Health Care System Retrenchment

(1)

HEALTH CARE SYSTEM RETRENCHMENT

EXPLAINING THE DIVERSITY OF POLICY RESPONSES TO THE GLOBAL FINANCIAL CRISIS

Master Thesis Presented to the

Nijmegen School of Management at the

Radboud University

In Partial Fulfillment of the Requirements for the Degree of

Master of Science (MSc) in Political Science

Supervisor: Prof. dr. M. Verloo

Second reader: Dr. K. Jacobs

Submitted by

Jens Pauw (4384431)

(2)

‘Thinking without comparison is unthinkable’

(Swanson, 1971: p. 45 in: Rihoux & Ragin, 2009)

Acknowledgement

At this moment of submission of the thesis, I will complete the master program in Political Science as well as I will end my life as a student. Handing in the final piece of work feels like the finishing touch to an intense but above all very valuable period in my life. I’m very grateful that I was able to finish the study with the topic of health care that particularly interests me.

This thesis would not have been possible without the time, guidance and support of many people.

First and foremost, I would like to thank my thesis supervisor Prof. dr. Mieke Verloo for her critical and encouraging support during the research period. I will consider her unconditional faith and the freedom she gave me, which has inspired me the most.

I have been privileged to experience first-hand how critical and passionate teachers in the Comparative Politics section can make a difference. Without the excellent academic fundamentals they taught me, it would have been impossible to accomplish this thesis. I also thank my fellow students, for the many discussions that frequently led to key insights.

I would also like to thank my friends for their understanding and interest, for helping me to enjoy my life-besides-thesis and for enriching my life. And of course many thanks go to my roommates, for creating the perfect environment to write my thesis at home and for their patience with me for all that nights out I missed.

And last but not least, I am very grateful for the warmth and support of my family. They gave me the opportunity to study and to complete the master. Without them, I would never have come this far.

Jens Pauw Tilburg, May 2016

(3)

Abstract

Recent studies have argued that European welfare states have been under tremendous pressure since the onset of the recent Global Financial Crisis (GFC). The health sector, in particular, has been affected by retrenchment politics aimed at reducing public spending. Cross-national variation, however, is clearly observed as the policy responses have manifested themselves unevenly across the landscape of welfare states. This thesis is the first to examine the relationship between welfare state retrenchment, health care systems and the GFC. Using crisp-set qualitative comparative analysis (csQCA), the study intends to identify patterns of welfare state retrenchment in health care system trajectories of European countries during the (aftermath of the) recent financial crisis. The results suggest that for the occurrence of health care system retrenchment various factors in different combinations are important. Three sufficient paths towards retrenchment are observed: 1) the presence of right-oriented governments combined with being a non-liberal market economy; 2) the absence of obfuscation tactics combined with being a liberal market economy; and 3) the presence of obfuscation tactics along with lower economic growth rates and EU-pressure. Further analysis was revealed to examine why certain countries cluster around a particular pathway. We suppose that these patterns can be explained by looking at the incentives for political action. The conclusion drawn is that three particular logics could be observed: retrenchment as a window of opportunity, retrenchment as a market-function approach and retrenchment as a spillover effect. Although the findings only permit for modest generalization and further work needs to be done, the results do reveal that health care system retrenchment a topic worthy of further investigation.

(4)

List of abbreviations

CEE Central and Eastern Europe

CEEC CME Central and Eastern European Countries Coordinated Market Economy

CEEC LME Central and Eastern European Countries Liberal Market Economy

CME Coordinated Market Economy

EC European Commission

ECB European Central Bank

ECJ European Court of Justice

EFSF European Financial Stability Facility

EMU European Monetary Union

ESM European Stability Mechanism

ESS European Social Survey

EU European Union

GDP Gross Domestic Product

GFC Global Financial Crisis

GGHE General Government Expenditure on Health

HFCM Health & Financial Crisis Monitor

IMF International Monetary Fund

INUS Insufficient but Necessary part of a condition which is itself Unnecessary but Sufficient

ISSP International Social Survey Programme

LME Liberal Market Economy

MME Mixed Market Economy

MSDO Most Different, Similar Outcome

MSZP Magyar Szocialista Párt

NHS National Health Services

OECD Organisation for Economic Cooperation and Development

OMC Open Method of Coordination

ParlGov Parliament and Government Composition Database

PASOK Panellinio Sosialistiko Kinima

PHI Private Health Insurance

PS Partido Socialista

QCA Qualitative Comparative Analysis

SHI Social Health Insurance

THE Total Health Expenditure

VoC Varieties of Capitalism

WFS Welfare State

(7)

(8)

Chapter 1: Introduction

1.1 Inducement

As one of the biggest features of national welfare state models, health care budgets have grown to immense proportions. Even since the end of the ‘Golden Age of Welfare’ – the 1970s – expenditures have continued to rise in all European countries (Freeman & Rothgang, 2010, p. 371-372). In comparison with other welfare state elements, it is assumed that at least three particularities – taken together – make health care a unique area of research. First, health care is mostly based on services rather than transfers, making them completely different from pensions and public housing. Second, health care is a capital- and human-intensive sector. It therefore demands huge investments in technology, infrastructure and salaries of medical professionals. And third, various mechanisms tend to offer a universalistic-like coverage, however, there is differentiation among its users ( Pavolini et al., 2013).

Health care should be seen as a complex and rather specific feature within the welfare state landscape that is in a constant state of political flux. Technological progress, demographic and political transformations, and radical changes in the guiding principles of governmental intervention have put traditional health models to the test. Despite debates regarding the increasing health care expenditures that have been around for decades, and despite the fact that cutbacks have taken place in almost all social security programs, health care services were spared from austerity for a while. As it is assumed that welfare state reform is very difficult in general (Pierson, 2002), the aforementioned characteristics make reform in health care systems even more difficult. It is seen as a ‘quadrilemma’, in which four objectives need to be combined (Pavolini et al., 2013). These are: 1) economic objectives – to control growth in expenditures; 2) social objectives – to guarantee universalism of access; 3) medical objectives – to guarantee the quality and 4) political objectives - to guarantee the overall responsiveness of the system. Trade-offs lie are looming when all objectives are addressed together, for example between universalism of access and cost reduction. The politics of health care system reform are therefore profoundly delicate and slowly moving forward.

The quadrilemma was expected to continuously keep the system in balance. We would see an upward trend in health expenditures and limited opportunities for reform (Freeman & Rothgang, 2010; Mladovsky et al., 2012). These stable expectations changed dramatically in 2008 when the European countries succeeded decelerating their health expenditures for the first time since the

(9)

1960s (Morgan & Astolfi, 2014).Since the Global Financial Crisis (GFC) affected most sectors, this is definitely the case for the health care sector. It is seen as a ‘critical juncture’, challenging the socio-economic and politico-socio-economic foundations of health care systems (Helderman, 2014). Although rather unpopular, some governments have responded with policies aimed at reducing public spending on health care. Some of the more common reforms across countries have endeavoured to moderate the growing budgets for health services, rethink the benefit packages, transfer health expenditures from public to private paying and implement far-reaching reforms in the pharmaceutical market (Mladovsky et al., 2012). Several countries reported cuts in their national health budget in response to the financial crisis, among others Bulgaria, Croatia, Iceland, Ireland, Italy, Greece and Spain. Other nations increased or introduced user charges for health services in response to the crisis, such as Czech Republic, Denmark, Estonia, France, the Netherlands, Portugal, Romania, Slovenia and Switzerland. On the other hand, social insurance revenues and expenditures continued to increase in Austria, the Czech Republic, Poland and Slovakia. So it is seen that the European countries have responded to the financial crisis in various ways, and that health care system retrenchment has been introduced to a lesser or greater extent.

This thesis is concerned with explaining under which conditions national governments have pursued health care system reform in the aftermath of the financial crisis. This will be done from a welfare state retrenchment perspective. Comparative welfare state scholars have long been concerned with explaining the nature and direction of cross-national welfare state developments. Since the early ‘90s, however, Paul Pierson (1994) opened up a debate in which he claimed that welfare state expansion theories are not able to explain (potential) dismantling. This ongoing debate has brought forward an ever increasing amount of literature and theories attempting to address the question of why and under which circumstances cutbacks take place in highly developed welfare states despite formidable political obstacles (Starke, 2006). Focusing on health care retrenchment during the financial crisis gives a unique opportunity to test several welfare state retrenchment theories that have been developed for their particular usefulness concerning health care.

1.2 Health care systems

1.2.1 Definition

Health care systems seem to be essentially contested (Gallie, 1994). Health care systems have been conceptualized and defined in various ways for different purposes. Starting with its uncontested parts, it comes as little surprise that its main focus is on people’s health. The systematic component of the definition refers to the ‘arrangement of parts and interconnections that come together for a purpose’ (Von Bertalanffy, 1968). This links all actors that have their influence - such as patients,

(10)

doctors, ministers, politicians, health providers, pharmaceutical companies, etc. – together with the roles and functions they play.

In essence, different definitions mainly reflect different understandings of the health system boundaries. Traditionally, it was referred to in terms of capacity indicators, such as numbers of beds. More modern definition, however, deliver an all-encompassing concept embracing any activity, resource and/or institution that results in health improvements. Roemer (1991) has defined a health system as ‘the combination of resources, organization, financing and management that culminate in the delivery of health services to the population’. He argues that it has five principal component parts: resources, organization, management, economic support, and delivery of services (p. 335). This definition is, however, problematic because it does not explain why and how the components differ in producing a set of outcomes (Hsiao, 2003). The European Observatory for Health Systems & Policies defines health systems as the ‘people, institutions and resources, arranged together in accordance with established policies, to improve the health of the population they serve, while responding to people’s legitimate expectations and protecting them against the cost of ill-health through a variety of activities whose primary intent is to improve health’ (Hoffman et al., 2012).

A more broadly recognized definition – used among others by the World Bank and the EU – was given in the World Health Organization’s (WHO) World Health Report (2000), in which it is stated that: ‘a health care system consists of all organizations, people and institutions producing actions whose primary intent is to promote, restore or maintain health’. Six categorical building blocks are defined: service delivery, health workforce, health information systems, medical products, vaccines and technologies, health systems financing and leadership and governance. It should be noted that this definition has some drawbacks, for example it lacks to highlight the relationship between the system’s key functions and its performance (Hsiao, 2003). On the other hand, it recognize health care as a service, in which interventions are not only curative but also preventive and promotional. Further, it emphasizes the stewardship role of the government. These two features are important considering the (political) purpose of this thesis. Therefore, from now on, the meaning of health care systems will implicate the WHO’s (2000) definition: a health care system consists of all organizations, people and institutions producing actions whose primary intent is to promote, restore or maintain health.

1.2.2 Six health care system objectives

Considering the widespread interest in the conceptualization of health care systems, it is a logical question to ask how one could describe the aim of a health care system. Why should there be an institutionalized system? Six objectives in national health care provision can be distinguished.

(11)

First, the system arranges the financing of health services through taxation, social insurance contributions or private means. The way health is financed depends on historical trajectories that have led to several types of systems. Second, the system steers the provision of health care, which can be carried out in state-run facilities respectively by state-based actors, in societal-based facilities, or in private-for-profit facilities respectively by private actors. Third, the system will regulate all actors of the various aspects of financing and provision (Wendt et al. 2009). Fourth, the system provides services, provided by all actors that are mandated to improve health. This can be done on an individual level (doctors, nurses) but also on higher ones (a ministerial campaign on anti-tobacco). Fifth, generating resources is a requirement for the creation and development of health care systems. These can be both personal- (medicines) and population based (education, infrastructure). Sixth and finally, a health system provides stewardship of health services to manage the well-being of the population. This can be done by the government or other actors that try to maximize the likelihood of achieving the system’s goals (Duran et al., 2011).

Every particular objective is important, as is their interconnection. For instance, services could be personal (every young child should be vaccinated) but also population-based (develop a vaccination campaign). Hsiao (2003) argues that there could always be trade-offs at stake, because governments want to achieve multiple objectives, but only have limited resources. These trade-offs could be inter-sectoral – do we fund education or health? – and intra-inter-sectoral - do we invest in medical technology or in waiting line reduction? The way governments deal with these trade-offs are influenced by the priorities of the health system concerned (Duran et al., 2011: 26). Together, the six mentioned objectives form the framework of each European health care system.

1.2.3 Different types of health care systems in theory

During the ‘Golden Age’ of welfare state expansion, the European states attempted to include more and more people under their health care system. By the mid-‘70s, almost all citizens had access to state-regulated health provisions (Leibfried & Zürn, 2005). Nevertheless, there were considerable differences between health care systems. A comprehensive conceptual framework was therefore needed, so that researchers were able to analyze health care systems and their transformations. It seems that classification is a common business in the field, which has resulted in the co-existence of numerous different classifications.

Some earlier typologies concentrate on financing and service provision as main indicators. These are, not surprisingly, the two main features every of the six described health care system objectives are linked to. In an influential OECD study, health care systems are categorized according to the dimensions ‘coverage’, ‘funding’ and ‘ownership’ (OECD, 1987). The result was that National Health

(12)

Services (NHS) have been contrasted to Social Health Insurance (SHI) and Private Health Insurance (PHI), which in fact ideally contrasted the United Kingdom, Germany and the United States. The NHS-system is tax-financed, organized by the state within a hierarchical NHS-system and hospitals are mainly publicly owned. In a PHI-system, by contrast, health care is subject to the constraints of the market. Doctors work mainly as entrepreneurs and are paid directly (‘out-of-the pocket’) or by commercial insurers. A SHI-system, then, is situated in an intermediate position between these two, which means that hospitals are generally owned by local-governments, non-profit organizations or private firms. The funding of health care could be regional, industry based, or through occupational schemes‐ (Wendt et al., 2011; Freeman & Rothgang, 2010).

Due to the wide variation of different systems within the typology, and the recognition that the aspect of regulation by the state should also be included, more modern typologies were developed (Freeman & Rothgang, 2010). Moran (1999), for example, introduced the dimensions of funding, service provision and governance. He emphasized that we should see the whole system as a ‘health care state’, resulting in four types: ‘entrenched command-and-control states’ (Scandinavian countries and the UK), the ‘supply state’ (US), the ‘corporatist state’ (Germany) and the ‘insecure command-and-control states’ (Portugal, Greece). By referring to Moran’s typology, in their influential article Wendt et al. (2009) suggested combining the dimensions of ‘financing’, ‘service provision’ and ‘regulation’ with the level of involvement by the state, non-governmental actors and the market. This resulted in a taxonomy of 27 theoretical possible health care systems, though, they identified three ideal types. These comprise ‘state health care systems’, in which financing, service provision and regulation are carried out by state actors, ‘societal health care systems’, in which social actors – for example social insurance funds – are dominant in all three dimensions, and ‘private health care systems’, in which all three dimensions fall under the auspices of market actors.

1.2.4 Different types of health care systems in empirical reality

Comparative studies that have applied the aforementioned methodological framework to the empirical reality have described a richer descriptive portrait. Böhm et al. (2013) have used the taxonomy of Wendt et al. (2009) for allocating different country systems to different typologies. By arguing that there is a hierarchical relation between the three dimensions, which implies that the superior dimension restricts the nature of the subordinate dimensions, they limited the number of theoretically plausible types to ten. Their empirical results show that after classification of 29 of the 34 OECD-countries, one could distinguish between five different health care system types, as presented in Table 1.1. These are the National Health Service, the National Health Insurance, the Social Health Insurance, the Etatist Social Health Insurance, and the Private Health System.

(13)

In comparison to other classifications, countries that are now grouped within the National Health Insurance system are specified from National Health Services for their political preferences for more private involvement under state control, particularly in the provision of health services. Further, Belgium, France, Israel and the Netherlands that have often been categorized under Social Health Insurance systems, are now specified for their minor role of social actors as an Etatist Social Health Insurance. Also the Central and Eastern European (CEE) OECD-countries fit to this latter cluster, because adequate societal actors to whom regulatory powers could be devolved have never been in existence since the breakdown of the Soviet regime.

Table 1.1: Dispersion of OECD health care systems in 29 of 34 OECD-countries (Böhm et al., 2013, p.263)

Health care system type Regulation Financing Service provision

Cases

National Health Service State State State Denmark, Finland, Iceland, Norway, Sweden, Portugal, Spain, UK

National Health Insurance State State Private actors Australia, Canada, Ireland, New Zealand, Italy Social Health Insurance Societal actor Societal

actors

Private actors Austria, Germany, Luxembourg, Switzerland Private Health Insurance Private actors Private

actors

Private actors USA

Etatist Social Health Insurance

State Societal actors

Private actors Belgium, Estonia, France, Czech Republic, Hungary, the Netherlands, Poland, Slovakia, Israel, Japan, South-Korea

It is important to note that each country-specific health care system has a unique set of characteristics. There are no two of a kind. The overview by Böhm et al. (2013) is therefore a typology – a classification in which the cases belong to one, and only one, of the types – rather than an ideal type, in which the question is to what extent the case belongs to the classification (Greve, 2012). This has consequences for Slovenia, not mentioned in the matrix, as the country conflicts with the author’s logic of hierarchy. Here, the state still provides most of the health care services with its own facilities while funding and regulation are dominated by societal actors. Nevertheless, Böhm et al. (2013) suggest that ‘the special case of Slovenia gradually evolves into a well-known type: the Social Health Insurance system’ (p. 268).

1.3 Financial crisis, retrenchment and health care system reform

The aftermath of the Global Financial Crisis (GFC) of 2008 inevitably marked a ‘stress test’ for the welfare state, as it fundamentally redrew the boundaries between the state and the market. How could we put the current politics of retrenchment in the right perspective? This paragraph deals with

(14)

the responses adopted across Europe since the outbreak of the crisis. It reviews in short how the crisis erupted and how it evolved over time. Subsequently, it reviews the actual consequences for the welfare state in general and thereafter for health care systems in specific.

1.3.1 The Global Financial Crisis: an introduction

The Global Financial Crisis is generally defined as the largest financial shock since the Great Depression in the ‘30s and could be distinguished for its four sequential episodes. The first phase is commonly believed to have begun in early 2006 when the subprime mortgage market in the U.S. began to display an increasing rate of mortgage defaults. These defaults led, in late 2006, to a decline in US housing prices after nearly a decade of exceptionally high growth, resulting in the collapse of the US subprime mortgage market in the spring of 2007 (Karanikolos et al., 2013). The second phase has been the international banking crisis that emerged from the US subprime crisis. Banks all over the world – both US and European - with investments linked to those mortgages started losing money. A couple of them (for instance Merrill Lynch, German Hypo Real Estate, FannieMae and FreddieMac) were rescued by the government or taken over by other banks. A climax was reached when Lehman Brothers filed for bankruptcy on 15 September 2008 (Starke et al., 2013). Governments around the world launched multi-trillion dollar rescue programs designed to prop up the international financial system, which only aggravated the panic. The fear that more banks could fall caused investors and banks to take extreme precautions: the interbank money markets froze when banks stopped extending loans to each other.

A third phase started late 2008, when banks increasingly affected national governments. The markets raised questions whether some countries could afford to rescue their troubled banks. At this time, the crisis ceased to be a simple financial crisis and spilled over to the ‘real’ economy (Starke et al., 2013). A credit freeze occurred, resulting in a collapse of consumer and investor confidence. The impact of the recession on the national economies was ubiquitous. Although the economic downturn started in 2008, the peak of the recession was seen in 2009. The real GDP fell across the EU-27 countries by nearly 5 percent. Unemployment rates soared to 7.7 percent on average by 2009 and rose even further in 2010. Continuously, the combination of huge costs for the banking sector rescue packages coupled with lower tax revenues resulted in large governmental budget deficits.

The fourth phase can then be marked in late 2009 and early 2010 when Greece, and subsequently Ireland, Portugal, Spain and Cyprus, came under pressure from sovereign debts markets. These countries were unable to borrow on financial markets at reasonable interest rates, forcing the European Union to step in. Large crisis resolution mechanisms were created, starting with the

(15)

European Financial Stability Facility (EFSF) in May 2010, a 750 billion euro package financed by the EU-member states and the IMF, to the European Stability Mechanism (ESM) in July 2012. The latter constituting a permanent instrument with the lending capacity up to 800 billion euro. Austerity policies, including large-scale cuts and public sector reform, were imposed as pre-conditions for the rescue packages in countries that needed them (Schweiger, 2014). Europe-wide, the Global Financial Crisis and its aftershocks have led to a divergent restructuring and remapping of the national welfare states (Starke et al., 2013).

1.3.2 Welfare state retrenchment in the aftermath of the GFC

Since Paul Pierson's seminal work The New Politics of the Welfare State (1996), retrenchment has become one of the most common terms to describe recent welfare state developments. Much energy has been spent in social policy studies on discussing the phenomenon and its possibilities (Palier, 2000). The term itself lends to a period in the history of welfare states: emergence (late nineteenth century until 1945) was followed by growth (the Golden Age, mainly until the 1970s); limits (or even crisis, the 1980s); and then retrenchment (since the late 1980s). Welfare state retrenchment is generally seen as a highly unpopular endeavor and, therefore, politically difficult to act on. The Global Financial Crisis seems to be path breaking as it paved the way for fundamental retrenchment.

It is therefore especially meaningful to look at the field of social policy during the aftermath of the GFC, as welfare state schemes often account for the bulk of public expenditures. As a major example of path dependency, the welfare state should exhibit a special pattern of incremental change during economic critical times (Van Hooren et al., 2014). The European welfare states have faced two main challenges during the recent financial crisis. First, the slower economic growth and productivity has raised concerns about the financing of welfare capitalism. Weaker economic growth has adversely affect social security systems. Due to the economic downturn, social expenditures increased because of a greater need for social support, like unemployment and housing benefits. On the other hand, GDP grew slowly or decreased. This challenge was therefore two-sided (Diamond & Lodge, 2013). Second, as a result of these developments, society is changing. New groups of winners and losers in society emerge, which shapes a new landscape of social inequalities and consequently asks for new priories in public expenditure (Busch, 2010). Welfare states have to react on that: on the one hand, they have to reconsider what kind of social problems they want to focus on, while on the other hand, they have to consider how to react on the shrinking public budgets.

Two distinct periods of governmental reaction in the aftermath of the GFC are noted. Initially, welfare state institutions were used to protect the citizens from the direct, harsh consequences, like unemployment. In the 2008-2009 period, countries responded to the crisis with Keynesian inspired

(16)

deficit spending. Welfare state spending as part of the GDP even increased in this period in almost all European countries, as social spending increased but the GDP did not. This situation changed significantly when governments turned from deficit-spending to austerity, such that ‘the 2010-2012 period came to be marked, in virtually all member states, by a succession of ‘structural’ reforms’ (Pochet & Degryse, 2012, p.1). One country after another adopted austerity packages, and this was not only done by countries with the weakest economies. Overall, as Hermann (2014) argued, ‘the scale of austerity adopted in response to the crisis is unprecedented in European postwar history’ (p. 10).

Despite country-specific variations, austerity and structural reform have fueled the erosion of welfare protection. Between 2009 and 2010, total social protection expenditure fell in several countries, among others Greece, the UK, Estonia and Hungary (Hermann, 2014). In 2011, this latter expenditure – as well as education and health expenditures - declined on average in the EU-27 and in most individual Member States, as showed in Figure 1.1. The unfolding sovereign debt crises in Europe together with slow economic growth has put additional pressure on public finances, resulting in many governments resorting to social retrenchment.

Figure 1.1 - Changes in real government expenditure, EU-27 and EA-17 (European Commission, 2014, p. 67)

1.3.3 Health care system retrenchment in the aftermath of the GFC

Health care systems are more likely to be affected by retrenchment in times of financial crisis than other features of the welfare state (Pavolini et al., 2013; Helderman, 2014). This is because public health spending accounts for a substantial proportion of the total governmental expenditure. It is

(17)

difficult arguing to maintain current levels of spending because there is more scope for efficiency gains in a health system – which is complex and often not clearly defined – compared to pensions, for example. Retrenchment could take various ways here, but generally cuts are established through policies intended to change the level of contributions for publicly financed health care (national health budgets, user charges), the costs of publicly financed health care (prices of medicines, salaries) and the volume and quality of publicly financed health care (population coverage, promotion and prevention) (Maladovsky et al., 2012).

Recent data on health expenditures in the aftermath of the Global Financial Crisis seems indeed to suggest a similar pattern across Europe. Budgets in several countries faced significant cuts in the years following the crisis, which reversed a long-term upward trend in public expenditure. Cuts were not always immediately apparent after 2008, since there was a time lag as the health care budgets adjusted to the financial crisis (European Commission, 2013). Public spending on health fell or slowed in many countries between 2008 and 2012, both in absolute terms and as a share of government spending. Some changes were fundamental, as for example in Greece where expenditure was reduced from 9.8% of the GDP pre-crisis to 6% of GDP post-crisis. In the latter country, and in Croatia, Ireland, Latvia and Portugal, public spending on health was still lower in 2012 than it had been in 2007. Further, in some countries budgetary changes were relatively small, but still impressive when taking into account that before the crisis budgets rose constantly. Table 1.2 gives an overview of countries in which the per capita public spending on health differed from historical patterns by more than two standard deviations. It is remarkable to note that not only the countries that were under the heaviest austerity policies are listed here. This seems to suggest that cuts in public spending are not only due to countries’ budgetary matters but also caused by non-economic factors.

Table 1.2: Countries in which changes per capita public spending on health were greater than historical rates (=two standard deviations) (Thomson et al., 2014, p.35)

Policy responses varied across health systems in Europe. Several countries increased or introduced user charges for health services (Czech Republic, Denmark, Estonia, France, Greece, Ireland, the Netherlands, Portugal, and Romania), while, by contrast, others (Austria, France, Ireland, and Italy)

(18)

reported expanding benefits. Several countries including Ireland, Romania, Slovenia and the United Kingdom froze or reduced wages in the health sector. Several governments restructured the Ministry of Health, health insurance funds or other purchasing agencies in order to reduce overhead costs. Examples of policy change are numerous (Quaglio et al., 2013; Thomson et al., 2014). Overall, the crisis accelerated reforms in some countries, while other reforms were resisted. At least, it is clear that the GFC should be marked as a ‘health system shock’: ‘an unexpected occurrence originating outside the health system that has a large negative effect on the availability of health system resources or a large positive effect on the demand for health services’ (Maladovsky et al., 2012, p. v).

1.4 The research problem

1.4.1 Scientific research puzzle

As we have just seen, the retrenchment process welfare states – and in specific health care systems – have gone through during the Global Financial Crisis and its aftermath, differs substantially among countries. Some were affected more than others, though there was wide variation in sphere and scope. In addition, I expect that these differences are not only due to economic factors. In this sense, some meaningful questions need to be asked. To start, why and under which conditions have some governments proposed for health care system retrenchment while others did not? Another relevant question is why have several countries faced a similar outcome – i.e. retrenchment, whereas they seem to experience a different economic, political and societal landscape?

The existing literature offers a number of theoretical frameworks attempting to explain welfare state cutbacks in general. Retrenchment has usually been studied from four approaches. These are all perceived to be theoretically relevant in the literature. Political conflict theory, institutionalism, neofunctionalism and social constructivism all aim to explain cross-governmental conditions of retrenchment. These frameworks will be further elucidated in the theoretical framework (Chapter 2) of this thesis.

I have observed, however, several critical problems with contemporary welfare state retrenchment research. First, the empirical comparative research till date is considerably dispersed. Most authors concentrate on one or two specific cases, resulting in a multitude of case study research. Conditions that can explain welfare state retrenchment could be valid for these particular cases, but do not have to be generalizable to other cases. Some authors argue that ‘blame avoidance’ (Pierson, 1994) is a key condition, while others argue for ‘partisanship’ (Green-Pedersen, 2002) or ‘institutional fragmentation’ (Jensen & Mortensen, 2014). A few researchers have made use of large-N studies, however, they could only make claims about the relative influence of the particular conditions and neglect the influence of the relationship between these conditions.

(19)

Secondly, welfare state retrenchment theories aim to be applicable to all welfare state features. This means that they, at least theoretically, could serve as explanation for health care systems. And that potentially all theories could have their influence on this specific feature. The proposed theories often contradict each other, but they are not mutually exclusive. They seem even supplementary to each other. Empirically, however, these assumptions are understudied in the case of health care and therefore deserves more research attention.

Thirdly, welfare state retrenchment theories generally study actor motives and orientations toward retrenchment at a given moment in time (Starke, 2006). It is, however, also important to take events at a higher systematic level into account. In political economy, is largely undisputed that macroeconomic shocks trigger policy change (Starke, Kaasch & Van Hooren, 2013). A crisis does not only affect economic policy, but also social policy. Some researchers refer to an emerging crisis as a ‘critical juncture’ (Helderman, 2014) or a ‘window of opportunity’ (Wilson, 2000). More scientific focus should therefore be paid to the recent Global Financial Crisis (GFC) and its implications for the welfare state.

Fourth, and foremost, there seems to be an academic ‘blind spot’ for retrenchment on health care – as feature of the welfare state – in recent crisis years. This research will include and combine three particular features: welfare state retrenchment, health care systems and the (aftermath of the) GFC. There is, however, a structural lacuna in the literature concerning the inclusion and combination of these interconnected themes. The three features are on the research agenda as a single issue, or in combination with one or the other, but they have never been addressed all three together (triple related).

(20)

Figure 1.2: Overview of the literature on the 3 interrelated themes

This latter lacuna is highly remarkable once one tries to map the debates within a broader scientific landscape, which is visually done in Figure 1.2. Conditional research on welfare state retrenchment falls within the richest tradition, starting with the classical debate between proponents of the ‘old politics’ (Esping-Andersen, 1990) and ‘new politics (Pierson 1994; 1996) and continuing work by among others Green-Pedersen (2002) and Starke (2006). Publications on health care systems (Jordan, 2011; Hacker, 2004a) and the GFC (Oldani et al., 2013; Benner, 2013) – in particular – are numerous. Literature on welfare state retrenchment and health care systems has mainly been written prior to the crisis but is not systemic as it is mostly country based (Appleby, 1999; Freeman & Moran, 2000; Hacker, 2004b; Giaimo, 2002). Studies on welfare state retrenchment (as general theory) and the GFC have dominated the field in recent years (Starke et al., 2013; Van Kersbergen, Vis & Hemerijck, 2014; Zartaloudis, 2014). The same can be said for literature on health care systems and the GFC (Pavolini et al., 2013; Helderman, 2014; Maresso et al., 2015).

(21)

Health care has long been an unstudied aspect of the welfare state (Jordan, 2011) and relatively little attention has been paid to how health care systems have been implicated by the larger shift of the welfare states into the era of retrenchment politics (Hacker, 2004b). Health care has been a significant and notable omission from the broader welfare state literature and particularly the retrenchment debate. It is striking that triple-related qualitative comparative research concerning welfare state retrenchment and the GFC – as this thesis intents to do - has been done in the field of labor market reform (Shahidi, 2014) and pensions (Pavolini et al., 2015) but not on health care. Therefore, this thesis aims to contribute to the welfare state retrenchment literature by addressing an unmistakably visible lacuna. It supposes to fill a gap in the literature by connecting the aspects of welfare state retrenchment, health care systems and the (aftermath of the) GFC. At this time of writing, there is no generally applicable pattern described for the conditions and causes under which health care change occurs in this specific context. This study aims to be the first to undertake a qualitative comparative approach in which all retrenchment conditions – in conjunction with each other – will have their place. This will be done by answering the following research question:

How can we explain patterns of welfare state retrenchment in health care system trajectories of European countries during the aftermath of the recent financial crisis?

(22)

1.4.3 Societal relevance

Health care system retrenchment does not imply a negative thing per se – efficiency cuts, for example, could lead to positive results – but it is negatively perceived by the citizenry and also researchers have focused on the negative side-effects for society. After the start of the Global Financial Crisis, several studies have reported that consequences of the financial downturn on health outcomes emerged due to rising unemployment, health care cuts and social instability. The recession has led to an increase in poor health status, increasing rates of anxiety and depression among the economically vulnerable (De Vogli, 2014; Quaglio et al., 2013). The number of suicides of people under 65 years has grown in the EU since 2007, reversing a steady decrease the years before (Karanikolos et al., 2013). In addition, decreases in accessibility were reported, for example increasing waiting times for hospitals in Slovenia and Sweden, or an acceleration of staff shortages in Bulgaria and Hungary (Dubois & Molinuevo, 2010).

It is not without reason that European Observatory on Health Systems and Policies – a partnership between the WHO, a number of European governments and several stakeholders – has established the Health & Financial Crisis Monitor (HFCM) that tries to collect (scientific) evidence on the impacts of the current crisis on health and health systems. A better understanding of the political decision making process could gain better insights for politicians and policy makers about the consequences for society.

1.5 Research strategy

In order to answer the research question, this thesis will rely on two research strategies. First, a literature review will be written in order to investigate which factors are mentioned as theoretical explanation for welfare state retrenchment. This will be the substantial part of the theoretical framework. Second, I will investigate to what extent these conditions are valid during the aftermath of the recent GFC, to what extent they are related to each other and whether it is possible to determine a pattern that could tell more about the (combination of) conditions under which governments are willing to cut on public health care.

To analyze the relation between the conditions and the outcome, I will use Qualitative Comparative Analysis (QCA) as a research method, which was introduced by social scientist Charles Ragin in 1987. QCA proposes to be a syntheses between qualitative case research and variable oriented quantitative research. The dominating principle is the examination of set-theoretic relationships between causally relevant conditions and a clearly specific outcome. These set-theoretic relationships are then interpreted in terms of necessity and/or sufficiency (Schneider & Wagemann, 2010). When comparing cases, the complexity of the particular cases and the diversity of the phenomenon will be

(23)

preserved (Ragin, 2009). QCA allows the researcher, on the one hand, to explain health care retrenchment by means of the combination of different causal paths, instead of the correlation of every individual factor to the outcome. On the other hand, these causal paths are linked to particular health care systems, which gives more insights in the country-based context of health care system retrenchment (Ragin, 2009).

The conceptual basis of QCA is grounded in three theoretical assumptions (Verweij & Gerrits, 2012; Schneider & Wagemann, 2010). First, it models equifinality, which assumes that multiple paths to a desired outcome may coexist. Secondly, QCA identifies conjunctural causation, which means that single conditions may not display their effects on their own, but only in conjunction with other conditions. Conditions do not exclude each other reciprocally or compete with each other. Thirdly, QCA implies asymmetrical relationships between causal conditions and outcomes, meaning that the presence or absence for the outcome, respectively, may require different explanations. For instance, arguing that developed countries are democratic, states in essence that developed countries are the subset of the set of democratic countries. Then, in QCA, the fact that there are also less-developed democratic countries does not undermine the set-theoretic claim (Ragin, 2009). QCA asks what conditions are necessary or sufficient to produce an outcome. Certain conditions, however, could neither be necessary nor sufficient but still play a role in the outcome. Such conditions, which occur regularly in real life, can also be revealed with CQA and are called INUS-conditions. These are insufficient but necessary parts of a condition which is itself unnecessary but sufficient for the outcome (Verweij & Gerrits, 2012).

QCA can be understood as both a research approach and an analytical technique. It is an approach because the process is ‘iterative’, meaning that both theory and data are in a constant dialogue during the analysis. It is equally an analytical technique, making use of algebra for its goal to discover empirical paths in the data (Schneider & Wagemann, 2010). How could one, then, systematically compares cases without losing sight of complex causality? Ragin’s original version (1987) is called crisp-set QCA (csQCA). At its core, there is a ‘truth table’, i.e. a data matrix that contains all values of the causal conditions and outcomes. These are all assessed in a binary fashion: they are absent (0) or present (1). The truth table shows all possible conditional combinations and their presence in the cases. The calibration of the set memberships, i.e. defining the threshold, is based on prior theoretical knowledge by the researcher, expert judgment and empirical evidence. Special software, based on fuzzy algebra, is then used to minimize the configuration in order to identify factor-relevant combinations for the outcome. These results will then finally be interpreted in order to ask more focused causal questions about the ingredients and mechanisms producing retrenchment.

(24)

Overall, csQCA seems particular suitable for the aims of this research. It allows me to investigate which combination of conditions could lead to health care system retrenchment, taking into account the relation of the conditions to each other and the potential existence of multiple causal paths leading to the same outcome. CQA is developed as a ‘macro-comparative approach’, meaning that research should be done on the macro-level, for instance on economies, countries or other complex social entities (Berg-Schlosser & Quenter, 1996). National health care systems are located at this macro-level. In addition, QCA is best suitable for small-N (2-15 cases) or intermediate-N (15-50 cases) research (Ragin, 2009). The amount of cases in this thesis will fit into this low-N range. At the final stage, to get insight in the causal pathways for health care system retrenchment, I will try to develop a new umbrella theory that could answer the question of why governments opt for health care system retrenchment at a certain moment.

1.6 Outline

This chapter has provided the reader with an initial introduction to the topic, the research strategy and the questions that are central in this research. In the next section, the theoretical framework will be revealed by means of an extensive literature study on welfare state retrenchment. Subsequently, the following chapter will declare the methodological fundamentals of the research and I will operationalize the relevant conditions. Building on that, in the subsequent section the empirical data will be gathered and the contextual case studies about national health care systems will be written. Next, I will be ready to present the results, give them interpretation and conclude with an overall assessment of the research question. Finally, the last chapter will reflect on the research process and the thesis’ contribution to the theory, as well as to provide some recommendations for future research.

(25)

Chapter 2: Theoretical Framework

2.1 Introduction

Since Margaret Thatcher (1979) and Ronald Reagan (1981) came into office in respectively the United Kingdom and the United States, the notion of ‘retrenchment’ entered the field of comparative welfare research. Although both leaders were willing to radically cut on the welfare state, at the end of the decade it became clear that they had not fulfilled their commitments. This wisdom led scholars to look at the underlying dynamics of dismantling the welfare state, which became the starting point of what is generally known as the ‘welfare state retrenchment debate’. Research within the field has provided a vast variety of approaches, concepts and mechanisms to explain why and how welfare state retrenchment is taking place.

For this chapter, I will work on the construction of the theoretical framework that enables the researcher to explain why retrenchment could occur. Both qualitative and quantitative studies have emphasized multiple factors that are significant predictors of welfare state retrenchment. It is generally assumed that we can distinguish between four theoretical frameworks in the literature that have explanatory power to picture retrenchment from their particular perspective (Swank, 2001; Starke, 2006). These are political conflict theory, institutionalism, neofunctionalism and social constructivism. Under these ‘building-blocks’, I will address the particular theoretical discussions on the dynamics of retrenchments, as well as the hypotheses they contain. It is important to note that one should perceive the different factors named in the literature as complementary to each other. In this sense, Qualitative Comparative Analysis (QCA) is the appropriate way to study them in combination.

Because this study is about welfare state retrenchment at a specific period in time, namely during a financial crisis, it is vital to get an insight to what extent these factors still yield explanatory power during economic hardship (Shahidi, 2014). Therefore, each perspective is supplemented with a consideration on this relationship. This will enable me to map the processes between economic shock and eventual changing policy as output. Further, it will help to focus on the really important factors in the specific context of this research. Finally, at the end of this chapter, an overview of all factors named in the literature for welfare state retrenchment is given.

(26)

2.2 Political Conflict Theory

2.2.1 Literature

The first explanation for welfare state retrenchment is given by a group of theories that emphasize social policy as an outcome of political struggles about distributive decisions. These structural studies have generally focused on the impact of partisanship on retrenchment, along with the influence of class-conflict in the political arena.

The origins of this first theoretical approach can unmistakably be traced back to studies that were ground-breaking in explaining the expansion of the welfare state. Most prominent in this debate is the ‘power resources theory’, claiming that the welfare state is part of a distributional struggle in the capitalist society between the working class and the bourgeoisie (Castles, 1978; Korpi, 1983). Due to the introduction of universal suffrage and the rise of the working class, social democratic parties and trade unions took advantage and were able to implement welfare legislation. The welfare state, here, is seen as an opportunity for the working class to protect themselves from the unequal distribution of wealth in society. Korpi (1989) argued, for instance, that welfare state variation can be explained by the extent of social democratic participation in the government.

The debate has been extended to more complicated accounts of this argument, centralizing on this deeply-rooted class-conflict. Heavenly influenced by the power resources arguments, Esping-Andersen (1990) famously distinguished between ‘three types of welfare state’, that can be explained by the relative strength of political parties in a country. He also showed that, despite the strength of labor parties that could support the social democrats, left parties still had to connect with other parties. Variation in welfare states could therefore be explained by the capability of left parties to form alliances with the middle-class. Similarly, Van Kersbergen (2003) argued that left parties have done well in cooperating with Christian democrats. These parties are supportive for the extension of the welfare state as well, however, policies should be less redistributive and more transfer- and service-oriented.

It may seem a little outlandish that these aforementioned expansion theories are noted here for retrenchment. However, it is not. Until the 90’s, there was a widespread consensus that the insights from the welfare state enlargement theory could also be applied to retrenchment. As Esping-Andersen (1990) strikingly wrote: ‘A theory that seeks to explain welfare state growth should also be able to understand its retrenchment or decline … [T]he class coalitions in which the three welfare-state regime-types were founded, explain not only their past evolution but also their future

(27)

prospects’ (p. 32-33). This assumption, however, came under severe attack by Paul Pierson (1996), who argued that retrenchment is ‘not simply the mirror image of welfare state expansion’ (p. 151). According to him, the logic of retrenchment is fundamentally different, introducing an age of ‘new politics’. This is because ‘the interests surrounding retrenchment are asymmetrically distributed to the disadvantage of retrenchment’ and ‘the negative bias of voters makes retrenchment politically dangerous’ (p.144-147) (see § 2.3 for more on Pierson’s institutional approach).

The contradiction between ‘old politics’ and ‘new politics’ has become the most disputed issue in the literature. Although it was believed by many authors that the ‘end of partisanship’ had arrived, some other authors disagreed with Pierson and tried to show that the partisan effect still mattered. Korpi and Palme (2003) and Allan and Scruggs (2004) have demonstrated empirically that the party color of the government remains an important indicator for the extent of which benefits are provided through social insurance schemes, with leftish politicians being more generous. Ross (2000b) has argued that the partisan logic works, however, counterintuitive. She claims that a ‘Nixon-goes-to-China logic’ is at work: left parties can reform the welfare state without being seriously punished by the electorate, this gives them more ‘degrees of freedom’ for reform. Left parties undertake retrenchment therefore more out of necessity than due to ideological motivations. This line of reasoning is consistent with Green-Pedersen’s (2002), who argues that unpopular retrenchment for leftist governments may be more acceptable to the voters because they enjoy more credibility in protecting the system than right-wing market reformers. This group of scholars assume that leftist governments are better at pursuing cuts.

Kitschelt (2001) has argued that not partisanship itself but the structure of the party competition influences incentives for political parties to pursue retrenchment politics. The key to this claim is that parties in office will only decide to roll back the provisions if they do not face competition of a creditable1_{anti-retrenchment opponent. If voters can punish a pro-retrenchment government by} defect to an anti-retrenchment party, the electoral risks will be (too) high. Contrarily, these risks will be lower when there is no credible alternative to vote for. According to Kitschelt (2001), retrenchment is not the least likely in countries with the strongest left party, but in a party system in which parties mainly compete over socio-cultural issues and where the main parties are established defenders of the welfare state fearing no opposition of an anti-retrenchment party.

Another explanation is offered by Green-Pedersen (2001) who emphasizes the spatial configuration of party competition. Borrowing from coalition theory, he distinguishes between bloc systems and pivot systems. Governments will only retrench if they can achieve a party consensus around

1 Whether a party is large enough to form a government, either alone or with other parties.

(28)

retrenchment. In bloc systems, there are two competing blocs of parties, for example the social democratic and bourgeois blocs in Denmark. When the right-wing bloc holds governmental power, they will be hindered by the left-wing. In this system, party consensus will therefore only emerge if the left-wing bloc is in power, as the right-wing opposition will not oppose welfare state reforms. In a pivot system, conversely, there is a strong center party that can cooperate with either side. The likelihood of retrenchment increases here, because the presence of the middle party forces other parties to move to the center. Retrenchment debates in this system are less polarized, and the blame of reform is more equally shared by political competitors. Green-Pedersen (2001) argues therefore that ‘unless only the left-wing bloc governs in a bloc system, more retrenchment is expected to take place in pivot systems of party competition than in bloc systems’ (p. 986).

2.2.2 Political Conflict Theory and the impact of economic hard times

An important question to answer is whether – or to what extent – the political conflict arguments can be applied during times of economic crises. Interestingly, there is very little on this particular issue in the literature and the results are mixed, at best.

Some theoretical arguments speak in favor of a partisan effect on crises responses. Scharpf (1991) has argued that economic crises are ‘moments of truth’ for left and right-wing party positions on the role of the state and the market within welfare provisions. These ideological differences will be more visible than ever before, resulting in the continuation of left- and right party positions. In addition, Cusack et al. (2008) have studied policy responses to short-term (economic) shocks in sixteen OECD countries from 1960 to 1995 and found a partisan effect on total governmental transfers.

Contrarily, other studies have argued against a partisan effect, claiming that the ability to create ideologically driven policies during hard times is limited by electoral constraints. During crises, one would expect a postponement of ideological differences, as policy makers will serve the benefit of the common good. During this ‘rally round the flag’, a centralization of decision-making or a technocratic government will be the rule (Starke et al., 2013). Lipsmeyer (2011) hypothesized that we should see a salience of partisan politics during times of permanent austerity, as his research found that economic problems diminish the partisan effect. It is further claimed that right-parties even soften their policies in the aftermath of economic hard times because a sizeable proportion of their electorate will be affected by the shock (Jensen, 2011).

I suppose, however, that there are good theoretical reasons to assume that the aforementioned political conflict theories even during economic hard times remain a powerful explanation for retrenchment. First, as earlier mentioned, the welfare state is often seen as a distributional struggle in the capitalist society (Castles, 1978; Korpi, 1983). Due to its redistributive character, low-income

(29)

households will be more affected during periods of crisis in comparison to other groups (Ahrend et al., 2011). It is therefore to expect that parties in the first place will respond in a way following their original constituencies of low- and high wage earners. Secondly, party responses are expected to be in agreement with the parties’ classical ideological positions. There are strong motivations to cherish their good reputation vis-à-vis their electorate, so they are expected to fall back on their established habits. Left-parties will be more eager to blame the market as a cause of the crisis. Contrarily, right-parties will prefer introducing retrenchment as a way to increase the competiveness of the private market sectors and will generally blame the size of the state (Amable et al., 2006).

2.3 Institutionalism

2.3.1 Literature

A second approach concerning welfare state retrenchment is given by a group of researchers that have emphasized the institutional configuration of the political system as powerful in explaining reform. This institutionalist scholarly debate intensified after Paul Pierson’s publications (1994; 1996) on the ‘new politics’ of the welfare state in which he claimed that politicians in this new era of austerity have switched from ‘credit claiming’ for popular programs to ‘blame avoidance’ for unpopular policies. In line with Pierson’s contributions, institutional approaches have focused on macro- and micro institutional settings.

Pierson (1994; 1996) first and foremost questions the socio-economic class approach that had long been a powerful explanation for welfare state expansion ánd retrenchment. Employing three lines of argument, he tries to explain why the logic of retrenchment is fundamentally different compared to expansion. First, psychological studies have shown that people are risk averse. They will respond more strongly to potential losses than to potential gains. So even when retrenchment will be compensated (e.g. through lower taxes), citizens will resist to program cuts. Surveys show that the welfare state is still highly popular, so retrenchment is assumed to be electorally dangerous.

Secondly, the previous expansion has created a huge number of interest groups trying to protect the specific welfare state beneficiaries. The ‘logic of collective action’ is here at work. While the costs of retrenchment are concentrated within certain groups, the benefits are diffused throughout society. As a result, opponents of retrenchment – often interest groups – have much more incentives to mobilize their grassroots than retrenchment supporters. This new landscape of activism has forced retrenchment advocates to no longer cut back provisions because they are just more numerous in the political arena. Only when political responsibility can be masked, or when voters are not able to punish them, cutbacks could take place (Pierson, 1994; 1996).

(30)

The third and probably most important reason that makes retrenchment so difficult, according to Pierson (1994; 1996), is that the welfare state institutions are interwoven with path dependency. It is said that ‘history matters’, meaning that past policy shapes present politics in a path-dependent fashion. So what politicians do today is shaped by what has been done (by their predecessors) in the past. This is what Pierson calls ‘lock-in’: past policy commitments often narrow present options. In addition, the contemporary configuration of the welfare state determines the availability of information politicians could count on. As Hay and Wincott (1998) have argued, resources such as information and capital are unequally distributed benefiting those who prefer the status quo. A new minister may have grand plans for reforms, but a lack of strategic knowledge will limit his possibilities for change.

Although it should be clear that Pierson thinks retrenchment is pretty difficult in practice, the ‘new politics’ thesis is not simply claiming that it is impossible. In his attempt to specify under what circumstances retrenchment may occur, Pierson argues that the major objective of governments in this new era is seeking to avoid responsibility, as reforms are highly unpopular. Budget-cutters therefore will try to confuse and divide the (political and societal) opposition through three main strategies of ‘blame avoidance’. Strategies of obfuscation tend to reduce the traceability, that is, to blur the line between the actual retrenchment and the policies that caused it. When reacting on policy change, voters have to construct a causal chain between governmental action and the positive or negative effects it has on themselves. Obfuscation aims to make it more difficult to construct these chains. A classic example is transferring responsibilities to the lower political level without adequate funding, so that the locals are forced to enact cuts. Another strategy that can be used is division. It aims to divide the political opponents by splitting them, which may reduce the political costs of retrenchment. A united opposition is of course harder to fight. The last strategy Pierson emphasizes is compensation. Side-payments can be used to mobilize the political opponents. The idea is to offer something positive to the victims of retrenchment. The use of these three strategies may make retrenchment more likely to occur, but it should be noted that not all of them are always available to policymakers (Pierson, 1994).

In line with Pierson, authors have found supportive evidence for the blame avoidance thesis ( Cox, 2001). Other comparative studies have broadened its scope by looking at the institutional fragmentation of the political system. The starting point for this group of scholar is the notion of veto points. A veto point refers to a situation in which a certain political actor can block a political decision (Immergut, 1992). These actors, called ‘veto-players’, must agree for legislation to be passed (Tsebelis, 1995). It is generally argued that as the number of veto points increase, a larger number of veto-players will favor the status quo. In other words: countries with the least institutional hurdles (or

(31)

veto-points), and therefore the highest degree of power concentration, should display the highest degree of welfare state reform (Vis, 2010).

From an institutional perspective, a first line of reasoning has therefore argued that in political systems with a high degree of horizontal and vertical fragmentation – for instance the United States and Switzerland – retrenchment is less likely due to the high amount of veto-players. Contrarily, in systems with a high degree of power concentration, like the United Kingdom, retrenchment is more likely (Swank, 2001; Van Kersbergen, 2006). In contrast to these arguments, a second line of reasoning has claimed for a reverse relationship: the higher the level of power concentration, the lower the degree of welfare state retrenchment (Ross, 1997; Starke, 2006). The argument here is that systems that concentrate political power also concentrate political accountability, so that the blame avoidance strategy is more difficult to apply and retrenchment therefore will be less likely. Conversely, in fragmented political systems it will be easier to avoid blame, and retrenchment will subsequently be more likely. These reverse claims obviously seem inconsistent. As Jensen and Mortensen (2014) put it: ‘the different effects […] neutralize each other or only work in combination’ (p.146).

Another group of institutional theoretical arguments has focused on the political institutions at the macro-level. In their famous article, Huber, Ragin and Stephens (1993) have empirically shown the existence of a strong effect of constitutional structures on the welfare state. Both presidentialism, bicameralism, the presence of judicial review and the use of poplar referenda constitute potential veto points by which retrenchment is more likely to be blocked. In addition, Crepaz (1996) has distinguished between Lijphart’s prominent typology of ‘consensus’- and ‘majoritarian’ democracies. While the former system tends to favor inclusion and power-sharing, the latter system is characterized by power-concentration and an antagonistic-style of politics. It is claimed that policies in a consensus democracy are the result of long processes of compromising. From the moment all actors have agreed on it, it will be difficult to reform certain social rights that have been given to the citizens. In other words, path-dependency is more likely in a consensus democracy, resulting in a decreasing likelihood of retrenchment. Conversely, a majoritarian democracy should experience more retrenchment.

A last group of theorists has emphasized the importance of micro political institutions, that is, the existing structure of the welfare state institutions and its schemes. In particular, these authors underline ‘policy feedback’ – the consequences of previously introduced welfare state programs – to be relevant. The starting point for this line of reasoning is Esping-Andersen’s (1990) well-known ‘types of welfare state’ classification, which assumes that one can order different country schemes on the basis of matching institutional characteristics. Swank (2001) has therefore argued that

Health Care System Retrenchment