Unraveling from Within? The Affordable Care Act and Self-Undermining Policy Feedbacks

Introduction

The tortuous path to enactment of the 2010 Patient Protection and Affordable Care Act (ACA) has been widely analyzed (Cohn 2010; Daschle and Nather 2010; Kirsch 2011; McDonough 2011; Starr 2011; Jacobs and Skocpol 2012; Brill 2015). Many ACA supporters anticipated that public support for the ACA would grow and partisan conflict would fade over time, with Obamacare gradually winning acceptance among its opponents – even if it was grudging and hostile acceptance in conservative quarters. Presidential advisor David Axelrod, for example, said in 2010 that “As the American people become familiar with what this program is and what it isn’t, they’re going to be very, very happy with it” (Baker 2012). As public support for the ACA developed and more Americans obtained benefits through health care reform, Democrats could anticipate that efforts to overturn Obamacare would recede.

In part, this expectation reflected prior experiences with programs that were controversial prior to enactment, but subsequently became much more politically popular and institutionalized – Medicare being the paradigmatic case in health care (Oberlander 2003). But the turn in political support has not occurred for the ACA. Congressional Republicans continue to pursue plans to “repeal and replace” Obamacare through legislative and legal challenges, and many states governed by the GOP have fiercely resisted the ACA’s implementation. Five years after enactment, public support for the ACA has been both quite stable and underwhelming, with more Americans opposed to (47%) than favoring it (40%), as of March 2015, though many of its individual provisions are more popular than the overall Act (Brodie et al. 2010; HuffPost Pollster 2015). Partisan differentials in attitudes toward the ACA remain remarkably high, as reflected not just in whether the law is viewed favorably, but also (and less plausibly) in self-reported perceptions of whether the ACA has helped or hurt the survey respondent (Jacobs and Mettler 2011; RAND Corporation 2014; Kaiser Family Foundation 2015a). In short, even as millions of Americans gained insurance coverage and other benefits due to the ACA’s policies – the uninsured population declined by about 9–12 million persons during the first year the law’s major coverage provisions were implemented (Blumenthal and Collins 2014; Long et al. 2014) – its political foundations remain shaky.

Anticipated feedback effects from the ACA played a strong role in the expectation that both support from the public and lawmakers would grow following Obamacare’s enactment. Several ACA provisions that provided concentrated benefits to individuals were “front-loaded,” slated to start shortly after the law’s passage, while some provisions that were likely to be unpopular or controversial were “back-loaded” to several years after the ACA’s major benefits started flowing. Moreover, tens of millions of Americans were expected to gain from the law’s expansion of access to health insurance coverage and new consumer protections. Given the scope of these benefits, and the experience of programs like Medicare that outgrew their controversial origins to attain mass popularity and robust political status, the Obama administration and Congressional Democrats could reasonably expect the ACA’s coalition to strengthen substantially over time. But to date that has not happened – and that is the fundamental political puzzle surrounding Obamacare.

The political logic of entrenching the ACA politically is consistent with a major stream of political science literature on policy feedbacks emanating from the seminal work of Paul Pierson (Pierson 1993, 1994, 2000). In this “historical institutionalist” approach to policy feedbacks, policies become entrenched by offering concentrated benefits to constituencies while many costs are delayed, diffused, or obscured. Because voters develop a sense of entitlement to benefits once they start receiving them, and politicians try to avoid blame for imposing cutbacks (Weaver 1986), benefits are unlikely to be rescinded once they start to flow because politicians fear the electoral consequences of doing so. Opposition fades because politicians, even those who oppose the program, make the political calculation that it is not worth alienating recipients of benefits or supportive publics.

The continued conflict and uncertainty over Obamacare’s fate suggest that an understanding of policy feedbacks as largely self-reinforcing is incomplete. In this article we sketch out a perspective on the political dynamics of policy feedbacks that is more comprehensive in the range of policy feedbacks that it considers. We discuss fiscal and administrative as well as socio-political policy feedback effects, and we discuss variation in the direction of those feedback effects (self-undermining, mixed, and self-reinforcing), in the timing of initial incidence of those effects (short-term, medium-term and long-term) and in their duration (temporary versus permanent). We also argue that the ACA’s political feedback effects are very different from that of entitlement programs like Social Security and Medicare in terms of the visibility and concentration of benefit flows as well as the fragmentation and political resources of beneficiaries. Understanding the broader range of feedback effects, we argue, helps to clarify why conflict over the ACA continues unabated and why its fate remains uncertain. A central argument is that, as the ACA experience illustrates, policies frequently produce feedback effects that undermine themselves as well as feedbacks that reinforce their base of political support. Many of these self-undermining feedbacks are initially unanticipated or underestimated, as politicians focus on “making a deal” that can win enactment in the gridlock-prone US system and on mobilizing public support. Our analysis also makes clear how sensitive policy feedback effects are to the political and policy environments in which those policies are generated, interpreted, and modified.

We begin by outlining existing approaches to the analysis of policy feedback effects. Next, we briefly review the ACA’s enactment and outline its core provisions. We then examine the ACA’s key provisions and the complex feedbacks that they generated. We conclude by discussing the implications of our analysis for understanding the role of policy feedbacks in public policy and for the ACA’s political trajectory.

Understanding Policy Feedback Effects

The dominant view of policy feedback effects associated with the historical institutionalist approach emphasizes long-term, frequently slow-moving, self-reinforcing effects on political support for programs. As constituencies grow up around programs and public support for those programs builds, the menu of alternatives considered is likely to shrink to incremental – and generally expansionary – adjustments to existing benefits, unless there are exogenous shocks to the “policy regime” (the set of policies in a particular sector) that is in place.

A second major literature on policy feedbacks, associated with the work of Frank Baumgartner and Bryan Jones (2002, 2009; Jones and Baumgartner 2005) and their collaborators, emphasizes the dynamics of punctuated equilibria. Because policymakers face more potential areas for action than they can handle, policy sectors are often dominated by “policy monopolies” of influence that derive benefits from the status quo. Such monopolies reinforce their control by developing “formal or informal rules of access [that] discourage the participation of “outsiders,” and…prevalent understandings of the policy [that] are so positive…they evoke only support or indifference by those not involved” (Baumgartner and Jones 2009). Thus policymaking “is very static but reluctantly changes when signals are strong enough” (Jones and Baumgartner 2005, p. 19), usually as the result of exogenous shocks to existing equilibria such as focusing events, which cause increased attention to be devoted to that subsystem by outsiders who are less likely to be concentrated beneficiaries of the status quo. Proponents of the status quo are often able to re-assert their monopoly as the outsiders’ interest in the policy wanes.

Recent research on policy feedback effects attempts to integrate and extend these literatures. Eight broad conclusions emerge from this literature. First, policies are not always self-reinforcing (Campbell 2011; Patashnik and Zelizer 2013). Policy reforms can be eroded or reversed (Patashnik 2008), and policies may even have feedback effects that endogenously undermine those policies over time (Weaver 2010; Jacobs and Weaver forthcoming).

Second, policy feedback mechanisms and effects vary both in terms of type and intensity. In Table 1, we divide policy feedbacks associated into three broad categories: socio-political, fiscal, and administrative with distinctive causal mechanisms (Weaver 2010; Jacobs and Weaver forthcoming; for a similar categorization, see Patashnik and Zelizer 2013). Socio-political policy feedback effects concern whether public and elite support for a policy regime is reinforced or undermined over time. These mechanisms and effects include feedbacks at the level of mass cognition (e.g., is the policy perceived to be successful or disastrous in achieving its objectives and does it lead predominantly to the mobilization of supporters or opponents?). Whether the policy leads to a menu of reform options that is truncated to solely incremental changes or stimulates a search for more dramatic “regime-shifting” reforms (usually in response to a perception of serious and unavoidable policy failings) also influences the odds that a policy remains stable or is subject to revision.

Fiscal effects concern whether a program creates budgetary strains that are likely to raise concerns among powerful actors, notably Treasury or Finance Ministries, weakening the autonomy of groups dominating a policy subsystem. Administrative feedback effects encompass design and operational elements that allow the agencies in charge of implementing a policy to retain or acquire the capacity to implement the program in a manner that is perceived as successful by internal and external constituencies. This in turn will result in the strengthening or weakening of agency morale, external reputation, and external political support. Self-undermining feedback effects of all three types (socio-political, fiscal, and administrative) are likely to involve developments in the program that disrupt relationships between policy “insiders” and “outsiders” in ways that increase attention from and involvement by outsiders because of perceptions by those groups of heightened political salience or costs, budgetary strains, or administrative failures.

Third, few policy feedback effects are “purely” endogenous. Feedback effects are more likely to be self-undermining rather than self-reinforcing under some conditions than others. Indeed, a fourth and related point is that feedback effects can change over time in their direction and intensity: feedback mechanisms that reinforce a program at one point in time may undermine it at another point in time as conditions change. Earmarked revenue sources illustrate both of these points. Earmarking revenue may insulate a program and foster incremental expansion when revenues are flush but spark retrenchment or even a more radical consideration of policy alternatives when revenues run short. And whether revenues in trust funds are flush or lean in turn depends on conditions such as macroeconomic growth and demographic shifts that alter the ratio of contributors to beneficiaries. Similarly, public perceptions of costs and risks will be higher when political elites have incentives to highlight those costs or risks. Self-undermining administrative feedbacks will be more prevalent when policy mandates given to administering agencies are intrinsically complex and difficult.

Fifth, policy feedback effects are often mixed in direction rather than unidirectional. Pre-2010 health policies in the US, for example, had a number of self-reinforcing feedback effects, notably engendering a sense of entitlement among Medicare recipients and a fear among those with good employer-sponsored health insurance plans that reforms to expand coverage to the uninsured that simultaneously altered their insurance arrangements would make them worse off. But Medicare policies also engendered concerns about gaps in its benefits and the reliance on employer-sponsored insurance produced anxiety among workers over the potential loss of coverage, particularly during times of economic downturn, and perennial concerns over affordability given rising premiums.

Sixth, policy feedback effects may be either fast or slow-moving. Major focusing events can provoke an immediate policy response. Long-term trends may be more akin to the famous (albeit apocryphal) Al Gore anecdote about a frog being boiled in water that is heated slowly, such as the slow but seemingly inexorable rise in health care’s share of the US economy. Slow-moving policy feedbacks are less likely to provoke an intervention by constituencies external to a policy subsystem, but they may also be less likely to experience waning attention by those interests in the latter stages of the issues attention cycle outlined by Anthony Downs (1972).

Seventh, self-undermining feedback effects are rarely a sufficient cause of policy change. While suffering concentrated costs may increase concern and an effort to find policy alternatives among affected groups, procedural rules that limit their participation and other factors can create barriers to acting on that disaffection. They may have to wait until sympathetic politicians gain power in order to redress their grievances, and in systems with multiple veto points like the US, even severe self-undermining policy feedbacks can result in the persistence of problems rather than in policy change.

Eighth and finally, when seeking to enact policy reforms, politicians and group leaders may seek to manipulate policy feedback mechanisms and effects to achieve policy and political objectives – for example by front-loading concentrated benefits and back-loading concentrated costs. However, they are likely to encounter a variety of constraints in their efforts to do so. Fiscal rules and constraints that require increased costs and benefits to be balanced within a particular time period, for example, can limit front-loading of concentrated benefits. The need to build political support coalitions for policy change, including actors who reject preferred options of coalition leaders, may require a political compromise that makes some strategies for manipulating feedback mechanisms untenable.

In sum, policy feedbacks are not always self-reinforcing, and such feedbacks can, under certain circumstances, actually undermine a program’s political support. These dynamics are visible in the political life of the Affordable Care Act.

The Politics of ACA Enactment

The history of health care reform in the US is largely one of failure (Starr 1982; Blumenthal and Morone 2009; Altman and Shactman 2011; Hoffman 2012). The Obama administration and Congressional Democrats overcame long odds to win passage of the Affordable Care Act. The extraordinary politics of the ACA’s enactment have been discussed extensively elsewhere (Cohn 2010; Daschle and Nather 2010; Oberlander 2010; Brown 2011; Hacker 2011; Kirsch 2011; McDonough 2011; Starr 2011; Jacobs and Skocpol 2012; Brill 2015). Here we focus on the constraints and opportunities that reformers faced during 2009–2010, as well as on the assumptions, decisions, and compromises that shaped the ACA, its implementation, and the resulting socio-political, fiscal, and administrative policy feedbacks.

Reformers presumed that the only viable reform model was one that built on existing health insurance arrangements and left as many insured Americans as possible in their existing plans. The Obama administration and Congressional Democrats believed that the ill-fated Clinton health care plan of 1993–1994 had imploded largely because it was too disruptive to the existing system and alienated the well-insured (Oberlander 2010). In addition, they were strongly influenced by the example of Massachusetts, which in 2006 enacted an ambitious expansion of insurance coverage that leveraged both existing public insurance and private insurance plans – and that drew wide bipartisan support (McDonough 2011). The Massachusetts model combined policies and ideas associated with conservatives (an individual mandate, private plan competition, individual choice of plan) with policies and goals (Medicaid expansion, expanding access to affordable insurance, regulation of private insurers) associated with liberals, a combination the ACA would emulate. The Massachusetts experience suggested that such a model could provide a political blueprint for passing health reform through Congress and produce a workable approach to expanding insurance coverage.

Another lesson taken from the Clinton health care misadventure was the necessity of winning over interest-group support. The Clinton administration had fought an unsuccessful multi-front war against health care stakeholders, including the insurance industry and business interests, a battle the Obama administration very much wanted to avoid (Oberlander 2010; Brown 2011). That in turn meant constructing a reform plan that accommodated important interests, for example by building on private insurance and making a deal with the pharmaceutical industry to not pursue stronger cost controls in exchange for their support.

Democrats’ reform vision in 2009–2010, then, ruled out more sweeping models, such as single-payer, Canadian-style national health insurance. Instead, the aim was to find ways to fold the uninsured into the prevailing US health insurance system, leaving its major pillars – Medicare, Medicaid, employer-sponsored private insurance – in place so as maximize the chances of securing both public and stakeholder support for reform. The ACA is ambitious in its scope and aspirations, yet its approach to health reform is also incremental in key respects. The ACA represents an effort to place an additional series of patches on the existing patchwork of US health care (Marmor and Oberlander 2011).

In covering the uninsured, reformers confronted a formidable fiscal challenge, namely how to pay for expanding insurance. The Congressional Budget Office (CBO) would not credit much in the way of savings to politically appealing measures such as promoting prevention or paying on the basis of value rather than volume (in other words, changing the payment system for medical services to reward doctors and hospitals for better health outcomes and adherence to clinical guidelines, rather than reimbursing medical providers for each service provided regardless of its necessity or value). That left Democrats with an array of politically difficult and controversial options to pay for health care reform, from cutting Medicare payments to raising taxes on individuals. The Obama administration decided that politically it could not afford to spend more than $1 trillion on health reform, which would have major implications for the type of coverage offered through the ACA (Oberlander 2014a). Further, the Democratic coalition in Congress contained a number of Blue Dogs, fiscal conservatives who insisted that any health reform plan not increase, and preferably decrease, the federal budget deficit. The administration promised to abide by PayGo principles, which meant they had to find about $1 trillion in revenues and spending offsets over a decade to pay for the ACA.

Finally, the ACA relied extensively on states to implement health reform (Starr 2011). Many of the uninsured were to gain coverage through the state-administered Medicaid program. And states were given the opportunity to establish health insurance exchanges, purchasing pools where the uninsured and small business could go to obtain coverage (Jones, Bradley, and Oberlander 2014). Liberals had pressed for a national exchange, but Democrats’ filibuster-proof majority of 60 in the Senate, which included Ben Nelson of Nebraska and others hostile to an expansionary federal role, made such a scheme impossible to pass; a Medicare-like public insurance option for the uninsured failed to clear the Senate for similar reasons. When the House passed the Senate bill in 2010 after Scott Brown’s victory in a special Massachusetts election deprived Democrats of their filibuster-proof Senate majority, and reconciliation was then used to pass a “sidecar” bill to revise the law, the Senate’s version of state-run exchanges remained intact. The federal government would instead serve as a backup, establishing exchanges in states whose governments declined to set up their own (McDonough 2011).

At the time of the ACA’s enactment, that arrangement did not seem crucial, since the vast majority of states were expected to embrace the chance to craft their own exchanges, nor did the reliance on Medicaid appear problematic since all states were expected to expand it given the alternative of losing all of their federal Medicaid funding. Subsequent events, though, would underscore just how crucial the reliance on state implementation was (Jones, Bradley, and Oberlander 2014).

The ACA’s Core Provisions

Given these assumptions, compromises, and constraints, what kind of health care reform did the US enact in 2010? The ACA contains a broad array of policies that aim to increase health insurance coverage, control health care spending, and transform how medical care services are delivered and paid for (McDonough 2011; Kaiser Family Foundation 2013; Oberlander 2014a). Obamacare seeks to enhance access to insurance by providing income-related subsidies to low- and middle-income Americans to help them purchase insurance (subsidies are available for persons up to 400% of the federal poverty level). Additionally, the ACA establishes health insurance exchanges (also known as marketplaces) where the uninsured can use their subsidies and choose from private insurance plans that offer varying levels of coverage. Private insurers are regulated so they cannot deny coverage or charge higher premiums to persons with pre-existing conditions. Annual and lifetime limits on benefits are prohibited. Minimum standards for insurance benefits are set and the ACA requires enhanced coverage of preventive services. Young adults can stay on their parents’ insurance plans until age 26. The law also requires most individuals to obtain and larger employers to offer health insurance coverage or pay penalties. Finally, the ACA called for expanding Medicaid eligibility to all persons who earned <138% of the federal poverty level ($16,243 for an individual in 2015), regardless of their demographic category. Previously, in order to qualify for Medicaid enrollees had to be in a specific demographic category, such as pregnant women and children, and income thresholds for eligibility for adults varied widely across states.

These provisions were expected to have a major impact on health insurance coverage. The Congressional Budget Office (2010) projected that the ACA would reduce the uninsured population by 30 million persons by 2016. The ACA funded this coverage expansion through a combination of new taxes and projected savings in federal health care expenditures (Kaiser Family Foundation 2013). Americans making over $200,000 a year ($250,000 for couples) must pay higher Medicare payroll taxes, and their investment income is now subject to the Medicare payroll tax. Penalties paid by individuals who do not obtain insurance and larger employers who do not offer coverage to their workers also are expected to contribute significantly to the ACA’s financing. The ACA additionally imposes taxes on the health care industry, including medical device manufacturers, for-profit insurers, pharmaceutical manufacturers, and tanning salons. A tax on high-cost private health insurance plans – the so-called “Cadillac tax” – is scheduled to go into effect in 2018. New limits on flexible spending accounts and the tax deductibility of health care expenses are also established (Kaiser Family Foundation 2013).

The ACA’s funding relies as well on substantial savings from reducing the projected rate of growth in federal health care spending (CBO 2010). The law restrains the growth rate in Medicare payments to hospitals and cuts payments to the private insurance plans (Medicare Advantage) that contract with Medicare. These policies, particularly the reduction in hospital payments, are expected to produce a significant slowdown in Medicare spending (indeed, there is evidence that these provisions are already generating savings). However, the ACA contained a back-up mechanism to restrain Medicare spending growth: the Independent Payment Advisory Board (IPAB). The IPAB was envisioned as a nonpartisan board comprised by health policy experts. If Medicare spending exceeded thresholds specified in the law, the IPAB would be empowered to make recommendations to reduce program expenditures and if Congress did not act, the IPAB’s recommendations would then become law (Oberlander and Morrison 2013).

While the ACA reduces the growth in Medicare payments to medical providers, it also aims to control costs by changing how hospitals, doctors, and other providers are paid, and by changing how medical care services are delivered, thereby impacting health care spending across the entire health care system. The ACA encourages a variety of experiments in health care payment and delivery, including Accountable Care Organizations, medical homes, penalties on hospitals for high readmission rates, payments to doctors and hospitals based on value (so-called pay for performance), and bundled payment (Oberlander 2011). Advocates of these strategies tout their potential to simultaneously control health care spending while improving the quality of medical care delivery and health outcomes. Additionally, within the ACA’s health insurance exchanges, reformers sought to control insurance costs by relying on a system of competition between health plans.

In sum, the ACA contains a wide array of policies that aim not simply to expand access to health insurance coverage, but also to control health care spending and improve the quality of medical care.

Conclusions

Examining the enactment and subsequent implementation of the Affordable Care Act provides important insights not only into its own difficult history but also into the complex dynamics of policy feedbacks in general, and how they can vary even within a single policy sector. Five years after the 1965 enactment of Medicare, few politicians could imagine the repeal of that program, and fewer still would advocate it. Five years after enactment of the Medicare Catastrophic Coverage Act of 1988, it was only a painful memory for those who enacted it. Five years after passage of the Affordable Care Act, its fate is still very much up in the air.

These differing fates reflect, in part, these programs’ different policy feedbacks. Medicare provided enormous immediate benefits to the elderly for a modest price, and followed the established and popular social insurance arrangements of Social Security, including payroll tax financing. The Medicare Catastrophic Coverage Act also sought to extend benefits to seniors, but it departed from traditional social insurance and Medicare arrangements by having seniors self-finance the benefits (rather than relying on contributions from the working public) and imposing high, visible, and immediate costs on more affluent Medicare enrollees while delaying benefits. As a consequence, it triggered intense opposition: even Medicare beneficiaries who would have benefited from the law were confused by its financing mechanism, and the program quickly imploded politically (Himelfarb 1995; Oberlander 2003; Patashnik 2008).

The Affordable Care Act lies in between these two extremes. In contrast to Medicare, the ACA is not a social insurance program with a clearly identified constituency that has a clear connection to benefits (Marmor and Oberlander 2011). The ACA treats different groups of Americans in different ways at different times, and that fragmented structure makes it harder to build support for the law, harder to recognize benefits that do exist, and easier to attack, since the law’s constituency is diffuse and many intended beneficiaries, including many of the uninsured, do not understand the benefits available to them.

Within this fragmented programmatic structure, the Obama administration and its allies did seek to manipulate the ACA’s feedback effects in ways that would benefit their political and policy interests, notably in frontloading some concentrated benefits and back-loading some important concentrated costs. However, they were highly constrained in how much they could do so by super-majority requirements to enact controversial legislation in the Senate, as well as by fiscal rules and Congressional Democrats’ internal divisions on substance and strategy. The complex history and many compromises needed to win congressional passage of the ACA embedded many self-undermining feedbacks – socio-political, financial, and administrative – and left it vulnerable to even more negative feedbacks encouraged by forces hostile to the ACA who could both challenge its legality and affect the way it was implemented. Reformers and the Obama administration, then, underestimated the impact of self-undermining policy feedbacks and overestimated the extent to which the ACA would be politically institutionalized quickly and transform its political fortunes as Medicare did. Why did this occur?

The degree to which these policy feedbacks, and the ensuing conflict, were inevitable or reflect poor strategic choices by the Obama administration and its political allies is certainly debatable. Blaming poor strategy suggests that if the Obama administration and its allies had been more prescient and more politically savvy, they could have avoided many of the problems that the ACA subsequently encountered. This explanation requires that the administration (1) failed to see warning signs of trouble ahead that could have reasonably been anticipated, and (2) failed to take actions in light of those warning signs that would have made a difference.

Certainly, if Democrats had noticed the problematic wording in the ACA’s text related to the availability of subsidies in federally sponsored exchanges, they could have taken steps to revise the legislative language in ways that would have never allowed a legal challenge like King v. Burwell to emerge. The cancellation of individual insurance policies during the fall of 2013 is one key area where the Obama administration appears to have overlooked the potential fallout of their own actions and the loss-imposing nature of some ACA provisions. Additionally, the ACA’s individual and employer mandates were highly controversial even before they went into effect. Indeed the political dynamic of the individual mandate strongly underlines how resourceful messengers can build strong opposition to a policy even before it has gone into effect. In a political environment that is both highly polarized and lacking slack resources to provide new benefits to attract voters, negative messaging becomes the major coin of political discourse in both politics and policymaking (Weaver 1986; Lee forthcoming). These negative messages can undermine the capacity of new policy to become popular and politically institutionalized, leaving program more vulnerable to challenges.

ACA proponents also underestimated the ability of critics to frame critiques (some more honest than others) about negative effects. ACA opponents in the Republican Party and their media allies repeated and amplified concerns from the real (cancellation of individually purchased policies, the troubled launch of healthcare.gov) to the mythical (death panels, cuts in Medicare benefits) and wove them into a broader narrative of Obama administration duplicity and widespread loss imposition (Eckles and Schaffner 2010).

The Obama administration, Congressional Democrats, and health reform advocates appear to have badly miscalculated the political environment. Partisan polarization explains much of the ACA’s enduring political problems, as partisans framed debate around ACA’s negative impacts, real and imagined, emphasized loss-imposing policy consequences rather than the law’s beneficial impacts, and resisted the ACA’s implementation at the state and federal level. Because of compromises made in the Senate and the 2012 Supreme Court ruling, states play a crucial role in health reform implementation – a role far greater than the ACA’s architects anticipated. ACA proponents underestimated state resistance to both exchanges, which had been an idea previously supported by many Republicans, and Medicaid expansion. After the 2010 elections and rise of the Tea Party, state politics moved to the right and GOP governors had a stronger incentive to boycott Obamacare and Medicaid expansion (Jones, Bradley, and Oberlander 2014). They have done so, despite, in the case of Medicaid, financial lures from the federal government: ideology and partisanship has trumped fiscal pragmatism in many states. Federalism clearly provided an opportunity for opponents to undermine the law’s implementation and operation, and to emphasize and indeed create both real and apparent loss-imposition.

It is less clear, however, what ACA proponents could have done to avoid the law’s federalism-related problems. While strategic mistakes certainly added to the self-undermining feedbacks of the ACA and added to its political and legal problems, most of its problems either seem difficult to anticipate or are due to constraints imposed by the political environment and the legacy of past policies that gave the ACA’s proponents limited room to maneuver. The rise of the Tea Party and its effect on forcing Republican officeholders into ever more intransigent stances on implementation of and legal challenges to the ACA could hardly have been anticipated in 2009.

Even if ACA proponents had anticipated these problems, it is unlikely that they could have done much about them that would have fundamentally altered outcomes. Perhaps their most significant failure was not realizing that the ACA really would inevitably impose costs (e.g., cancelled insurance policies) as well as benefits in the short term. As noted earlier, politicians tend to minimize negative feedbacks at the time of a program’s enactment because they are focused on getting legislation through the Congressional gauntlet and on mobilizing public support. The ACA’s architects faced extraordinary obstacles to enacting health reform, which created strong incentives for them to minimize the law’s self-undermining feedbacks. But if Democratic lawmakers had perfect political foresight – given the large number who subsequently lost their seats – it is likely that some never would have voted for the ACA in the first place and it would not have passed Congress.

As noted at the outset of this article, public opinion on the ACA has been remarkably stable over last 5 years: despite initial problems with the rollout of the health insurance exchanges during 2013–2014 and subsequent successes in enrollment, public support for the ACA is quite similar to that in 2010 and consistently more negative than positive. Even if healthcare.gov’s opening had been smooth and nobody had individual policies cancelled, it is not clear that the ACA be much more popular and under significantly less attack: healthcare.gov is working well now, millions of Americans have gained insurance coverage, and yet the ACA still draws tepid public support and remains in a precarious political position. The importance of partisan polarization (McCarty, Poole, and Rosenthal 2006) in producing this prolonged uncertainly and conflict surrounding the ACA is impossible to overstate. Polarization has served as a crucial mediating factor in the politics of Obamacare, amplifying self-undermining feedbacks while obscuring self-reinforcing policies and benefits.

In sum, we have emphasized that (1) policies may generate self-reinforcing and self-undermining feedback effects simultaneously, (2) the balance of these feedback effects may shift over time; (3) it is the balance of these effects that matters politically, with blame-avoiding and blame-generating politicians paying particular attention to policy consequences that appear to generate losses for particular groups; (4) feedback effects, especially socio-political feedbacks, are in part social constructs that reflect the framing that key shapers of opinion convey to the public about the impact of those policies, and they are mediated by external factors like the level of partisan polarization; and (5) while self-undermining policy feedbacks may shape the political debate over revising a policy, the final outcome depends heavily on the political bargaining leverage of key actors and the institutional rules involved in changing policy. This case study of the Affordable Care Act suggests a number of extensions of this argument. For example, not all self-undermining effects are equivalent: some, like the costs imposed by the individual mandate, risk undermining an entire reform package of which they are one component. Others, like the Cadillac tax, mostly undermine themselves without jeopardizing the whole law.

The Democrats’ initial calculation in passing the Affordable Care Act was that if they could get a policy of “near” universal coverage in place, offer benefits to both large numbers of uninsured and insured Americans, pay off the states, neutralize key interest groups, and backload at least some of the concentrated costs, the policy would be impossible to get rid of. Rather than being a slam dunk, however, the ACA’s avoiding repeal or major rollback for the past 5 years has depended on maintaining veto power in the presidency and (through the super-majority required for cloture) the Senate.

Republicans continue to push for Obamacare’s repeal. But Republicans have their own political problems: a pre-ACA status quo ante with many highly visible defects, and no credible alternative that would not make many individuals worse off (Pear 2015). If the Supreme Court upholds the government’s position in King v. Burwell, the GOP will confront a political reality where millions of Americans, and the insurance and hospital industries, are benefiting from subsidized insurance coverage and Medicaid expansion. The longer the ACA goes on, the more beneficiaries it gains, and the more the medical care industry comes to depend on it for revenue, the harder it will be for Republicans to dislodge Obamacare, even if the GOP wins the White House in 2016 (Oberlander 2014b). Still, the past 5 years have shown just how difficult it is to institutionalize a major new social program with complicated and mixed policy feedbacks in a polarized partisan environment. The fight over Obamacare is not over.