policy choices across the states in the wake of the ACA. Within political science, scholars have come to define policy diffusion as occurring when one government's decision about whether to adopt a policy innovation is influenced by the choices made by other governments (Graham, Shipan, and Volden 2013: 675). With this definition in mind, each essay in this issue takes a different angle, ranging from the roles of framing and rhetoric in the spread of reforms to the importance of changing public opinion, from the important political and policy drivers of ACA enrollment to how advocates could use the diffusion process to their greatest advantage. As such, each contributes in different ways to our understanding about the future of health policy in the United States. In total, these contributions tell the story of policy makers across the ideological spectrum struggling with the implementation of a landmark policy reform.
Moreover, the timing of this special issue could not be better. With Section 1332 “State Innovation Waivers” becoming available to the states in 2017 (McDonough 2014; Howard and Benshoof 2015), innovation and policy diffusion under the ACA may soon be more robust than ever. More profoundly, if the new Trump administration and Repulican Congress follow through on their promise to repeal and replace Obamacare, a new era of state-level experimentation will begin. Rather than merely summarizing the authors' contributions, I instead encapsulate them in a broad discussion of what we know about the diffusion of policy choices across the American states, how that knowledge may help us understand the past, present, and future of the ACA, and how both the ACA and policy diffusion more generally have been significantly affected by the polarized times in which we live.
Across the past two decades, scholars of policy diffusion have moved significantly away from the narrow view of diffusion being the geographic clustering of policy choices to a much more robust and theoretically grounded understanding of the complexities of policy choice across governments. We have come to: (1) identify a small set of driving mechanisms of diffusion; (2) highlight how the attributes of policies themselves have influenced their diffusion; and (3) explore how policy diffusion has been conditional on politics within potential adopting states. In this essay, I will focus on each of these three lenses for viewing policy diffusion, assessing how ideological and partisan polarization shapes their applicability to our understanding of state adoption and diffusion of the Affordable Care Act.
The Mechanisms of Policy Diffusion
Although various authors have invented more than one hundred adjectives and metaphors to colorfully describe policy diffusion processes (Graham, Shipan, and Volden 2013: 690), scholars have recently coalesced around a small number of concrete mechanisms through which policies spread across governments (e.g., Simmons, Dobbin, and Garrett 2006; Shipan and Volden 2008; Maggetti and Gilardi 2016). Here, I focus on the four mechanisms of learning, imitation, competition, and coercion, each of which comes into play in the spread of ACA policy choices across the states.
When states experiment with solutions to policy problems, they present an opportunity for other governments to learn from their experiences. Policy successes will spread, while failures will be abandoned, with little cost to those who do not engage in the experiment (e.g., New State Ice Co. v. Liebmann, 285 U.S. 262, 311  [Brandeis, J., dissenting]; Volden 2006). Such learning has been identified in health policy areas ranging from children's health insurance programs to antismoking policy choices. Under the Affordable Care Act, states had a variety of choices to make—most notably whether to set up a health care exchange and whether to expand Medicaid. Within these broad choices were many smaller choices, each of which could be made in light of the decisions of other states and the successes or failures of those policies.
Polarization within and across the American states affects policy learning in a number of ways. Potentially limiting learning and policy diffusion is the possibility that some policy options are completely dismissed due to ideological biases. Regardless of a policy's success or failure, its perceived ideological location may be so divergent from what is deemed acceptable to state policy makers as to render the policy's effectiveness immaterial (Volden, Ting, and Carpenter 2008). However, recent evidence from survey experiments of local officials across the United States suggests that such ideological biases can be substantially reduced and potentially overcome (Butler et al. 2017). If the earlier policy has been adopted by co-partisans or has a high degree of success, policy makers are at least willing to find out more about the policy, if not adopt it in the end.
Polarization might also lead to a proliferation of policy models from which others can learn. Rather than merely gravitating to the first policy that seems to address a problem, ideologically diverse policy makers may seek out something quite different. In the extreme, two or more quite divergent models may be attempted across the states, potentially with important lessons to be learned across key evaluative criteria. Such starkly different models may be very beneficial in helping policy makers understand the impacts of policy choices and ideological viewpoints. Recent competing models to respond to gun violence provide a case in point. Liberal policy makers have been pushing for gun control while conservatives have advocated for expanding gun rights. Such dramatically different policies may make each side very uncomfortable with one another's choices, but they certainly provide the opportunity to learn about the effects of different policies.
When the federal government takes a major new role in a policy area, such learning and experimentation may be diminished. The implications of health insurance experiments from Oregon to Massachusetts may not yet have come into full view by the time of the adoption of the ACA. As much as advocates of universal coverage argued that the law had been too long in coming, proponents of a federalism approach instead suggested that there had yet to be sufficient state experimentation on which a national law could be based. Whether the ACA gives states sufficient flexibility to still experiment and learn from one another is an open question ripe for additional research and examination.
Separate from learning, states may adopt policies found elsewhere through the diffusion mechanism commonly referred to as “imitation.” Imitation involves the copying of a policy found elsewhere without regard to its effectiveness. Largely what is involved in imitation is a sort of herding activity, wherein states are hesitant to be first movers but also do not desire to be left behind and appear out of sync with others, especially if doing so casts a negative light on elected policy makers.
Polarization adds a wrinkle to the classic imitation model. No longer are policy makers solely looking to do what is popular or widely accepted. Now they are looking to do what is widely accepted within their (potentially isolated) ideological community. If most other Democratic governments are adopting state health exchanges and Medicaid expansions, it becomes very difficult and politically dangerous for other Democratic policy makers to seek a different course. Likewise, a potentially treacherous road lies ahead for Republican governors and legislators who wander in such a liberal direction.
The role of polarization and imitation pressures as part of the ACA are very much on display in the essays within this special issue. For example, Callaghan and Jacobs (2017) show a strong negative relationship between Republican state policy makers and the adoption of state exchanges and Medicaid expansions. Grogan, McMinn Singer, and Jones (2017) highlight the challenges faced by conservative policy makers who nevertheless wanted to adopt a Medicaid expansion. They could not do so in a cookie-cutter manner of imitation that would make them seem to endorse Democratic proposals, but instead had to find the modifications and rhetoric needed to differentiate themselves from such a liberal position.
When states compete with one another, they may take actions to try to attract business from other states or to offer lower taxes and more efficient services. Sometimes such competition produces poor results as the quality and generosity of services are significantly reduced along with tax cuts. In an ideologically polarized environment, such competition may be reduced somewhat. States may differentiate their policies from others more for ideological reasons than to attain nonpolitical competitive advantages. Diminished competition therefore reduces both the benefits arising from adding market discipline to government policy making and the race-to-the-bottom costs found in competitive federalism.
Competition across states in their health policy choices under the ACA may therefore be limited. While citizens moving across state lines may find greater access to Medicaid benefits in certain circumstances, the ACA largely serves to diminish state-by-state differences. Nevertheless, the combination of learning, imitation, and competition led to a robust environment in which states made a variety of important decisions about how to implement the ACA. Early adoptions set the stage for what could be learned by others, and for the broad understandings of what would be seen as prototypical liberal or conservative policies.
In this special issue, Boehmke, Rury, Desmarais, and Harden (2017) establish how crucial the order of such initial policy adoptions was. They show that the right mix of innovative initial states with neighbors engaging in learning, imitation, or competition can dramatically speed up the adoption of many policy innovations. With regard to Medicaid expansion typically opposed by Republicans, which states' conservative policy makers act first, what arguments they make, and how others react will go a long way in determining whether key elements of the ACA take broader hold across the country or leave the health care landscape deeply divided. Policy advocates interested in a broader impact on that landscape may wish to target their state-level efforts carefully in light of the authors' findings.
Coercion is common in the spread of policies due to attempts of one government to force the hand of another, such as with trade restrictions across countries. In the US federal system, coercion mainly comes through top-down pressures from the federal government on the states in the form of regulations, preemptive policies, and intergovernmental grants.
The ideologically polarized environment both within Congress and in the states at the time of the passage of the ACA deeply influenced the coercive nature of the policy. Looking to attract bipartisan support and a sizable enough coalition in the Senate to overcome a possible filibuster, the percentage of federal funding for Medicaid expansion increased to very high levels. This made Medicaid expansion attractive to most Democratic governors and state legislatures. But it put Republican policy makers in the states in a difficult position. They could continue to promote small government policies and the private sector by limiting their Medicaid expansions. But, in so doing, they would be turning down a substantial amount in federal subsidies, while still likely supporting medical coverage for poorer families through other policies designed to help doctors and hospitals with unpaid bills.
More generally, when designing intergovernmental grant conditions, federal policy makers need to be mindful of the responses of the states (Volden 2007). During polarized times, the responses of one set of states may differ from those with different ideological positions. A grant sufficient enough to induce liberal policy makers to adopt a programmatic change may be insufficient to bring along conservatives. And yet, grant conditions targeting conservative lawmakers in the states may be inappropriately set for liberal officials. On top of such policy considerations are the political calculations, such that the majority party in Congress may be more concerned about state responses only by their own co-partisans, or may look to put politicians of the other party in a difficult political position, as with Republicans deciding how to react to incentives to expand Medicaid under the ACA.
The Role of Policy Attributes
Beyond the mechanisms that influence the spread of policies across states and localities, attributes of the policies themselves help determine which policies diffuse, and at what speeds. Makse and Volden (2011) apply the attributes typology of Rogers (2003) to policy diffusion. They establish that greater relative advantage, compatibility, observability, and trialability of policies enhances their diffusion across states, whereas policies' complexity reduces their spread. Moreover, they show that the mechanisms of diffusion are influenced by the policies' attributes.
“Relative advantage” captures the extent to which a new proposal is perceived to be an improvement over existing policy in a particular area. “Compatibility” characterizes the consistency between the new policy and prior laws, values, and experiences. “Observability” is the degree to which policy makers can see the choices of others and their effects. “Trialability” captures the possibility that a policy can be adopted and later abandoned at a low political or budgetary cost. And “complexity” characterizes whether a policy is difficult to adopt, understand, or implement.
Each of these policy attributes may be understood differently during ideologically polarized times than absent such divisions. For example, a policy seen to have a relative advantage by some policy makers may be perceived quite differently by others, who evaluate that policy according to different criteria. In this special issue, Karch and Rosenthal (2017) describe how Medicaid expansion is seen as less of an advantageous policy approach by Republicans than by Democrats. Until they started framing the policy as allowing lower taxes due to the increased federal subsidies, the expansion was perceived in a mainly negative light as an undue step away from the free market and toward big government.
Compatibility is also relevant in implementation of the ACA. Depending on their prior policy choices, some states could more easily select health care exchange models that matched their needs and values (e.g., Noh and Krane 2016). Such choices would in turn affect enrollment levels (Callaghan and Jacobs 2017). Grogan, McMinn Singer, and Jones (2017) highlight how Section 1115 waivers have allowed states to enhance the compatibility of Medicaid expansion to both past policy and the current political environment. More generally, as liberal and conservative policy choices diverge, many new policy proposals may be seen as incompatible with existing laws by one group or another. Increasing the flexibility of policy options may be crucial to the diffusion of such policy ideas.
Observability may be simultaneously enhanced and undermined given political polarization across the states. On the one hand, there may be many new experiments and models that states could observe and learn from. However, as discussed above, some policy makers may be hesitant to look to ideological opponents for such policy ideas. Such trade-offs in observability may apply not only to political elites but to the public as a whole. Pacheco and Maltby (2017) demonstrate how citizens in neighboring states take note of other governments' experiments and begin to demand change at home. They establish that the resulting shifts in public opinion in turn influence policy choices.
Trialability in the ACA may be enhanced by policy options, including state-by-state waivers such as those explored by Grogan, McMinn Singer, and Jones (2017). Yet, ideological polarization and the high-profile nature of state ACA policy choices make trying and then abandoning a policy politically perilous. Policy makers may be hesitant to abandon policies that seem to be failing in their states if doing so means they need to admit to their own failings in their earlier policy choices (e.g., Volden 2016).
Finally, there is little doubt that the Affordable Care Act is complex, both in itself and in how it affects state policy. Whether judged by the nearly two-thousand-page law itself or by the thousands of pages of regulations that followed during the implementation process, the details that needed to be spelled out were immense. As state policy makers waded into this area for their own purposes, they began to struggle with the complexities of matching the ACA's provisions to their own existing policies. From this point of view, difficult policy choices regarding implementation of the ACA and adapting the law's provisions (and modifications) to different states' circumstances will continue for many years to come. Adding to this complexity is the polarization across state policy makers, with their different perspectives influencing how each new development in the health care debates is perceived.
The Conditions for Policy Diffusion
Interacting with the mechanisms of policy diffusion and the attributes of the policies themselves are the natural differences across states that lead to conditional patterns in which policies spread and where they are adopted. For example, Shipan and Volden (2006) highlight how interest group strength and state legislative professionalism influenced whether states adopted various antismoking policies. In a similar manner, Callaghan and Jacobs (2017) note the role of administrative capacity for understanding the rollout of the ACA across the states and for subsequent enrollment levels.
Whether policies spread also depends on the role of public opinion (Pacheco 2012). Given the polarized political climate, not every governor leapt to embrace the ACA, accept grant funding, and develop a state health exchange. Rather, some governors only came to that position following other states' actions and the public opinion shifts that arose in their own states as a result, as Pacheco and Maltby (2017) establish.
Moreover, the spread of policies across the states is conditional on finding or developing a political climate in which such a change would be broadly accepted by politicians as well as the public. The framing of Medicaid expansions in terms of tax relief (Karch and Rosenthal 2017) and the rhetoric surrounding ACA waiver usage (Grogan, McMinn Singer, and Jones 2017) both offered conditions under which health care policies could spread from one state to the next. Especially when moving toward a policy change that may be in poor ideological alignment with state policy makers' predispositions, political choices are very much conditioned by policy perceptions.
Throughout this special issue, the authors have wrestled with state policy choices under the Affordable Care Act. They have explored the nature of policy diffusion within this landmark program. In so doing, classic themes of the mechanisms of policy diffusion and the attributes of policies have come into view once again. However, both the mechanisms of how policies spread and the role of policy attributes in that diffusion play out somewhat differently given how ideologically polarized US state governments have become. Policy makers may be loath to learn from or imitate those from the other party. Policies that are seen as relatively advantageous by one group may be viewed with suspicion by others.
As policy makers take further steps to implement or modify the ACA, they will need to navigate treacherous political waters. The authors writing here offer some sage advice and some words of caution along these lines. Rhetoric and framing have become crucial to building a coalition for policy change. The early actions of some states influence both public opinion in other states and their likelihood of adopting similar policies. And each of these choices has major implications for enrollment levels, medical care provision decisions, and ultimately the degree of success or failure that states experience under the ACA. Hopefully, this special issue will guide future scholarship and an understanding of such important issues for years to come.