1. Introduction
The European Union (EU) is an ambiguous health policy actor. It is powerful, but not enabled to act in the name of health. Its leaders are often eager to claim that it improves health in Europe, but it is also under powerful pressures to promote other agendas that might not improve health, such as budgetary surveillance or deregulation. It is implicated in the changing politics of left and right, populist and liberal, of its member states, but how its health policies do and do not figure in those politics is often unclear.
This paper analyzes the EU's health politics in the context of comparative federalism. It first discusses the EU's distinctive political form and dynamics, which shape its health policy around three faces of health, markets and fiscal governance. It then presents some results of a large study of social policy in the decentralized states of the Organisation for Economic Co-operation and Development (OECD), identifying the shape and nature of their federal welfare states. The third part compares them. It finds that the EU is quite unlike all the other federations, with far more central regulation of matters such as professional qualifications that in the federations are left to lower tiers and far less risk pooling and redistribution such as the finance of health care. The conclusion suggests that even while the EU has done much for health outside its health treaty bases, a structure in which the EU creates and regulates without compensating individuals who lose out has serious imbalances. That no other federation has combined such a level of regulation, including of health care access and administration, with such a lack of redistribution within units or among people suggests the EU's unique approach might not not be stable.
The case for comparing the EU to federations, rather than to other international organizations, has been made more extensively elsewhere as a general approach (e.g. Fabbrini, Reference Fabbrini2004; Obinger et al., Reference Obinger, Leibfried and Castles2005; Menon and Schain, Reference Menon and Schain2006; Kelemen, Reference Kelemen2017) and in the specific area of EU health policy (Fierlbeck and Palley, Reference Fierlbeck, Palley, Fierlbeck and Palley2015; Vollaard et al., Reference Vollaard, Van de Bovenkamp and Martinsen2015). Part of the reason is that international organizations, as a rule, are far less consequential than the EU. Member states can disobey rulings of United Nations bodies or regional agreements far more easily than they can disobey EU law. Part of the reason is that the most classic EU theories from international relations, such as neofunctionalism and intergovernmentalism, are theories of integration rather than of policymaking (Rosamond, Reference Rosamond2000); they are about the development of EU powers and competencies rather than its current policies and are therefore relatively limited as explanations of what the EU does. More positively, many of the key attributes of federal politics can be found in the EU: commonalities “concern the division of rule between the central and the subunits' levels of governance, the autonomy of each level of governance in at least some tasks in their respective territories, formal arrangements to change the division of rule, and the representation of the sub-units at the central level” (Vollaard et al., Reference Vollaard, Van de Bovenkamp and Martinsen2015: 160). This paper, without engaging in bigger comparative federalism debates (for which, Bednar, Reference Bednar2011), will draw on findings from comparative federalism research on these topics, above all morphological comparisons: insofar as the EU is a federation, it is a very unusual one.
2. A neofunctionalist regulatory polity
The EU is, by consensus, a regulatory state (Majone, Reference Majone1996). It is characterized by its reliance on one key policy tool, which is regulation. It is weak in other actions of the modern state, whether that is taxation, coercion, redistribution or operation of a welfare state. Its regulatory powers and efforts, reflecting that one-way bet, are more developed than the regulatory powers and efforts of most federations, reaching deeper into issues such as workforce credentialing than comparable federations. It also explains how the EU can be so influential with so few staff and such small resources (Page, Reference Page and Menon2001). The EU, structurally, uses member state policy capacities to develop policies implementing its directives and member state bureaucracies to enforce regulations. For example, member state labor inspectorates inspect compliance with working time rules, whatever their origin in EU law or domestic politics. It also can rely on decentralized enforcement via member state courts, in which litigants can call on member state courts to enforce, and develop, EU law (Obermaier, Reference Obermaier2009; Kelemen, Reference Kelemen2011).
It is not just any kind of a regulatory state, of course. It has a pronounced constitutional asymmetry toward market-making measures, notably the removal of barriers to cross-border flows of goods, services, capital and people (Leibfried and Pierson, Reference Leibfried and Pierson1995; Scharpf, Reference Scharpf2002). Broadly, Scharpf's argument runs, the EU is better at striking down national regulations than at expanding regulatory protections and market-compensating programs. Positive EU actions, such as higher regulatory floors, must pass through the complex and veto-filled EU legislative system, which is slow and often waters down protections, and actual redistribution, outside the structural funds and similar capital schemes, is effectively nonexistent. By contrast, the EU's regulatory approach (with member states, rather than people or firms, being the targets of regulation), as enforced by courts, means that it is relatively easy to strike down member state regulations that might have fulfilled a useful social role, while increasing European integration can create downward pressures on, for example, social benefits or jobs. Member states are left with the tasks of compensating those who lose from deregulation and market creation, through welfare programs that include health systems. In this reading, the EU is constitutionally fated not just to be a regulatory state, but a market-promoting one. Even if member states are increasingly effective at reining in EU law (Martinsen, Reference Martinsen2015; Pavone and Kelemen, Reference Pavone and Kelemen2019), that might slow market-making integration but does not change the basic asymmetry.
The EU has developed in this way through the interplay of neofunctional integration dynamics and its regulatory state form. Neofunctionalism is a theory of European integration, not politics or policymaking or Europeanization. It explains how the EU has gained the powers that it has, but not the politics of what is done with those powers. It is often and inexcusably misunderstood to mean some sort of agency-free process in which the EU inexorably gains powers and loyalty as European social and economic integration progresses, but that is not how it works. Rather, social and economic integration turns into European integration via the agency of EU institutions and policy entrepreneurs who put things onto the European agenda and thereby start to create a EU competency. Its leading progenitor was Ernst Haas (Reference Haas2004), who argued that:
the central institution will affect political integration meaningfully only if it is willing to follow policies giving rise to expectations and demands for more or fewer federal measures. In either case, the groups concerned will organize across State boundaries in order to be able to influence policy (Haas, Reference Haas2004 [1958]: xiii).
Note that this quote puts the agency on the ‘central institution,’ namely the EU institutions such as the Court and Commission. They can trigger integration, including by spearheading actions that prompt the construction of European coalitions against them, as happened when the Court started to engage in health law (Brooks, Reference Brooks2012). The history of much of EU health care is that of EU health care law and its development by EU and member state courts (Obermaier, Reference Obermaier2009; Hervey, Reference Hervey2016).
Social integration producing demand for European integration also happens and includes the repeated ‘crises’ that loom so large in a history of EU public health policy. In public health, shared European food safety crises such as vCJD show how cross border trade outpaces regulation and produce a demand for European regulation to match European integration, which is why the EU has a specific treaty competency to regulate blood and blood products. Haas later repudiated neofunctionalism, but it has since seen a revival with some modifications to its scope conditions by scholars who saw, in particular, the ways in which EU political institutions could slowly enter an area and start its integration (Niemann and Schmitter, Reference Niemann, Schmitter, Weiner and Diez2009; Lefkofridi and Schmitter, Reference Lefkofridi and Schmitter2014). The implication of neofunctionalism is that the EU can expand slowly into new areas as entrepreneurial institutions such as the Commission – or entrepreneurs within institutions – open up space for EU action and start to create a European politics of a given policy area.
This neofunctionalist process works regardless of the substantive importance of the issue (Kleinman, Reference Kleinman2002: 121). Cross-border mobility within the EU is a good example. It is not a major issue for most countries and is an even less important practical issue in concrete manifestations that catalyze EU activity. The issue of cross-border patient mobility involves the very specific cases of patients who pay out of pocket for care in another member state that was not pre-authorized and then seek reimbursement. Very few Europeans actually seem to want cross-border health care at all, fewer still shop around, and even fewer want to shop on the much less preferential terms under the Directive rather than through preauthorization (Glinos, Reference Glinos2012). Europeans resident in another member state, even multilingual ones in border regions, will try to go home for care, inflating statistics on border-crossing without actually shopping across borders (Glinos et al., Reference Glinos, Doering and Maarse2012). The latest estimate from the European Commission, for 2016, is that 0.1% of EU health care expenditure is cross-border, including pre-authorized care under social security coordination procedures that effectively date to 1971 and are separate from the patient mobility issues that emerged in 1998 (European Commission, 2018). That makes it a trivial policy issue for health systems rather than the threat to necessary closure that many expected (Ferrera, Reference Ferrera2005). It was nonetheless the basis for the development of much of EU health care jurisprudence, a Directive, and a series of efforts to develop an EU framework for health care with different objectives and from different political quarters. The reason was not the scale of the policy problem; the reason was the legal and political disruption caused by the prospect of regulating health care through the EU internal market (Greer and Rauscher, Reference Greer and Rauscher2011).
3. The three faces of EU health policy
EU health policies stem from this basic morphology. They are regulatory, and secondarily built around grants and ‘soft’ harmonization mechanisms, they derive from its powers over the market above all, and they do not resemble the policy tools or priorities of a member state.
3.1 EU health policies
EU health policies produced in such a system have three faces (Greer, Reference Greer2014). The first is the overt public health face, authorized by Article 168 TFEU, headquartered in the DG responsible for health (DG SANTE), and focused on meliorist public health initiatives that mostly use the EU's non-regulatory soft powers: production and comparison of data, capacity-strengthening initiatives, research and an agency to coordinate public health efforts (the European Centre for Disease Control and Prevention) (Guigner, Reference Guigner, Elgström and Smith2006, Reference Guigner, Baisnee and Pasquier2007; Greer, Reference Greer2012; Deruelle, Reference Deruelle2016; de Ruijter, Reference de Ruijter2019).
This face is by no means the most important form of EU engagement in health. It can contribute ideas, networks and some regulation, but the health treaty base (Article 168) is deliberately weak and the budget is a fraction of the running costs of any member state health care system. The entire EU budget, excluding European Investment Bank (EIB) loans, is capped under 1% of EU GNI (0.84% in 2018); member states spend between 5% (Romania) and 11.5% (France) of gross domestic product (GDP) on health (Eurostat, 2019). The EU does not and is unlikely to ever get the money to be a major part of health care financing. Even if we include structural funds and EIB lending, which unlike research or agriculture spending are intended to compensate losers, we do not approach the sums of money that running any health system requires, and both structural funds and EIB loans are allocated via processes not specifically focused on health and subject to pulls from other directions (side-payments to member states, for structural funds, and both caution and alternative missions such as climate change for the EIB).
Further, much of the policy that looks most like public health is actually done under other treaty bases – especially internal market ones. Article 168 is weak by design, with language that clearly authorizes only consensual and supplementary EU policy and specifies that EU policy “shall respect the responsibilities of the Member States for the definition of their health policy and for the organisation and delivery of health services and medical care” (Art. 168.7 TFEU). Thus, many policies, including ones that seem highly relevant to health policy or have clear health benefits, are not legally ‘public health policies.’ The General Food Law Regulation (178/2002) is very clear that its main treaty bases are not in Article 168, and medical devices, medicines regulation and professional qualifications regulation are built on internal market law. Environmental policy, which has probably had enormous health benefits, is still more legally and bureaucratically separated. Much of what promotes public health is not done by formally identified public health actors on public health legal bases and viewing it as public health policy might misrepresent the actual policies and politics.
3.2 EU market policies and health
The second, and much more consequential, face of the EU in health is single internal market legislation and jurisprudence. This face is where the EU has had the most impact on health systems, for it is where the EU has the most powers and the strongest legal and enforcement capabilities. The only EU legislation really about health care, Directive 2011/24/EU on patients' rights in cross-border health care, stemmed from the Court applying law on the freedom to provide services to decide that countries could not confine the purchase of publicly financed health care services to their domestic providers. The Court thereby broke the legal closure of national health care systems (which had previously required pre-authorization for non-emergency care) (Ferrera, Reference Ferrera2005), precipitating a political argument in Brussels, and EU legislation where there had been none. In other words, the energy of the central institutions – EU courts – using internal market principles led to European integration in a new policy area despite near-total opposition from interested groups such as member states and providers.
This story is particularly clear, but its outlines are found throughout the history of European integration: once something is treated as part of the market, it is easy to integrate into the EU, but in the form of market-making activity. This typically takes the form of deregulation (elimination of member state policies on the grounds that they discriminate and thereby interfere with the internal market) coupled with a level of EU re-regulation (to ensure minimum standards). It helps to explain, for example, the energy put into the European regulation and promotion of cross-border mobility among health care professionals (Kuhlmann et al., Reference Kuhlmann, Maier, Dussault, Larsen, Pavolini, Ungureanu, Hervey, Young and Bishop2017). The EU's internal market law treaty bases allow it to promote mobility, attack discrimination on grounds of national origin and set a regulatory framework for credentials and their recognition, which makes those the preferred EU policy tools in the area.
One of the most important areas of this second face is in medicines, and secondarily medical devices. In these areas, the EU is a functional, if often very strange, single market (pharmaceutical pricing and trade in the EU is not well understood since much of the data about pricing are kept opaque in the interests of sellers, middlemen and states). The European Medicines Agency, latterly of London and now located in Amsterdam, is a globally credible central regulator and the hub of a network of member state regulators (Hauray, Reference Hauray, Greer and Kurzer2013). European medical devices regulation is much less stringent, though there is pressure to change (see Jarman and Rozenblum, this issue). Regulation in these areas was part of the 1992 single market program, and the treaty bases for both are in the promotion of the single market. Given that pharmaceuticals and devices loom large among the EU's most innovative and high-value-added industries, these policy areas are important and viewed by policymakers as being industrial policy as well as public health policies. Other policies that are often more commonly viewed as part of other policy areas, such as research or data protection but which have health care dimensions include governance of clinical trials, data governance and life sciences research. In many of these policy areas, e.g. research, there is a tension that will be familiar to health researchers between those who see promotion of the sector as industrial policy and those who are more concerned about other, less commercial, goals [for discussion of the policies in this paragraph, see Part II of Hervey et al. (Reference Hervey, Young and Bishop2017)].
3.3 EU fiscal policies and health
The third face of EU health policy is, like the market provisions, powerful and not driven by actors or priorities from the health sector. It is fiscal governance and it is an outgrowth of decades of efforts to create and maintain a common currency for Europe. Essentially, the logic of fiscal governance (which is the same as within federations) is that monetary unification creates incentives for member states to free-ride by issuing too much debt, taking advantage of the low interest rates created by a shared currency and exploiting the soft budget constraints created by markets' conviction that the politics will always lead to a bail-out for overextended governments. To avoid such free-riding, the EU always had a set of rules about, in particular, deficits and debt as well as, for aspirant members, currency stability and inflation. These were formalized, with the move to currency union, in the stability and growth pact (SGP). Notably, the SGP demanded that member states have debt-to-GDP ratios below 60% and deficit to GDP ratios below 3%.
This set of requirements had their maximal leverage immediately before the Euro accession, when member states had to demonstrate, if not compliance, at least a credible trend toward compliance in order to join this important political project. The push to qualify for the Euro led to a host of interesting experiments in coordination and reform in different countries (Parsons and Pochet, Reference Parsons, Pochet and Dyson2010; Hancké, Reference Hancké2013). Most member states had some manipulation in their figures (St Louis, Reference St Louis2011) but the push to reform was strong, universal and led to a degree of policy change in most of the EU.
The Euro is “a real federal currency with underdeveloped institutions of governance” that is “devoid of functional crisis-resolution mechanisms” (Underhill, Reference Underhill, Payne and Phillips2014: 154, 164). The structural imbalances between surplus and deficit countries that it created were ignored by the architects of the Eurozone. Scholars seeking to explain the crises of the Eurozone since 2008 point to two key factors. One is the influx of speculative money into peripheral countries, exploiting newly low interest rates by blowing up speculative bubbles (Pérez, Reference Pérez2019). The other is the superior competitiveness of the coordinated labor market economies (Germany and its satellites such as the Netherlands, Belgium and Austria). The most convincing version of this argument is that the coordinated economies could contain wage growth and prevent sectoral distortions such as suffered by countries with less coordinated labor markets, where the private tradable sector was seriously damaged when public and service sector inflation was inflated by the influx of speculative money from the creditor countries (Hancké, Reference Hancké2013; Johnston, Reference Johnston2016).
The upshot of the economic crisis is that EU policymakers, led by Germany and other creditor countries (e.g. the Netherlands), sought to tighten the rules, in part in order to credibly argue that bailouts of lenders to Greece, Ireland, Portugal and Cyprus did not create soft budget constraints for the future (Dyson, Reference Dyson2014; Bastasin, Reference Bastasin2015). The result was legislation (the six pack and the two pack, in Brussels argot) and a new intergovernmental Treaty on Stability, Coordination and Governance (TSCG) entrenching the SGP in domestic constitutional law and creating an elaborate mechanism of policy surveillance called the European Semester whose stated goal is to keep member states from adopting budgets that threaten SGP compliance or macroeconomic stability (Schelkle, Reference Schelkle, Zahariadis and Buonanno2018).
Health care is very expensive so the Semester almost immediately took an interest in the topic (Azzopardi-Muscat et al., Reference Azzopardi-Muscat, Clemens, Stoner and Brand2015; Stamati and Baeten, Reference Stamati and Baeten2015). There were no health-related objectives involved, and the Semester was initially run by finance ministries and the Economic and Financial Affairs DG and Council of the EU. Initially, the Semester recommendations, reflecting these origins and this organization, were about constraining health expenditures, and some of the recommendations seemed to show poor understanding of health economics and policy.
For example, the 2015 Country Specific Recommendations (CSR) for France suggested, with a threat of penalty given that France was in the Excessive Deficit Procedure, that part of the solution would be a review of admission procedures for health professional education (Council Recommendation 2015/C272/14). Increasing supply of labor in most cases does lead to lower prices, but in the case of health care the effect is often wiped out by supply-induced demand (Dranove, Reference Dranove1988). Doctors, in particular, are very good at exploiting information asymmetries to make work for themselves (a version of ‘Roemer's law’: A built bed is a filled bed) (Shain and Roemer, Reference Shain and Roemer1959). Supply-induced demand suggests that French restrictions on professional education might be a cost-containment measure rather than the cost-increasing measure that the 2015 CSR suggested.
Over time the Semester, like previous exercises in fiscal governance, began to lose its focus on budgetary control. Internally, its goals were expanded, its data contested and scope of conflict expanded (e.g. to include health ministries) (Scott and Eleanor, Reference Scott and EleanorForthcoming). Broadly, it also became the vehicle through which ‘social’ lobbies at the EU level sought to promote a social Europe agenda that had previously been pursued in venues such as the Open Method of Coordination (Vanhercke, Reference Vanhercke, Vanhercke and Natali2013; Zeitlin and Vanhercke, Reference Zeitlin and Vanhercke2018). As the EU, associated with austerity, sought legitimacy by enunciating a broader social agenda, the Semester started to acquire positive social policy objectives and some researchers even began to view it with optimism as a tool for improving social policy in Europe (Vanhercke et al., Reference Vanhercke, Zeitlin and Zwinkels2015).
The actual implementation and effect of CSRs is difficult to measure in any subtle way, and the credibility of the entire fiscal governance system could not be tested until the economic downturn of 2020. In 2020, one of the first EU actions was to activate the "general escape clause" and suspend the SGP's fiscal rules on the understandable grounds that the COVID-19 pandemic was a major emergency (https://www.consilium.europa.eu/en/press/press-releases/2020/03/23/statement-of-eu-ministers-of-finance-on-the-stability-and-growth-pact-in-light-of-the-covid-19-crisis/). If, as seems likely, the Semester has been increasingly instrumentalized by social interests to promote goals such as better and more equitable health care, then it is likely to be downgraded or discarded by its original advocates, who were primarily interested in austerity.
Regardless of whether fiscal governance, in the form of the Semester and the new legislation, is capable of either controlling government expenditure or improving European health policies, it is certainly an integrating move. Article 168 makes it clear that the EU shall have no powers over the organization and finance of health care, but just as health was regulated under internal market law, health policies are now subject to surveillance and potential discipline by the EU. The third face of EU action is very ambitious.
3.4 A democratic politics with three faces
The caveat to this analysis is that the EU is also democratically responsive to a continent where political parties of the right have often dominated politics and where there was a substantial shared commitment to a basic neoliberal program since the early 1990s. Neoliberalism is justifiably regarded as a problematic term, often used as an insult, with unclear shared content, but there is still clearly a core to it that scholars have identified: a faith in markets shared with classical liberalism, an esteem for the expertise of academic economics, and specific policy commitments to a limited role for the state in employment law, state-owned enterprises and welfare provision (Ban, Reference Ban2016; Offer and Söderberg, Reference Offer and Söderberg2016; Slobodian, Reference Slobodian2018; Vail, Reference Vail2018). Neoliberalism could come in forms preferred by right governments, which found it took little adaptation to endorse an agenda of deregulation and a smaller state. But, it also came in a sort of social democratic form often seen in the work of the OECD. This latter version found space for redistribution through the tax system, and often included private sector provision of welfare, arguing that liberalization of the economy (and the end of dualism in some cases) would produce the employment and revenue needed to build a welfare state suitable for changed times. Perhaps the epitome of that approach was the British New Labour approach of encouraging the London financial sector and using the considerable tax revenue it spun off to expand a welfare state, and good health service, for the rest of the UK as part of clearly progressive fiscal policy (Hopkin and Wincott, Reference Hopkin and Wincott2006; Belfield et al., Reference Belfield, Blundell, Cribb, Hood and Joyce2017). Regardless of one's explanation for neoliberlism's rise to influence, it is a way of viewing the world that was common across European policymakers for decades.
In other words, we should not use a purely structural, deterministic analysis of the EU. It has developed over decades in a context of shared elite agreement on neoliberalism, and, since around 2010, in a context of left party weakness that means most governments have been on the right. The EU treaties evolved in ways compatible with shared and mostly global government ideologies and commitments, and so did legislation. The image of the economy and economic management in the oldest versions of the EU treaties are quite unlike the market-focused EU of today.
4. The EU and health in comparative perspective
One basic way to learn about the dynamics of a system is to compare it to another. To what is the EU comparable? Not the member states in general; it is more federal, more complex, newer and much larger than most EU member states, and as noted above it does different things (Dehousse, Reference Dehousse1994). The EU is, however, more comparable to other federations: in the tensions between units, in its complex horizontal structure and vertical structures; in its veto points; and its history of bargaining over the roles of different tiers of government.
How does the EU, with its three faces, compare to the actions of other federations in health care? Quite differently. This section reports the conclusions of a multi-author comparative project on the interaction of federalism and social policy in the decentralized OECD member states: Australia, Austria, Belgium, Canada, Germany, Italy, Mexico, Spain, Switzerland, the United Kingdom and the United States (Scott and Heather, Reference Scott and Heather2019). The project brought together contributors responsible for each country. It started by trying to address limitations of much of the existing comparative literature on federalism and social policy, which is at best rigorously comparative but legalistic (Hooghe et al., Reference Hooghe, Marks, Schakel, Osterkatz, Niedzwiecki and Shair-Rosenfield2016) and is more commonly impressionistic, leaving readers without a clear sense of how federal institutions, politics and programs actually interact (as in Loughlin et al., Reference Loughlin, Hendriks and Lidström2011). Contributors described the breakdown of financing, legislative and delivery roles in health, primary and secondary education, tertiary education, 18–65 benefits and services and pensions. Two of the questions they asked are particularly relevant to a study of the EU as a federation in health: to what extent are distinctive regional welfare states found in places such as Quebec or Scotland; and to what extent do the structures of federal welfare states vary?
The answer to the first question is that variation in the programatic structures and priorities of regional welfare states is limited compared to the variation between countries. Variation is mostly in minor differences in covered services, resource levels and occasionally organization (e.g. primary care). Scotland runs an NHS program, Quebec a Canadian Medicare program and the Basque Country and Catalonia have regional versions of the Spanish NHS model. And these are among the most politically distinctive regions with some of the strongest fiscal autonomy and strongest political incentives to differentiate their policies from the rest of their states! The constraints on them include financing systems, policy legacies, statewide political parties and media, technical networks of experts who believe in ‘good practice,’ and interjurisdictional competition conducted within an existing set of rules. Thus, we can find identifiable differences, but the striking thing is how few of them there are – Scotland and Wales are less inclined to pluralistic health care provision than England, Quebec has a separate pension plan under federal rules and used it for a widespread industrial policy, the Basque Country has a universal basic income scheme – but in most cases the programatic design, coverage and priorities of their welfare states mirror those of the rest of their country. The Basques might enjoy the most distinctive welfare state in Spain, and the best-funded, but it is still a Spanish welfare state that they enjoy.
The implications for the EU of this finding are somewhat speculative due to timing. There are welfare states built after the basic federal institutions were created, in places such as Canada and the United States. In those places many of the key features of policy, such as heavy use of conditional grants, clearly reflect federal politics even if they then went onto shape federal politics. There are welfare states created before decentralization, which means politicians are separating linked or even unified systems, which is the case in the UK and Belgium. There are also welfare states that grew up together with federations, whether over a long time (as in Switzerland, Germany, which was federal in different ways except for during the National Socialist regime), or a shorter time (as in the simultaneous construction of the welfare state and decentralized state in Spain in the 1980s and 1990s). What we could not find outside the EU was a federation engaged in uniting countries with established welfare states. A reluctance to share benefits between countries is also understandable given the enormous economic as well as social and political disparities between EU member states: for example, per capita income in Luxembourg is almost exactly three times that of Bulgaria (Blanchet et al., Reference Blanchet, Chancel and Gethin2019) and healthy male life expectancy at birth ranged from 73.6 in Sweden to 51.5 in Latvia in 2014, according to European Core Health Indicators (ECHI) data. In such a context, leveling up would involve a very large flow of resources from the richer to the poorer member states, and granting the best available health care to all EU citizens would be far beyond the fiscal capacity of some member states.
The second finding of the project is perhaps more powerful. That is that there is limited scope for variation in the shape of a federation. The economics of fiscal federalism is a field that is both normative and descriptive, and its core policy recommendation is that sustainable programs put the responsibility for raising revenue and pooling risk at the highest possible level (the largest possible risk pool) and the responsibility for delivery at the most local level possible, e.g. the local area for primary care delivery and the region for the organization of advanced hospital services (Oates, Reference Oates1999; Boadway and Shah, Reference Boadway and Shah2009). Put another way, various arguments about risk pooling, local responsiveness and the benefits of competition boil down to the dictum that policy should be delivered at the smallest level that can internalize externalities.
We found that this rule actually describes the federations in our study (see also Peterson, Reference Peterson1995; Adolph et al., Reference Adolph, Greer and da Fonseca2012). Financing is organized through shared budgets with strong equalization components (in the NHS systems, Canada, Australia, Italy and to some extent in the USA through Medicaid), and in the social insurance countries the social insurance system and its equalization mechanism are set up with national legislation and scope. The federal social insurance countries are if anything more prone to separate territory and health financing. Germany's constitutional court makes a point of asserting that the logic of social insurance is alien to the logic of territory (Mätzke, Reference Mätzke, Font and Greer2013), and as if to make the point the territorial distribution of health expenditures in social insurance countries is basically unreported (Scott and Heather, Reference Scott and Heather2019). Austria, the most thoroughgoingly Bismarckian of European welfare state, reports basically nothing with territorial breakdowns despite the practical importance of the interaction of its federal states and their social insurance funds in actual operation of the system (Trukeschitz and Riedler, Reference Trukeschitz, Riedler, Greer and Elliott2019; also Mätzke and Stöger, Reference Mätzke, Stöger, Fierlbeck and Palley2015). The main, and still partial, exception to this finding was that the welfare states with strong stateless nations had more powerful regional governments than others, and were more likely to have some kind of asymmetry, as in the separate financing systems for the Basque Country and Navarre, or the complex of asymmetries that make up the UK devolution arrangements.
In other words, the territorial organization of health financing is surprisingly similar across federations as distinctive as Mexico and Austria. The central government either raises the money through taxes or orchestrates the social insurance framework that raises the funds. The money then moves to either regional governments in NHS-model systems, which might have to co-finance or choose to supplement from their own tax bases, or social insurance funds, which purchase care from providers which might have received support for capital costs from the regional government out of its budget. In either case, the core of the running expenditure of the system is central and has a large risk pool. The regions will then typically centralize hospital service planning and capital investment, regardless of system.
This model is almost the opposite of the EU's developing health federalism – rather than trying to place the EU on a continuum it is almost more effective to see it as a sort of photographic negative of the other rich federations, doing what they don't in regulatory areas, and not doing what they do in social policy. Even if we consider the supplementary capital financing from the likes of the European Investment Bank, the entire EU budget pales relative to the average costs of health care in EU member states, which are 7% of GDP in 2017 (Eurostat, 2019). Insofar as there is any larger support for member states, it takes the form of a promise that the European Central Bank will step in to defend member states that are compliant with the fiscal governance regime should they come under speculative attack in bond markets (Schelkle, Reference Schelkle, Zahariadis and Buonanno2018) – useful, but quite unlike the financing models that sustain health care services. In other words, the EU, as a federation, is leaving risk pooling at an intermediate level, member states, and some of the member states, and their risk pools, are quite small.
The federal politics of this regulatory state form are also therefore quite different. Most federations are marked by a considerable vertical fiscal imbalance, in which the central state enjoys greater and more flexible taxation and debt-issuing capacity, and to a greater or lesser extent uses it to redress horizontal fiscal imbalances between richer and poorer regions or richer and poorer people. In the EU, the vertical fiscal imbalance runs downward, and the member states control the resources (more like the United States under its Articles of Confederation than any contemporary federation). Unsurprisingly, the rich states do not show much appetite for redressing horizontal fiscal imbalances .
The EU's (de)regulatory efforts, by contrast, go far beyond what other federations do (Matthijs et al., Reference Matthijs, Parsons and Toenshoff2019). In particular, the market-making efforts in areas such as professional qualifications are much more thoroughgoing than the professional mobility found in other federations. The United States, for example, has no interstate mutual recognition of most qualifications, so a nurse who wants to move from one state to another has to start the qualifications process anew. In the EU, by contrast, even the use of language tests to control health care practice across borders is circumscribed in the name of preventing discrimination.
5. Conclusion
This paper compared the EU to existing federations, based on new comparative research in federalism, and found that the EU is very unusual among federations. Its second and third faces, of market making and fiscal constraint, are more important and powerful than its first face, of health competencies. This means that the history of EU health policy is mostly of policies made on other treaty bases, for other reasons, and extended into health by people who might not be interested in or knowledgeable about health. This makes the EU quite unlike all other wealthy federations, where the federation exercises less regulatory authority over the operation of health care and does much more redistribution and risk pooling between people and territories.
The EU is, then, unusual among federations. It is strong where others are weak, as in the harmonization of professional qualification and promotion of cross-border mobility, and it is weak where others are strong, as in efforts to finance services or redress the consequences of markets. Insofar as its law creates what seem to be attractive new rights, such as the right to cross-border patient mobility, it has created fiscal and regulatory pressures on member state governments that do not clearly contribute to health. Such is the case of patient mobility law (Martinsen and Blomqvist, Reference Martinsen, Blomqvist, Magnussen, Vrangbaek and Saltman2009; Greer and Rauscher, Reference Greer and Rauscher2011). Even where it does to some extent redress inequality, its distinctive history and politics mean that it does not do so in the ways that member states do. Financing capital expenditure in Hungary might buy the acquiescence of Hungarian politicians who can use it for clientelism (Magyar and Vasarhelyi, Reference Magyar and Vasarhelyi2017; Bozóki and Hegedűs, Reference Bozóki and Hegedűs2018). It does not constitute any kind of needs-based redistribution, or necessarily address within-country health inequalities, or, by definition, finance health care.
At the same time, the sort of local restrictions on trade (e.g. licensing rules) that often maintain a middle class in, for example, the United States, are under relentless attack by EU policies with their legal theme of antidiscrimination, and their substantive content of market-making through liberalization of restrictive labor contracts and economic rules (Matthijs et al., Reference Matthijs, Parsons and Toenshoff2019). State aid and competition rules further limit, by design, the ability of public authorities to use their power to promote favorite businesses (Guy, Reference Guy2018). The process might produce better economic performance than we would otherwise see, but it also might alienate the local middle classes of small businesses, licensed notaries, hairdressers and others whose protections and political strategies are threatened by EU law.
The risk is that the EU is playing out the same processes that turned Southern Italy – the mezzogiorno – into the problematic region that it is today. Scholars of Italian economic history have a long-standing debate about the economic consequences of Italian unification, with many arguing that the apparently permanent underdevelopment and corruption of the south is a function of the way Italian unification worked (Toniolo, Reference Toniolo2013). To put their argument briefly, the disparities between north and south were not big or growing until unification, though the economies of north and south already worked quite differently. Unification in a liberal state with a single currency and massive internal imbalances (like the EU) put the southern industries at a permanent disadvantage, leading to outmigration, decline and turbulent politics. State expenditure to shore up support from southern elites and develop the south were, in the context of seemingly structural southern decline, routed into clientelism and corruption, which entrenched local authoritarians and organized crime still more without redressing the regional inequalities. The combination of liberalism, a currency union, peripheral decline and side-payments produced the apparently intractable corruption and underdevelopment of the Italian south today (Sassoon, Reference Sassoon2013). The EU might be enacting this process on a continental scale, with parts of its southern and eastern peripheries showing the combination of stagnation and corrupt, authoritarian use of EU funds that the mezzogiorno saw (for the distorting impact such funds can have on recipient state politics, modeled as a case of the ‘resource curse’, see Huliaras and Petropoulos, Reference Huliaras and Petropoulos2016). If the history of the mezzogiorno and the carnival of national stereotyping seen during the debt crisis are any guides then we can expect the southern and eastern countries to be roundly blamed for their future misfortunes, and the development of still more cultural explanations for their failure.
Karl Polanyi taught generations of scholars that the market, to survive, requires cushions, policies that compensate for the social disruption caused by the commodification of different aspects of life (Polanyi, Reference Polanyi1944). The EU has been a particularly elegant and powerful machine for the extension of market logics, since both current politics and inherited law point to greater liberalization and fiscal discipline as its key goals. Its second face is hostile to local regulations, such as restrictive licensing and union contracts, that can be challenged as discriminatory but help compensate for exposure to the market. Its third face was designed to be hostile to policies that might compensate with additional spending, though its impact and clarity is being undermined as it acquires health goals and objectives.
The basic deal of European integration, like most expansions of free trade, is that the overall society will win, with lower prices, higher productivity, and a better quality of life, and government will compensate losers through welfare state programs, including good health care, that help them manage the effects of economic dislocations and, when feasible, move into more successful sectors (what economists associate with the Kaldor–Hicks improvement) (Hübner, Reference Hübner2016). In the case of the EU, the asymmetry is constitutional in that the EU promotes and regulates markets and the member states compensate for them. In health policy, that means the EU regulates member state budgets and health care systems to comply with the demands of the Eurozone and internal market, but the member states are still obliged to finance and provide health care.
Given this situation, and the slim likelihood that the EU will ever finance health care delivery in any meaningful way there are three possible useful futures that would contribute to health if not approximate the structure of other federations in health care. The first might be to focus on developing useful EU capacities such as research, data collection and comparison and communicable disease surveillance and control, including the activities of the ECDC. Much of the most useful EU activity for health has been in this kind of first-face activity. A second would be to emphasize health as part of activities grounded in other treaty bases. The second face of EU policy has not just been the market-making one seen in patient mobility, but has also underpinned a variety of useful policies that promote health under different names. Environmental protection (Art. 191) and social policy (Arts. 153 and 156) mention health as objectives and have contributed to health in the EU. Medicines and medical devices regulation are both squarely grounded in internal market (second face) law, but there is a sustained campaign to regulate them in the interests of health as well as industry and the market. The European Semester's gradual shift toward constructive and more egalitarian health policy ideas probably marks its decline as an instrument of austerity or coercion of member states, but that means it is an opportunity to promote health policy objectives now and an invitation to consider what might happen in the increasingly ambitious third face policies in the next economic downturn.
These policy ideas would build on existing developments and compensate for the EU's unusual structure. Even so, the comparison with other federations suggest the EU stands particularly exposed among federations, with a large and largely pro-market regulatory arm and very little redistributive capacity with which to compensate people displaced by market or regulation. No other developed country federation works that way. They all exploit the greater risk pools and fiscal strength of the larger unit to compensate people and places for their losses. It is not hard to see why the EU failed to develop that way. It faced (and largely avoided) the unique challenge of integrating developed welfare states. In the hands of governments with a neoliberal bent, its reactivation since the 1980s since extended its regulatory state features and constitutional asymmetry. Perhaps that was the only way to make the EU the powerful polity it is today, but it is also one that works nowhere else, and that might give us all pause.
Acknowledgements
I would like to thank the editors of the special issue, two referees, my colleague Holly Jarman and an audience at Johanns Kepler University Linz for their helpful comments.
