The American Democratic leadership in the White House and Congress in 2009–10 and the British Conservative/Liberal-Democrat Coalition government in 2010–12 each pursued a strategy of rapidly assembled multiple adjustments to the prevailing policy framework for health care rather than attempting a “big-bang” strategy of sweeping institutional change. Despite their relative modesty, each set of reforms encountered a highly conflictual and tortuous process of legislative passage. Subsequently, the reforms failed to gain broad public acceptance and were variously hobbled (in the United States) and transformed (in the United Kingdom) in the course of implementation. These two cases thus offer some common lessons about the potential and the pitfalls of such complex “mosaic” reforms.
To paraphrase a clever proverb about history, the politics of health care policy does not repeat itself across national regimes, but on occasion it rhymes. One such rhyming couplet occurred in the United States and Britain at the end of the first decade of the twenty-first century. The American Democratic leadership in the White House and Congress in 2009–10 and the British Conservative/Liberal-Democrat Coalition government in 2010–12 each pursued a strategy of rapidly assembled multiple adjustments to the prevailing system rather than a “big-bang” introduction of a sweeping new institutional regime. Despite their relative modesty, each set of reforms encountered a highly conflictual and tortuous process of legislative passage. Subsequently, the reforms failed to gain broad public acceptance and were variously hobbled (in the United States) and transformed (in the United Kingdom) through the implementation process. These two cases thus offer some common lessons about the potential and the pitfalls of such complex “mosaic” reforms.1
Mosaic reform strategies represent one of four possible strategic approaches to the scale and pace of policy change (Tuohy 2018). By scale I mean the degree and scope of change in the prevailing institutional logic, that is, change in those dimensions of the policy framework that set the balance of influence among key interests, such as the medical profession, the state, and private finance; the mix of instruments, such as hierarchical control, market exchange, and peer persuasion; and the organizing principles defining the function of the state and the obligations and entitlements of citizens. By pace I refer to the pace of enactment: are the changes enacted all at once in a single piece of legislation or through successive bills over time? The intersection of scale and pace yields four tendencies: large-fast, large-slow, small-fast, and small-slow. The two principal cases under review here are examples of a strategy of small-fast “mosaic” change (multiple adjustments, enacted as a piece, that do not fundamentally alter the prevailing institutional logic).
The scale and pace of policy change depend on political actors' assessments of their ability to overcome vetoes in the present and over time: how can a reform coalition be built, and how long can it be sustained? In each of the American and British cases reviewed here, the parties in power had strong partisan incentives to move quickly to enact a health care reform that could be presented as a bold achievement. In each case the political leadership judged that it had the necessary institutional and electoral resources to overcome vetoes in the short term but risked losing its position of advantage within a brief time frame. Despite the radical differences in their political institutions and health care systems, each set of political leaders found themselves faced with similar questions of political strategy: how to bridge differences across independent actors within the governments of the day, in order to assemble a winning coalition for a major health care reform within a shrinking time horizon. In so doing they created micro-constituencies of satisfaction with particular elements of the resulting reforms without building understanding or support for the overall package.
American and British Similarities
In 2009 the Democrats under Barack Obama achieved their best electoral results since the Lyndon Johnson's landslide of 1964. They took the presidency with a solid majority of the popular vote and won control of both chambers of Congress, including for a brief period the razor-thin 60% supermajority in the Senate technically necessary to overcome a Republican filibuster by closing off debate. Obama chose to capitalize on these advantages by tackling a push to universal health insurance coverage that had remained “unfinished business” for the Democrats since the enactment of the targeted Medicare and Medicaid programs (for the elderly and social assistance recipients respectively) under President Johnson in 1965. The window of opportunity was seen to be very tight: Obama and the Democratic leadership initiated the process of drafting health care reform legislation in his first budget within 2 months of his inauguration.
While his position was arguably more favorable than that of any of his predecessors, including even President Johnson (Peterson 2011), Obama still lacked the cohesive coalition necessary for a sweeping institutional change. The independent electoral bases of members of Congress meant that, although the Democratic congressional caucuses were more united than they had been since the beginning of the twentieth century, the leadership still could not rely on the full compliance of their Senate caucus to forestall opposition filibusters. Moreover, the Democrats gained and lost even the nominal cloture threshold in the Senate within 9 months—just long enough for legislation to pass both House and Senate. Reconciling the two bills and completing the process would require yet more bargains and adjustments within the constraints of the process of budget reconciliation requiring only a simple majority.
The result was a layering on of additions to the existing employer-based system: expanding Medicaid to include all those below a set income threshold, establishing state-based health insurance exchanges (later “marketplaces”) to facilitate the operation of the individual and small-group markets and to administer income-scaled federal subsidies, and broadening the regulation of private insurance to ban or constrain various underwriting practices. In addition to these three principal components, the reforms were studded with myriad provisions added as the price of securing the votes of individual legislators. Some of the multiple compromises were aimed at luring moderate Republicans; the rest were reached entirely among Democrats after those bipartisan efforts proved fruitless. Among the deals reached were the deferral of some provisions to come into effect over a lengthy 8-year implementation period. These various effective dates were fixed in the legislation—a strategy that would haunt the rollout of the reforms. Despite these various concessions, the legislation attracted no Republican support and triggered an ongoing series of efforts at repeal.
Although such mosaic strategies are less likely to be adopted in the supposedly “veto-free” (Hacker 2004) Westminster system of unified government and party discipline in Britain, they may occur as the result of political calculations. Strategies of scale and pace are determined not by political and institutional conditions themselves but by the ways in which political actors, individually and collectively, read those conditions to assess their prospects for success in the present and over time. In 2010, such calculations led the new Conservative/Liberal-Democrat Coalition toward a mosaic strategy for reform of the National Health Service (NHS). The partners in the Coalition saw the projection of a radical stance as a partisan imperative: they were eager to put a strong and positive face on what otherwise would be seen as a dispiriting election outcome for both parties, in which the Conservatives failed to gain a majority of seats and the Liberal Democrats fell well short of the expectations of only a few months prior. The formation of their unprecedented coalition allowed them to present their partnership as capitalizing on a historic opportunity to embark on new directions. Developing a common agenda, however, required negotiating myriad differences of opinion both between and within the parties. And it would need to be done rapidly: the electoral horizon was even shorter for the two parties in the Coalition than is normal for a Westminster majority government. The parties had agreed to a fixed date for the next election 5 years hence, but they knew they would have to distance themselves from each other increasingly as that deadline approached.
Throughout the process of developing a coalition agreement and a program for government, negotiators at the center of government agreed on a set of reforms to the English NHS. The reforms were essentially an amalgam of Conservative desires for increased competition and greater autonomy within the NHS and Liberal Democrat support for enhanced local authority. They would maintain the logic of the “purchaser/provider split” within the NHS (established two decades earlier by the Thatcher Conservative government and preserved and adapted by subsequent Labour governments) while reconfiguring the “purchasers” essentially to invert the formal relationship between physicians and managers. General practitioners, who had formerly advised local purchasing authorities through professional executive committees, would now form consortia of general practices (subsequently known as clinical commissioning groups) to take over the purchasing function, supported by managerial staff. In addition, numerous other changes were made to the structure and mandate of regulatory and support agencies.
As in the United States, the Coalition government launched the process of health care reform almost immediately and sought to do so with a single piece of legislation. Little over 2 months into its mandate, in July 2010 the government released a white paper fleshing out the reforms—a pace “far faster than any previous health white paper” (Timmins 2012: 15)—and quickly embarked on a period of consultation followed by the introduction of the Health and Social Care Act (HSCA) into the House of Commons in January 2011. However, passage would require numerous compromises during a protracted 14-month legislative process of extraordinary contentiousness, during which the government called an unprecedented “pause” for renewed consultations by a quasi-independent commission, which in turn led the government to support almost 400 substantive amendments (not including minor technical edits) and to adapt its announced approach to implementation (Timmins 2012: 121). The process was fraught with conflict throughout: implacable resistance from the opposition Labour Party was exacerbated by “rebellions” and threatened rebellions in the legislative ranks of the Liberal Democrats in the House of Commons and strong resistance from the more independent members of all parties in the House of Lords. The legislation passed on almost entirely party-line votes in each house, and immediately upon passage the Labour opposition, like the US Republicans, announced its intent to repeal the legislation “at the first opportunity” (Jowit 2012).
Assessing the Politics of Mosaic Reforms in a 10-Year Perspective
What can we learn about the politics of mosaic strategies of policy change by comparing these two cases, not only at the time of their enactment but also in the decade following? The principal observation concerns the self-undermining nature of mosaic strategies, in both process and substance. Both the politics of mosaic strategies and the resultant policy designs have features that leave the reforms open to attack and/or substantial modification in the postenactment period.
Politically, the fast pace of mosaic strategies exacerbates conflict. Enacting changes all at once compresses all veto points into a single episode and forfeits the possibility of dividing the reform process into a series of contests with particular interests while building broader coalitions of support. This phenomenon is especially apparent if we compare the politics of the British coalition reform process with those of the two periods that immediately preceded it: the “big-bang” introduction of the purchaser-provider split under the Thatcher Conservatives in 1990 and the incremental adaptation of that model by Labour from 2000 to 2009. Over that full time span the principles underlying the functional role of the state moved progressively from those of an owner-operator to those closer to a model of a single payer of independent providers. But these phases were marked by very different politics. The fast-paced big bang generated sharp conflict between parties and between the government and affected interests in the health care arena, whereas the incremental Labour phase was a “gradual, step by step process . . . providing little opportunity for a confrontation on the principles underlying the model that finally emerged” (Klein 2013: loc. 7143 of 9801).
Moreover, because they involve coalitions of actors with independent power bases, mosaic strategies typically generate a spectacle of sprawling consultations, inviting continual interventions and feeding public perceptions of incoherence. In addition to the multiple consultations held by the Democratic leadership in Congress and the administration with affected interests, the legislative process for the Affordable Care Act (ACA) comprised 79 hearings involving 181 witnesses in the House and “approximately 100 hearings, roundtables, walkthroughs and other meetings [as well as] 25 consecutive days [of debate] in continuous session” in the Senate (Jost 2017; see also Cannan 2013). Even in a Westminster system, the UK coalition government was forced into an extraordinarily extended and open legislative process, including 50 days of debate plus its midway “pause” for further public consultation.
In terms of policy design, the prospect of losing a position of advantage in the near future that spurs a rapid pace also requires reformers to specify implementation timelines in the original legislation rather than leaving hostages to the fortune of future administrative discretion. This imperative is further complicated when different timelines are attached to different discrete elements of the reforms, often as a result of ad hoc compromises in the process of passage. Both the ACA and the HSCA allowed for phased implementation—in the former case, extending beyond the life of the enacting Congress and the president's first term. In each case, this phased implementation was partly a matter of deliberate design—in the UK case, to allow for the commissioning consortia of general practices to be developed by the numerous local actors involved, and in the US case, to back-end load the costs of the program, given the fiscal constraints of the enactment period. In both cases, however, additional deferrals also resulted from compromises necessary to build the legislative coalition for passage, such as the effective dates for employer mandates, targeted taxes, and payment regimes in the United States, and for the establishment of new bodies and the phasing out of their predecessors in the United Kingdom.
The piecemeal policy design, ad hoc compromises, sprawling negotiations and staggered effective dates characteristic of mosaic reforms set the stage for self-undermining dynamics in the postenactment phase. These dynamics play out in three arenas: the legislature, the implementation process, and public opinion.
Both the ACA and the HSCA encountered vehement opposition from opposition parties that persisted long after the legislation was passed. The multiple attempts, both symbolic and real, of the US Republicans to repeal all or parts of the ACA in the decade following its passage are dealt with elsewhere in this issue. While the Westminster system does not allow for the sort of ongoing repeal attempts mounted by US Republicans, the UK Labour Party took various opportunities to signal its ongoing opposition. When the regulations under the act were published in April 2013, Labour sought to have them annulled by the House of Lords. The regulations easily survived the legislative hurdle, although the government had already tempered them in response to the challenge. In the general elections of 2015, 2017, and 2019, the Labour Party manifesto pledged to repeal the HSCA.
However, in neither the US nor the British case could the opposition accomplish its objective of full legislative repeal. The very point of the fast pace of the reforms was to hardwire them in such a way that they could be repealed only if the opposition could recreate for itself the very conditions that allowed for passage in the first place. In the United Kingdom, this would mean that Labour would have to form a majority government. In the United States, it would mean that the Republicans would not only have to control the White House and both houses of Congress but also be able to forge an internal consensus. Neither British Labour nor the US Republicans were able to clear these bars in the years following passage.
However, the design of the reforms offered other possibilities for opponents, short of outright repeal. Unable to come to an agreement on a legislative package to repeal and replace the ACA despite gaining control of the White House and both houses of Congress, the US Republicans were nonetheless able to defang its most symbolically contentious element, the individual mandate, by repealing the tax penalty for those who did not have health insurance, as part of a broad package of tax reform. The resilience of the ACA reforms was nonetheless demonstrated when, despite fears to the contrary, the individual mandate proved not to be the linchpin whose removal would bring down the entire package but rather to be, in practice, a severable element.2 British Labour has not yet found itself in a position of advantage similar to the US Republicans. But despite continuing to vow to repeal the HSCA, by the time of the December 2019 election, Labour had narrowed its focus to two symbolically significant elements with modest material effect, pledging specifically to “reinstate the responsibilities of the Secretary of State [for Health] to provide a comprehensive and universal healthcare system” and to “end the requirement on health authorities to put services out to competitive tender” (Labour Party 2019).
The Implementation Process
The piecemeal nature of mosaic reforms can be both a strength and a liability, offering opportunities for various political actors to either chip away at or adapt them in the implementation process. Consider first the opportunities for opponents. The multiple disjointed elements of mosaic reforms are likely to vary in their vulnerability to legal challenge and administrative discretion in implementation. They therefore provide a number of footholds for those who would seek to overturn or modify the reforms even without building the coalition that would be necessary for broad legislative change. These opportunities are multiplied to the extent that political institutions allow for “venue shifting” as quintessentially afforded by the checks and balances and federal division of powers in the American political system, in contrast to the more unified British Westminster model. Republican opponents of the ACA were able to use the courts and state governments to frustrate or constrain some of its aspects, notably the expansion of Medicaid. State refusals to establish health insurance exchanges in all or part also resulted in a greater federal role in that aspect of the reforms by default than had been envisaged at the outset.
In the United Kingdom, the piecemeal nature of the HSCA reforms allowed them to be recombined in implementation without legislative change, to an effect quite different than what the foundational model of the “purchaser/provider split” would imply. A new chief executive of NHS England, appointed in 2014, moved progressively to reintegrate activities that had been separated in the name of increased autonomy and choice. New models of integrated care that would blur the distinction between purchasing and provision functions have been promoted, and new structures and processes that effectively reintegrate executive functions at both central and regional levels have been put in place—all through administrative action.
A final vulnerability of mosaic reform strategies stems from their characteristic staggering of the effective dates of various elements. In theory an attenuated implementation timetable can allow for the development of technological and other capacity before reforms come onstream. But when implementation dates are specified in the legislation up front, implementers are tied to a fixed timeline whether or not capacity development has proceeded as anticipated, thus providing opportunities for opponents to seize upon and exacerbate administrative difficulties. Because of their nature as packages of elements that are more the result of hasty compromises than of policy design, mosaic strategies are particularly vulnerable to such problems. This was particularly apparent in implementing the ACA—most notoriously in the case of the problem-ridden launch of the federal enrolment platform, HeathCare.gov, in October 2013, but also when the effective dates for a variety of provisions arrived before those at whom they were targeted (e.g., employers and insurers) had developed the capacity to comply. Unable to adapt the legislation itself in the polarized partisan Congress, the Obama administration controversially took administrative action on more than 20 occasions between January 2013 and March 2015 to provide “significant” delays, extensions, exemptions, provisions for retroactive payments, and other deviations from the strict provisions of the ACA in order to smooth its implementation (Redhead and Kinzer 2015).
Tactical choices can nonetheless mitigate such problems. Faced with deferred effective dates, the British Coalition government put in place a number of the new entities in “shadow” form to operate de facto through various forms of delegation, to ready them in advance of the dates at which the legislation would provide them with de jure authority. Conversely, some legacy bodies were retained longer than anticipated, as administrative support while the new or transforming entities were developing capacity.
Because mosaic reforms touch multiple parts of the system yet lack a coherent overall frame, they are vulnerable to being portrayed as both chaotic and overweening, generating and perpetuating public confusion and hostility. In the case of both the ACA and the HSCA, public reaction to the legislation divided along partisan lines. In the United Kingdom sheer confusion dominated that partisan effect, whereas in the United States voters appeared to take their cues from their parties in forming their views. Table 1 shows opinion in each of the two countries 1 month before passage of the legislation. In the United Kingdom, partisan opposition was most sharply apparent among Labour voters, whereas supporters of the two parties in the governing coalition, and especially the Liberal Democrats, were more divided and much more likely than Labour voters to register a “don't know” response. In part this may be because neither the Conservatives nor the Liberal Democrats had given any signal in their election manifestos that they would enter into coalition, much less embark on major reform to the NHS, whereas American Democratic voters had been in no doubt that health care reform would figure prominently on the Democratic agenda. Notwithstanding their differences regarding the merits of the legislation itself, however, Democrats, Republicans, and Independents in the United States were alike in their propensity to see the politics of its passage as less about “fundamental disagreements on what would be the right policy for the country” than about “both sides playing politics with the issue” (Kaiser Family Foundation 2010).
What is more interesting from a decade-long perspective is the extent to which these attitudes persisted. In part, this persistence may have been fed by difficulties and contention in rolling out the reforms as discussed above, reinforcing the perception of incoherence. In the United States, the persistence of the partisan divide in public opinion has been well documented, including elsewhere in this issue. In the United Kingdom, there is evidence of a similar phenomenon. The sharp decline in satisfaction with “the way in which the NHS runs nowadays” that coincided with the parliamentary debates over the reforms in 2011 reversed a 10-year trajectory of increasing satisfaction and marked the onset of a pattern of fluctuation around a general declining trend (Robertson et al. 2019: 6). Partisan differences continued: in 2018, Conservatives were more likely to be satisfied; Labour supporters, less so. Evidence is mixed as to the impact of the reforms themselves on public satisfaction, at a time when the coalition and subsequent Conservative governments were also constraining rates of annual increases in NHS funding after a decade of funding increases under Labour. Initially the new austerity appeared not to trigger strong opposition: a majority of respondents in the 2014 British Social Attitudes survey supported the current level of spending, with only about a third arguing that the government should “spend more” (Curtice and Ormston 2015). Later, however, respondents to the survey from 2015 to 2018 were consistently more likely to cite insufficient government spending than to cite “government reforms” as a cause for dissatisfaction (at about 40–50% and 20–30%, respectively) (Robertson et al. 2019: 15).
A Within-Country Comparison
The ACA is not the only case of a mosaic strategy in the history of American health care reform. Given the multiple veto points inherent in the US political system, it is perhaps not surprising that the very founding of the modern US health care state—the Medicare/Medicaid legislation of 1965—required a mosaic strategy, producing the original “three-layer cake” of additions to the employer-based model as disparate proposals for physician and hospital services for the elderly and health care services for low-income families were amalgamated to attract a winning coalition. This process of negotiation yielded “a model of unintended consequences [that] . . . incorporated features that no one had fully foreseen” (Marmor 2000: 58–59).
The parallels and differences between Medicare/Medicaid and the ACA suggest some qualification of our propositions about the dynamics of mosaic reforms. Each reform package was born in open legislative conflict as well as behind-the-sciences negotiations. The ACA process brokered factionalism within the Democratic Party to yield a purely Democratic result against unanimous Republican opposition; the Medicare/Medicaid process deeply divided the Democratic Party, even as it attracted some Republican support. These conflicts played out in public view as various bills ran the legislative gauntlet (Marmor 2000: 46–61). Moreover, although the Medicare/Medicaid package was a relatively more elegant result than the rather jumbled ACA, even it was a pastiche of competing proposals.
In both cases, reformers focused on what was at the time a relatively small segment of the private insurance market and on the medically indigent. Medicare and Medicaid targeted the elderly and social assistance recipients, respectively. The ACA was aimed at covering those uninsured or underinsured under either public programs or private insurance. In each case the degree of institutional change within these segments was significant. Although Medicare/Medicaid built on the established Social Security model, the extension of that model to health care represented a significant increase in the weight and a change in the role of the state. Similarly, by expanding Medicaid the ACA built on an established program but substantially increased its redistributive effect, and the establishment of health care marketplaces introduced a significant innovation in the regulatory role of the state.
The scope of the two sets of reforms varied, however, in ways that had important implications for their future dynamics. While Medicare imposed an obligation on most employers and workers to contribute to Part A of the program, those contributions built on the well-accepted Social Security program. And while Medicare and Medicaid potentially affected all physicians and hospitals, the two programs essentially added another (public) payer to the mix of private payers to which those providers were accustomed, and those relationships with other payers were left alone. No grand accommodation, such as that binding the medical profession and the state into exclusive or near-exclusive relationships in the founding models of the health care state in Canada and the United Kingdom, was brought about. Meanwhile, employer-based coverage for the bulk of the population, as well as the norms and commercial reach of the private insurance industry, were left intact.3
In contrast to this delimited mosaic, the ACA made small adjustments to the rules of the game that touched almost all actors in the health arena. In addition to the changes to Medicaid and to the individual and small-group insurance market, the reforms addressed underwriting practices and profit margins of private insurers more broadly, in addition to many other provisions added to the legislation as the price of passage. The changes thus affected, at the margins, not only all providers of covered services but also all insurers and many employers. (In this respect the ACA reforms were more similar to the HSCA reforms in England. Although the latter effectively accelerated changes already under way under the previous Labour government, the organizational changes simultaneously affected all purchasers and providers of NHS services. In both the ACA and the HSCA cases, the pervasiveness of the changes conveyed the impression of a gargantuan reform, while the complexity and incoherence of the package made it difficult for proponents to build support by communicating its overall purpose and intended benefits.)
The postpassage politics of the two American sets of reforms were starkly different. There is no parallel in the Medicare/Medicaid case to the unrelenting opposition of Republicans to the ACA and their ongoing attempts at repeal and constitutional challenge (Patashnik and Oberlander 2018: 663). Nor did public support decline over the course of the Medicare/Medicaid legislative debate and thereafter, in further contrast to experience with the ACA (Blendon and Benson 2001: 35–36). Medicare and Medicaid have not only persisted but also expanded on balance over time, in part because of demographic change, but also as a matter of deliberate policy choices.
The contrasts between these two American cases suggest some further nuance to the propositions offered here about the postenactment dynamics of mosaic reforms. Mosaics largely confined to a segment of the health care arena may prove more durable than those whose effects are more widespread. Conversely, multiple small changes that are spread across the arena exacerbate the problem of conveying the coherence and point of the reforms and provide multiple footholds for opposition. In short, the scope of change may be at least as important for postenactment politics as is its degree. Our few cases are not sufficient to explore this hypothesis further, but they raise questions for further research.
A final caveat: any comparison of strategies of scale and pace in policy change, across or within nations, needs to situate those cases in their historical-institutional context more fully than space allows here (Tuohy 2018). In the American context, comparisons between the 1960s and 2010s are complicated not only by increasing partisan polarization over that period but also by the fact that Medicare and Medicaid have become part of the historical-institutional context in which the ACA played out (Patashnik and Oberlander 2018). Indeed, it can be argued that the feedback effects of those two programs have been to reinforce the structure of the programs themselves while making it more politically difficult to adopt broader-based designs. Nonetheless, similarities in political dynamics of mosaic strategic across historical-institutional contexts as different as the United States and the United Kingdom in the early twenty-first century suggest that choices of scale and pace have significant effects at the margin.
A mosaic strategy may well be a reasonable choice in political circumstances such as those confronting the framers of the ACA. Other political actors facing analogous circumstances would nonetheless be well advised to be alert to the potentially self-undermining properties of such strategies and to consider how they can be mitigated.
The severability of the universal mandate as a matter of law nonetheless continued to be litigated.
There is also a significant tactical difference between the Johnson and Obama mosaics. Perhaps owing to Johnson's famous aversion to any delay that would allow opposition to build (Blumenthal and Morone 2010: 190), legislative compromises did not include the deferral of effective dates of various provisions. With Johnson's personal involvement, implementation proceeded rapidly (Blumenthal and Morone 2010: 198; Ball and Hess 2001: 8–9), in contrast to the staggered ACA implementation discussed above.